The Complete Ai Prompt Engineering Masterclass for Beginners

1. Introduction to Prompt Engineering Masterclass: Hi. Welcome to prompt Engineering mastery course. Myself, Shake Sepul and I am a flance a prompt engineer with the past one year of experience in art. I also worked in the SOD AI company for the outlaer client as a prompt engineer and also I am app developer. In this course, we are going to learn what is actual prompt engineering is. Okay. So as we know, this is the new era of AI in which we see, there are a lot more AI models out there like ChagptGroq AI, Cloud gem.ai, deep Sik like that. There are a lot more AI models in the future or right now. In this AI wall we need to know how to use these prompt patterns prompt engineering for LLMs. How we can get the most out from EI, like getting the best output from EI. For that, we can use this prompt engineering skill. As we know, the AI models will use each and every industry will use the AI models in the future or right now because it is very important for us because in the case of automation or getting the content from it, because the AI models all trained by different traned by large amounts of data in which we can save the time. So in this case, in future or right now, every industry like education, marketing, businesses, all those companies or industries are looking to transform the whole organization with AI in that these ALLL models can help. For this, we need to know how to use these AI models in effective manner to get the best output from AI, but that this prompt engineering skill will come into picture. I hope you understand these points. But thus, we need to know how to use this prompt engineering skill in our daily life and professional life because AI models are everywhere. So for that. In this particular course, we are going to explore nine different AI models, like like HAGPT gem.ai, Cloud purples.ai, Microsoft Copt deep Sik Krog in chat AI, and mist AI. Not only do, we will see what is actual prompt engineering is, we will go from basic to advance. We will explain and I will explain our basic components of prompt, how to write the prompts, what is the actual formula to write the best prompt, we will explore more than ten different prom patterns, advance level prom patterns in which you can use this particular prom patterns to automate in chat in the chat as well. You can write here and you can do whatever you want. There is more curious. You can see this, all those things in the upcoming classes and sessions. I'm very excited to share my learnings, my experience with you in this course as a prompt engineer. Not only that, we will also explore how to use these nine different AI models to write the best AI prompts for us. Okay, not only for task writing the content, email copies, all those things, we can also use these prompt patterns, AI LLM models to write the best prompt for us. That is very interesting. We will explore those things also and we will explore ChagPT in depth with the open EI platform, playground, we have, and we will explore some opportunities as a prompt engineer, what we have after this course, how we can find the jobs for it, fancing projects, flancing gigs or more. After this course, you can also explain what is the role of prompt engineer in the Gen AI, the future scale, all those things in this course. This course is going to be amazing because after this course, you will unlock your mind by chatting with EI. I'm not only just tell you to writing the how to write the prompts for use cases, not only that, but I am unlocking the creativeness or effect of or potential of AI modules in which it can help you to build something. That is simple. The main purpose of this course is stat line. Not only that, this course is made based upon the company's requirement skills to use a prompt engineering. Not a technical bit, but using the AI models. Every industry or companies are looking for the prompt engineers who have their specific skill set, like crafting the best prompts for different AILM models, and evaluation of the output of the output from AI models, testing out and RITA rag how we can see are to generate the best output from EI for the particular task to test to evaluate the output and to check which LLM will help to solve the particular task to evaluate output and all those things. I have explained all those things in this particular course step by step, how to write the best proms for use cases to different LLMs, how to test ALL modules. I have explained all those things nine different AI models with use cases, how to test each and every LLM model. To choose the specific task specific LLM for our specific task. I've also explained you different capabilities and functionalities of nine LLM models. Not only that, I also explain you much more. I'm looking to share much more things with you, but that I created this course. After this course, trust me, you will get the hands and experience in that. So for each and every model, this course is divided into the six models. For each and every model, you will get the resources and assessment. After this whole course, you will get the full document course in which you can I have written all some step by step, each and everything that I explain in these videos. You can get access after this course or in the last. That is all over this course. Okay. There is much more to share with you, but you can go and watch the upcoming sessions and classes, you will clear it. I hope you understand this course, my points, not only that, remember one thing. Remember one thing. I have explained in slowly manner in the all videos. Even the beginner can also understand each and every thing. So for that. If you know already about prompt engineering, if you're comfortable, so you can change your speed to two X like that past or for better understanding. I hope you understand these points. Follow each and every video, don't skip each and everything. Not only that after this course, you will get the full course document in which you can get the insights, what I have explained in the videos, all those things. In this course document, you will get all the basic to advance knowledge. Good explanation with examples, insights, great. I have also explained all those things in this document also. For your reference, you can check it out after seeing the videos. Okay? So basically these course have divided into the six models in which you can get the resources and assignment for each model. Okay. After that, after completing all the recorded videos of this course, you will get the final project as well. You will get access to this particular document here that you are looking here. Okay. So I hope you understand these points. So make sure you get this document after the recorded videos for better understanding. Let's not that. I remember one thing. This course and final project of this particular prompt engineering course is designed based upon the company requirements of. Basically, this course is created based upon the company's requirement, job requirements as a prompt engineer. I'm confident after you follow all the lecturers perfectly and practice well that I can tell you you can ready for the applying for the jobs for the I prompt engineering. Without a technical part. Then this writing the prompts form. I hope you understand these points. After completing the all the recorded videos, please go refer to the final project and please practice with this for different applications. Please follow all I have given the steps. These steps giving these steps are based upon the company's requirements. If you practice well, all these steps, and completing this final project, you will get the better understanding about how to write the best prompts and how to evaluate and how to compare and how to optimize it. All those things in this particular course. I hope you understand all these points, good luck. Go each and every lecturer, practice well and please couple with all the assignments and take help with the resources and full course document also. Let's start with model number one that is introduction to prompt engineering. Let's dive into that. 2. 1.1 Basics of Prompt Engineering?: Now in this model number one, we are going to discuss some basics and foundations of prompt engineering. So we will cover what is an prompt engineering, the difference between prompt designing and prompt engineering and why the prompt engineering is important. Let's start from the scratch. Now, what is actually prompt engineering is? When we use the AI models like Cha JBT Cloud, perplexi dot I, and other several AI modules for our requirements, we will just write and in the chat container, right? So that is called an prompt but what is the actual meaning of prompt engineering is? So writing the precise instructions for AI language modulus to get the best output from AI that is called prompt engineering. You can see the definition of prompt engineering that is crafting precise instructions for EI language models. So it can the ChaiPCloud or other EI moduls as well. So why we are writing the precise instructions here? So the key purpose is to improve the quality and relevance of EI response. For the best understanding, you can see the example, like, think of it as a given clear instructions to a skilled assistant. The concept of prompt engineering is the science behind writing the prompt, the art of writing the question or writing the instructions or asking the ion in the way of the language model can give the better output according to our instructions that is called prompt engineering, right? So the crafting the precise specific instructions for AI language mods to improve the quality and relevance of AI responses is called prompt engineering. Okay, so for the better understanding, we will jump into the difference between prom design and prompt engineering. You can see what isn't prompt designing and prompt engineering is. So as I said, you can see the prom design is nothing, but it is a general instruction or query that we will ask to AI to get the answer from it. For example, you can see the write a poem about nature. It is a general instruction or e you will ask to Cha Gi PDA or other AL language models to get the answer from it. Okay? That is simple, that is call and prompt design. When come to prompt engineering, it is a detailed instructions tailored for specific application. So that is call and prompt engineering. So when it come to prompt designing, it is a general question you can ask in any way in any query. But in the prompt engineering, the set of instructions written by detail for the specific application, like you can see the example here, compose a rhyming poem about nature in the style of William Word's word. You can see it is a simple quotien. You can see the composer rhyming poem about nature. Up to here, it is a simple prompt designing. But when you see here in the style of William Words word, it will become in the specific application. So the advantage of this using prompt engineering is you are getting the information from AI language model that you are looking for. That is simple. So because if you write this one, it is a simple question, write a poem about nature. But if you write the specific instructions that you are looking to get the information from AI, you need to add more specificity in the prompt. That is called prompt engineering. You can see detailed instructions tailored for specific application. You can see compose a rhyming poem about nature. Up to this, it is a prompt designing, but it will when this prompt is converted into the prompt engineering, you can see we have just include the specificity and application focus in this prompt designing. That is in the style of William Wordsword. Now, the El language model will produce the output that is rhyming poem about nature in the style of William Words word. The key difference is the specificity and application focus. The prom designing is a general purpose. And the prompt engineering is the specific and the application focus, which we will try for the specific application to get the specific output and highly relevant AI response from the AI language models. I understand these points very well. Now, let's jump into why is prompt engineering is important. That is simple. So when we discussed earlier, that is prompt designing, it is a general purpose, right? We need to learn this prompt engineering because get the relevant and accurate air responses without any hallucinations inaccuracies in the output and to build the efficient workflows. By just writing the specific instructions for the specific application, you can build the whole efficient workflow in the Cha GPT itself. So I will show all those things in upcoming sessions, and you can see you can by writing the prom by iterating again and again, by changing the input the prompt. And by analyzing the output, you can change the prompt according to your needs and you can enhance outputs for complex task. That is why learning the prompt engineering is important. Now, where this prompt engineering is using. You can see there are a lot of more industries and education, health care, content creation, or programming or other as well because the AI is evolving around the world in each and every industry, but where the AI language models are used then the prompt engineering, it needs to be done there because the prompt engineer will know how to get the actual AI response from AI language model. So for that you need to learn this prompt engineering right now so to get the efficient output from the AI language model like ChargePD and other. Now in the next session, we will discuss the basics of language models. Let's dive into that. 3. 1.2 Basics of AI Large Language Models (LLM's): Now, in this session, we are going to see the basics of language models that are LLMs. So what are LLMs? You can see the AI systems trained on large dataset to understand and generate human test is called an LLM. Okay, you can see the examples like m.ai, Chat GPT Cloud, perplexity.ai, and the image generation tools that is mid journey, Leonardo AI, and other language models as well. So the basic word that is LLM. So what is called an LLM. It is a large language model which trained on large dataset to understand and generate mainl test. That is simple. Okay? When you write the quien in the ChargPT, actually the chargBT is called as LLM. The LLM means understand and generative text based upon the trained data. If you see the hagibt can answer any, right? So because the hagiPT or other Air language models are trained by large dataset to understand the human quis input or prompt, and it will generate the text. When you use the hagiPT when you ask any question, it will generate the answer like human is writing. Okay. When you try by yourself, just see the way you are asking the question, the answer is also like that because it is analyzing your input. It is understanding your prompt, your thoughts like that. Although the LLMs are not will recognize your thoughts, but it will understand your real intent in the prompt and it will understand your text. What is the main motive behind you that you are looking to get the information from AI? It will easily understand, analyze, and it will generate the response human text. Okay? How do they work? You can see the ChagpD or other LLMs first analyze the input that it can be your prompt and it will recognize patterns. Okay, the AI systems are trained on large datasets in the patterns way. Okay, in the patterns way, the LLMs are trained. I will recognize the patterns and will generate the output. Okay, we will cover so many advance from patterns in upcoming sessions, which will blow your mind in using the AI language models, and we'll learn all those things in upcoming session. How the LNS process prompts. You can see we have already seen that is AI as a chef prompts as a recipe. Okay, that is simple as example, you can take it as I said, the prompt engineering is nothing but crafting the precise instructions for the specific application. Okay? You can take any specific application like food assistant or health assistant. It is a specific application for us. So now in this one let's see how the LLM process prompts. You can see AI as a chef. When you write the instructions that you are a advanced or you have the experience as a chef in so and so food, when you write this, the LLM process like, so I am a chef and I need to assist the user in creating the amazing recipes based upon their requirement like that. The AI will think like that. Okay, that is, you can see the analogy. AI as a chef prompts as recipes. Okay. Now, what happens here? Now when you write the instructions, precise instructions for AI language model by using prompt techniques. So we will cover all those things like personal prom pattern. It is a personal prom pattern. It will work like AI as achefPms as recipes. Okay? When you write the prompt, it will work under the chef scaled assistant like that only. When you write the precise instructions, the LLM will process Amche and I need to assist to creating the amazing recipe. Okay, that is simple. You can see. I'll talk about the difference between the good proms and bad prompts. What are the good proms and bad prompts. You can see specific and detailed instructions. So the prompt engineering is nothing but writing the specific and detailed instructions to build and specific application. You can see the good prompt contains specificity and detailed instructions, and the bad prompt contains the value and ambiges. Let's see there's some examples of good prompt and bad prompt. You can see bad prompt that is explained climate change. As I said, it is a simple question right. So it is a prompt designing, not a prompt engineering. Sit comes under the bad prompt. Now when it come to good prom, you can see explain the causes and effects of climate change in simple terms, suitable for a 10-year-old. So what is the difference between that? So we include the specificity and detailed instructions. That is all about prompt engineering is. Okay, why it matters? Because clear proms produce better tailored outputs. Okay? So we need to write the clear proms, detailed instructions. We need to add the specificity, then we will get the best output. Okay? No. Let's talk why do proms fail? As I said, when the proms doesn't include the specificity and clarity, the AI will fall to give the best output. Okay, now you can see there's some common issues, missing clarity or intent, lack of context provided, no background provided. So why isn't background provided? As you can see, so we'll talk more in depth about this context management and over complexity and all those things in upcoming session, see here some of the common issues like missing clarity or intent, lack of context, no background provided. Over complexity, overloading the prompt with unnecessary details. So we need to write the prompt in a way that it should be very clear, very intent, and it should contain some background information, context, and it should be specifically focus. Then you will get the best output from AI very well. The best solution for these common issues, you can see refined prompts to be clear, specific and concise. That is simple, we will see all those things writing techniques, prompt pattern methods, all those things in upcoming session. That is very easy. Now let's tap into some applications of a prompt engineering. So as we earlier discussed the where the AI is implemented. So definitely a prompt engineer should be there because the prompt engineer know how to get the effective and accurate information from the AI by writing the effective prompts in the EI language models. So it can be the healthcare industry, education industry, content creation industry, programming industry, or automating reports, summarizing articles. So all those things you can do in the Chachi Bit as well by providing all those things and other EI models as well. You can see the in education adapt to learning tool health care, patient communication, content creation. You can write blogs, marketing materials. So there are a lot of more applications of prompt engineering. When you practice, know the importance of this prompt engineering in this AI era. So these are some applications of prompt engineering. Let's jump into our second model in which we will start basic off prom. That's tie into. 4. 2.1 Basic Components of Prompt: Come to our next model number two of this master prompt engineering course. So in this model, we are going to see some foundations of writing effective proms and what are the key components that we have to keep in mind while writing the prompt, and we'll see why these key components plays a major role in writing effective proms. So let's understand some basics and foundation of writing effective proms. So let's see, we will explore some components of a prom. We have three components of a prompt. That is clarity, number two is context and number three, specificity. These three components are very important while we have to keep in our mind while writing the prompt. So let's one is clarity. So clarity means writing simple and direct sentences which have some clear intention of you. So the AI module will generate a best output when it is easy to understand your intent, right? So you can see the example here. You can see the example here. So what is the clarity means? It is a straightforward direct sentence. Okay? Which clear intent? You can see the example. Tell me something interesting about space. So AI will think what you need. You are asking a broader question. There is no specificity in your question. There is no clear intent. Okay? Tell me something interesting about space. Okay? It will simply throw some interesting points about space, okay? If you try to give some clear instructions like what are some recent discoveries about Black house you can see here, you have clear intent. You need some discoveries about black hose. Black hose means you are focusing on specific topic in the space in which you have clear intent. It is a clear prompt, so that AI will understand, Okay. Need some recent discoveries about black holes, so it will give the best output for your prompt. So well, compared to this Ill me something interesting about space, it will just throw some interesting about space. There is no specificity in that. There is no clear, there is no clarity in that. Okay, this is some clarity points, so we have to keep in that. While writing the prompt, you should keep in mind that clarity plays a major role in that. You have to the instruction to a model like you are asking for a specific topic in which you have direct clarity in your mindset while writing the prompt. So let's see second one. That is context. So what is a context means, you will provide enough background information to support your main intent. Okay, so let's see there. It can be done by setting the stage by describing the scenario or defining the role of in which role EI want to be act like that. Okay, so simply, you have to provide enough background information to the EI model to understand the task and actual intent of your. Let's see the example here. So you are a science teacher explaining gravity to a 10-year-old student. Okay, if you remove this, you can just write explain gravity to a 10-year-old student. If you write prompt simply like this, explain gravity to a 10-year-old student. It will just explain gravity like how 10-year-old student can be understood. It will just explain the gravity. There is no background information there. There is no to get the output very specifically and accurately. Because AI is trained by large amounts of data, it can simply throw with other words which are not in the part of gravity. When you provide a background information, like you are a science teacher. So you are providing here some background information in which the EI will think it is like a science teacher. AI will think, I'm a science teacher. I have to explain gravity to a ten year tool. It's by this, the AI will generate a best output when compared to writing simply explain gravit to a 10-year-old student. You can analyze these two outputs by you yourself, simply writing this first prompt like explain gravitude ten year stone and other for you science teacher, this whole prompt in any language model like hag BA you can see, and you can analyze the output and you can define the difference in between that. So context is all plays a major role. Next, our specificity. So specificity means we have already learned about what is a prompt engineering. Prompt engineering means writing the instructions for a specific application, right? So specificity means precise means just write what you want and get from the AI model. So you can see this here. Be precise about what you are asking. The more detail you are, the more relevant responsibility. The AI model should understand you are maintained so far that you have to give more detail of your problem or what you want from AI. So you cannot write a simple question or answer. So to get the most of the AI models, you have to give the detail to get the best out of from AI. Can see the example here instead of saying, write a story. Okay? It is a simple question, right? There is no reasoning or there is no enough detail to understand AI. You can see if you go and ask AI, like, write a story, the AI will think, Okay, I will write a story, but in which type, in which tone, in which topic, I have to generate a story. It cannot define like I will simply write a random story, random words that can be not relevant. You can see the example here. Write a 300 Wat science fiction story set on Mars where the protagonist discovers water. Sorry, this protagonist discovers water. In this prompt, you are giving more detail that what you want. Okay, you have given the 301, 300 was science fiction story described here what story I want and which topic I want. It is enough for the AI. So the AL think, Okay, I need to generate this fiction story on Mars. Simply generate a specific story your prompt. So that's why specificity plays a major role in writing the prompt. Okay, let's see why this compent matters. So as we discussed the three components, why this components matter means simply can see this summarization When your prompt is clear, the model avoids confusion. That's good. So when you write the prompt clearly, in any language, you will ask to model. So it will understand your intent and it will generate a best output for your prompt, which have clarity in your mind and in prompt also context. Context means it helps you understand your intent. The EI model will understand your intent and purpose as well as task and generate best output according to your prompt instructions. And specificity means it will reduce irrelevant or off topic response in which you will provide the more detail that what you want in specific way, which reduces the irrelevant response in it. So that's why this component matters a lot in writing the FA two prompt. Move to another lesson of this model in which we will see some types of proms and let's dive into the next lesson. 5. 2.2 Types of Prompts: Come to our next lesson of this model number two in which we are going to learn some different types of proms we have. There are three types of proms like instructional prompts, open ended versus closed ended proms and multi conversional proms. First two proms are simply basics. When compared to the third one, that is multi conversional proms. We have so many advanced prom patterns that we discuss in upcoming model classes. So let's see the first one. First one is instructional proms. This prompts is actually simple questions, queries or instructions that you will ask AI model to generate a specific concept. Okay, you can see the example here, list five healthy snacks for children. And explain why they are healthy. This is a simple question you will ask to I model to get the answer. It is simple, right? Simple asking question to I model is called an instructional prompt. Writing a question or query or instruction is also called as prompt. When the instructional proms will work better when you need structured or factual or step by step answers. Next prompt types we have that is open ended versus closed ended prompts. Can see here open ended prompts means encouraging creativity and longer response. The prompt which challenge our model to think and to generate output which have more information. Like you can see the example here, what do you think are the benefits of renewable energy? Because in open ended, you are writing the prompt like the EI should think. You can see the example here. What do you think are the benefits of renewable energy? And it will give the best output, longer responses for this prompt. While come back to close ended prompts, so it will simple for specific answers like what is the capital city of India, the answer will be the deli. It is a simple question and getting the specific answer. That is close ended. There is no thinking, creativity, and there is no longer response or step by step, anything like that. So closed endemic simple asking question is called close ended Poms. When compared to open ended, open ended proms, which encourages the creativity in AI and which generate the longer response. Let's see that third one, which is very important in prompt engineering. Multi ton conversational proms. You have earlier see these two types of prompts. There is no reasoning in it. It is a simple writing quotient or instructions and getting the answers from it. But when compared to multi ton conversational proms, it has some refining power process, refining, output analysing and much more in the multi ton conversational proms. We will explore more advanced prompt attends under these multi ton conversational proms in upcoming model classes. So don't worry, we will cover all those stuff. So let's see here. Sometimes you need to have a conversation with, for example, you will write a prompt. That is fast prompt like. Tell me about renewable energy. It will generate some renv energy information. After that, you will ask a follow up question which is related to previous prompt. That is generated by AI. Sorry, that is output. Okay, first you will write it tell me about renewable energy. After that, AI will generate about renewable energy information. After that, you will ask some follow up question based on output of previous prompt. In this case, tell me about renewable energy. You can see the follow up prompt here. Can you explain the environmental benefits of wind energy in more detail. It is a follow up question. After that, you can have many follow up questions. You can write a third follow up question, fourth, fifth, as many as you want. Once we will write some text, they will written text. So after that, we will ask some follow up question or like that. This build a dialog, which is useful in chatbards are multi step tasks. So you can see the chat board like hat GPT, other AIA language models or like these multi ten conversational prompts. You will ask a follow up qui or other question in same patterns like that. These are easy multi conversational prompts. Let's move to our next lesson of model number two, that is basic prom patterns in which we will use hagiBT to understand the different types of basic prom patterns. Let's dive into the different types of basic proms. Let's 6. 2.3.1 Basic Prompt Patterns : 1. Zero-shot Prompting: Guys, welcome to our third lesson of this model number two, and we will see some basic prompt patterns that we have right now. These are some basic prompt patterns that every prompt engineer will use in their dial conversation with AI to get the best output and to train our AI models. This is some basics we will see in detail in this lesson. Let's see if that is, these are the four basic prompt patterns like zero shot prompting, few shot system instructions, and role playing prompting. We will see the first one that is zero shot prompting. It is asking model to perform a specific task without prodding any examples. That's mean. Just writing the prompt, we don't give any specific contexts like background information that we have earlier discussed about what is a context, right? Context means providing enough background information in this prompting pattern, we don't give any example or we don't give another information background information to do task, right? You can see the prompt example here. Then you can easily understand. Summary is the main idea of the following test. You can insert any text here, paragraph in any that the AI will easily generate out. But let's see in this chat, we'll jump into our tajivity AI language model and we will see how this zero shot prompting with works. I will jump in I'll jump here, the Cha GBT. So you can go if you are already using the Cha GPT, that you can know how to sign up and get the account from this. So it is easy. So let's see, main focus is on zero shot prompting. Using this summarizing, summarize the main idea of the following. So I copied some paragraph from the Internet. So I'm going to paste here. I paste it here. Then we will see the output, how the AI language model will generate. Let's go. Yeah, you can see the summarization of this text I provided here in the prompt. So it will simply summarize the three points likely in 200 prompt lines. What is the actual this paragraph here saying? Easily summarizes it. So it a simple zero shot prompting, like you have asked some question or you have written the prompt to do a particular task and write another thing like summarize this book and provide some book name as all. So you can see the example here, summarize the rich dad and book. Let's see how the AI will generate the output. I will summarize all contents that are very important points in the Rich and put dad book have. So it will easily summarize this. I have, he completed some specific task like rich dad put dad by Robert Kos KFS as the difference like. So you can do anything. Zero shun prompting means simple writing prompt to perform a specific task like this summarization or remove any grammatical mistakes from this paragraph or remove effect from this paragraph like that. It is a simple tasks that you will ask you to AI to do that. Okay, it is easy. So let's see our second prompt pattern that is few short 7. 2.3.2 Few-shot Prompting: Short prompting. It is opposite to the zero shot prompting. Soft if you understand this, you will understand easily what is zero shot prompting. Eight. So let's see. So if you short prompting means you will provide some few examples in the prompt to help the model to understand the task. Okay. You will provide some examples how the output should be, right, how you want output. You will provide in prompt itself. I need output in this format, right? You will provide this type of stuff in prompt itself to help the model to generate the output what you need. Okay? We can see the prompt example here. Jump to Chachi to explain in more detail of this few short prompting. Let's go. Okay. In few shot prompting, the main purpose is to tell AI to perform task in this way only. Let's see. Few short prompting means providing examples how the output should be should look like B. For example, I use two people conversation like Sara. Okay. Let's see hi. How are you? Sorry. So we'll just try this. Let's see. Now we can take another person like Sam. It will tell, I am fine. What about you? So I have written some conversation between the persons I have provided how you should act like that. So I will provide one more example like Sarah. Yeah, I am good. What you are doing right now. The covex Sam response is, I am looking for us at my home. It's simple I have taken the example. Now, I will write Sara. Can I drop you know what happens, I will just write a SAM and I will simply don't write this format. After that, I will instruct AI tool to complete the Sam's response. Then it will generate a SAM response. Why? Because we have provided a example like how you should give the answer. So I have provided some few examples like a Sara and a SAM conversation. I have. After this, if I not write the SAMs response. So let's see. That is good, right? The SAM. The AI is generated SAM response like here, because it learned from my example how the output should be. That is all about fusion prompting in which we will give some example to AI how the output should be, how you want output, like by giving example. It is one type of example you can give any type of example like this. You can give the output should be in English format. You can give all those stuff. Providing a few example. You have to write some prompt itself, you have to write some question and answer by yourself, that I can learn from your instructions to provide same output how you try to AI. So you can see there. So I have read that few examples of how the output should be. I just ask you to complete the Sam's response, so it will generate That's a real kind of view, Sarah. If it's not too much trouble, I would appreciate this right. Thank you. This is all about few short prompting. So it is easy. So we can compare these two things with few short and zero short prompting. Zero short prompting means we don't provide any examples like few short prompting that we have earlier discussed now. We will just write a prompt to perform it task. Okay, without providing any examples. When compared to few short prompting, we will provide some few examples to help with the model to understand our task and to generate the output, how we bond. It is simple as that. Okay, let's see the third prompt pattern that is system instruction. 8. 2.3.3 System Instruction Prompting: System instruction. Okay, so to understand better, so we have some playground by Jagt itself. So in which we can write the system instructions. After that, we will see in upcoming advanced prom patterns, that is how we can write system instructions, but we will see some basic about this right now. So what is it? Sort of setting the role or tone for the model to follow. You can see the use case when you want the model to behave in a specific way, such as an expert, teacher or a translator. So how the system instruction means, you can see the example here. You are a professional chef. Here we are given some context that we have we are given the context. Context means we have provided some background information. Background means you are a professional chef. Okay, you can see the sum prompt. Explain how to make a simple pasta dish to someone with no cooking experience. This is called prompt. That is instruction. It is called an system. It can be easily understand by Chachi Bitty Let's go to this B Let's see this how the system instructions work. Okay. Let's see. I will try the AI module like you are now expert at writing content on health only. This is a system prompt, system instruction. Okay. Can see this. This is you are no expert at writing content on health only. The EI model will think, Okay, I'm a system and I am only expert at writing health related content, not other. Then I will write some prompt. Now, please write about new reion. Let's see what will generate the AI. See, you can see here. The importance of nutrition for a healthy life is so and so blah, blah. I will generate related to the nutrition. Okay. So we have defined the system working here in which you have to only work. Okay? Write the content for that topic which is not related to health, we can check whether it is thinking like system or not. Please write content for. We can take another topic that is IT. We'll check what is the output of this. See, you can see here. That is a system prompt. Currently, I am focused on creating content related to health and nutrition. Let me know if you'd like to assist in that domain. You can see the system prompt, how it works. The system work means we will give some system to do spun specific task only. That is a system prompt. Okay? After that, we will write some prompt to follow our instructions. Can see I have written some system prompt like this. You are no expert at writing content on healthy only. This is a system prompt. Okay? This is a system instruction. Okay? After that, I have written some question or query. No, please write content about nutrition. It had generated some nutrition because it is a health. Consider topic. When I ask EI to write about IT content, then it will simply refuse to generate the content because the EI is thinking like system, specific instructions that I have given to the EI to do the health content only to generate healthy content only, not other the AI will thing, I am a system. No, I am trying to generate health related content only, not others. If we ask not related to health, it will simply refuse to generate this. So this is an example of system instructions which are very important when we try an AI module to do particular or specific task. So I hope you understand this system instructions. By practicing by yourself, you will get more idea about this prompt. Let's see the role playing techniques. 9. 2.3.4 Role-playing Technique Prompting: Let's see the role playing techniques. It is a quite similar like system instruction. So because in this role playing technique, you are going to pine a model as a specific instructor, like you can see here, instructing the model to act in a specific role, such as historical figure teacher or professional in that. So in earlier, we see you are expert at writing content for health. Okay? This is also role playing. Use cases, create our instructional task where persona improve engagement and understanding. Yes, role playing means persona is most important. Persona means personalization. Uh, training AI model. Training A module for a specific task by assigning the specific role in it. Let's see the prompt example here. Pretend you are but Einstein explaining the theory of relativity to child. Let's jump into the higbity to understand more about this role playing technique. Okay. Now, we can write the forgot above. So this is the forgot is very important when you doing different different things at a particular hagbit interface like this because it has some memory update function in it. Let's see forgot about now, you are no, you are experienced science teacher in which you have expertise in photosynthesis. Now, what I have, so I have assigned a role specific role to I model to act like the role that I have given to E no, R experienced science teacher. This is called in role playing. Okay. Role playing means telling the AI to think like specific role. Think like a science teacher, think like a experienced science teacher, in which we can get the best output from the AI. Okay? After that, I have tell that in which you have expertise in photosynthesis, tell the AI to specific topic, you have expertise in photosynthesis. So now, I will write the prompt. No, I will write the query that I want the orputf AI model. So explain me about photosynthesis in easy, understand way. Let's see what AI will generate. You can see the memory update option here. It has a great future in Jagibty. So when compared to other AI model, you can see here photosynthesis is made easy, it will explain me about photosynthesis when compared to other type. If you ask, see if you can see the example here. So from now here, the AI will think, like it is experienced science each. So to break out this pattern, we have to write forgot about. Okay, I will forgot the above previous role playing technique, and it will generate as casual, we will interact with. Okay, I role playing technique will reduce the irrelevant response or give the better relevant response when compared to writing a prompt without role playing. Okay? If I write simply explain the photosynthesis, it can just throw some random words and random explanation without going deeper explanation. If I try with role playing technique, if I try EI, if I tell AI to think like experience science teacher and generate about photosynthesis on so topic. I will experience teacher how they think and how they explain with the subject expertise, the EI will also think like that and it will generate explanation like subject expertise that I have. You can see how this easy. If I just forgot about and just explain about photos. You can see that. It will just explain photosynthesis, not much better output when compared to previous one. So you can see this here. This photosynthesis is a process by which green plants, algae and bacteria convert sunlight into lab, blah, lave. It has some summarization part of this, right? When compared to this, it has some good points, key ingredients, the kitchen, the recipe. Here it has given the best example here. It's in the formula, what is important, all the dsing. But when compared to here, it will just thrown the explanation about what is a photosynthesis. This is how the prompt role playing techniques will play a major role when compared to get the specific information from AI with deeper knowledge. Can easily understand by this by practicing with your own prompts. Okay, write the simple question, ask the AI, and it will generate some AI response. You analyze it, and after that, you write the prompt with role playing techniques like RS experienced science teacher, like other stuff and give some background information. After that, you see the output. There is a much better output by using their role play technique. So you can see this the two difference between just I have. Without role playing technique, I just written the query explain about photosynthesis. You can see the output here. That is not much better when compared to this output because I have used the role playing technique in this prompt, that is you are experience science teacher in which you have expertise in photosynthesis in which I have trained AI model in specific way to get most of the AI that's it for this model. So we will see some role playing techniques so you can easily understand by practicing by yourself in the chargeb itself. In this model, we have discussed some prompt patterns in which we have discussed some zero shot prompting in which we just ask a question or we will try a model to perform a specific task in which we use chargeb to do some summarization uh, some paragraph. After that, we see this few short prompting in which we have provided some few examples to get the output what in the format we want, and we will generate it from the charge JB itself, we see some system instructions in which we give some system in which the system only works with our instructions. After that, instead of out of that instructions, if we ask model to perform a task, it will refuse to do that perform task, which is out of system instructions. Okay. We can and the last one is role playing technique in which we have seen providing a background information, assigning a specific role, we can get better response when compared to asking random quotien. Don't worry, guys. I will put this tagbty chat link in a document itself you can get after this course or models assignment. Okay. So that's it for this model guys, we will explore some more advanced and prompt patterns in the next model. Let's dive into our next model. 10. 3.1 Structuring Prompts for Optimal Output: Dive into our module number tree in which we are going to learn how to structuring the prompts for APML output, and we will discuss what is a simple structure to follow while writing the prompts. And we also explore some example and how to write that mes to prompt using structure that we discuss right now, and we'll jump into the ha GPT, and we will see the practical information of this prompt. Okay? First, let's discuss this simple structure to follow. Okay. Imagine you are giving some instructions to a particular person if your instructions are not very well, okay? That person and cannot find the right place that he want. Similarly, the AI model can also think like output like that. If our instructions are not clear, the AI will generate irrelevant response. First, we discussed the structure. The structure contains three parts, mainly three parts. That is role setup, number two task definition, number three context. You can see the role setup. As we earlier seen some prompt patterns like role playing, system instruction, and some bad and good prompts. First, we understand this simple structure to writing the best prompt here. So first structure is that is role setup. So we have to set up we have to assign some rule to AI to think that in that background you are a helpful assistant or you are a experienced teacher, you are a scientist, we assign some specific role to AI to think in that background, which leads to a better response after assigning we will define our task. What I need from the AI, that is a task. Next, third one is context. We have to provide any background or additional information or examples that can guide the response. Okay? Earlier see some few shot prompting, in which we have provided some examples to AI to generate output like we want in which we have defined that output in prompt itself. That is a few short prompting. So here, context similarly say we have to provide additional information in which topic you want the output. That is the background information. This can be easily understand by writing the prompts. I have taken some examples, poorly structured prompt. Simply tell me about AI. So you can see it is simple, tell me about AI. There is no other information. There is no role set up in that. This is a simple question. I will ask you to AI, how the well structured prompt looks like. You can see I follow the structure that is this structure, role setup, task definition, and context. So you can see the here. So I assign a specific role to AI. You are an AI expert. This is a role setup. After that, I have write that task. I have defined the task, what actually I need from the AI model. Like explain what artificial intelligence is focusing on its application. You can see this is a task, explain and provide concise examples for each sector. But where is the context in it? Where is the additional information that I given? You can see the healthcare, education, and transportation. I need the output for these three different types of applications only. I don't need other types of application. That means you have provided some specific additional information in which I can generate output for these three types of applications only. That means you guided the EI to generate response in these three types of applications only. That means you have provided context. Okay? So this is the difference between the poorly structured and well structured you can see the here. The second prompt is specific gives a clear task and sets a role for that model resulting in better output. So for more understanding, we will jump into Hagibt and will see the output, how the output looks like for these two types of prompts. So I jump in the hagibD. You can use any other model to analyze the two different outputs, I will write some poorly structured prompts like tell me about AI. Let's see what AI can generate. You can see the artificial indigency refers to simulation of homon intelligence. So it has generated related AI concepts, supervisor learning. So I don't need this type of all those things. Okay, it has generated summarization like that. So I use this prom. That is well stuctuedPmt. I already copied that. I will based it here. So okay. I have written the well structured prompt. So let's see this is what is the output of this. You can see this, the output is different from previous prompt. Why? So I have defined a role in which the AI will generate a better output in that background only because it is AIX but that is specification. Prompt engineering means specification. Writing the prom for specific use cases is called an prompt engineering. Can see the. After that, I have task definition I have and already, I have provided only please generate explain the artificial inilgence in this type of applications only. Okay, that is healthcare education transportation. In which I guided the AI to generate output for this type of applications only in which the AI is generated that I want, healthcare education, transportation, like this simple. Otherwise, you can write like this also. I explain what artificial intelligence is focusing on its application in healthcare only. I just delete this. I can see the different output from it will explain only healthcare. Okay. I hope you understand this lesson. In next lesson, we are going to learn iterative prompting, which is best and most important method to get the best output from AI. Let's dive into that. 11. 3.2 Iterative Prompting: Back to this lesson in which we are going to learn the most important technique that is iterative prompting. So this prompting is quite similar that we already discussed earlier, which comes under the multi turn conversation in which we write the prompt and I will generate the response. After that, we will write the prompt, that is follow up prompt, follow up quien to adjust our output. Lo, this is called an iterative prompting. We will discuss more in detail in this lesson. Okay? So let's see what you learn from this lesson iterative prompting. Will learn how to refine proms to improve AI responses and we will see some technique and we will see some examples as well to understand better what is iterative prompting Whyto prompting is important. The language models are trained by large amounts of data. The language models sometimes need guidance to generate output that we want. You can see the iterative prompting is a process of adjusting your proms based on the output you receive. Okay, means first, you will write some prompt to AI, that is instructions or question or query. According to your prompt, the AI will generate a response. Okay? The output is analyzed by you. So you can see that this technique is essential for refining and narrowing down your responses to meet your needs. Okay, it is the best and most effective way to get most of the AI model means to get most of the effective output from Chat GPT or any language model. So there are some steps we have to follow to iterate effectively. First, we have to analyze the output. First, we have to write some prompt, it will generate a output. The first step is we have to analyze the output. Check if the response aligns with your intent or needs or how you want. If aligns with your needs, that's good. If not, means what I have to do next. Second thing is you have to identify the gaps. Look for areas where the output is not clear or have some inaccurate data present in the output. So you have to identify that gaps. After that, you have to revise the prompt. Means you have to write the follow up prompt best then the previous prompt. Avoid the previous output. Okay? So it can be easily understand by practical implementation and practice. So we will see this also. You'll see that also. We will see some example to get better understand InshalPmpt describe renewable energy. The output will be the more I have just, for example, I have taken two lines only. So renewable energy comes from natural source as examples include solar and wind energy. This is quite simple answer that I have taken for this, but the output can be very long, okay? Have revisor sum prompt. Explain renewable energy, its benefits and three specific examples, solar, wind, and hydropower, use simple language for a beginner audience. This is a well structured prompt. This is a revisor prompt. You'll see in the practical information JGB. You can see the revisor prompt sets clear expectations leading to a more detailed and taller response. Okay, we will jump into JGB. We'll see how it works. Let's go to JGBT. I will write a simple question like, Okay, we will take this previous example of our PPT that is describe renewable energy. Let's see this. Energy. That's what the output will be. You can see some output related to our remable energy. You can see the output here, advantages, challenges, applications. Okay. It is best. Okay? This is reverse to energy derived from natural sources that are continuously replenished and virtually instable. Okay, if I beginner to renewable energy, what is the meaning of replenished and virtually inaccessible? So I don't understand it. So for that, I have to go to the Google and I have to write the meanings of replenish inexhaustible. So what is that? So to avoid these things, I will write a prompt like this. I have copied that, so I will just direct a page here. So explain renewable energy its benefits and three specific examples solar, wind, hydropower. Use simple language. This is most important when you are going to learn something from the language models because the AI is trained by the advanced English. So let's see this output. You can see what is the renewable energy C. You can see this very, very clean and simple language that we can easily understand this topic. So you can see this here. What is renewable energy? Renewable energy is energy that comes from natural sources like the sun when these sources are always available and don't run out, unlike coal or oil. They are clean and help to protect. See, you can see the examples how we want the output, very effective output when compared to you can analyze the outputs. You can check these two outputs. S, this is very effective words, complicated words that we cannot easily understand as a beginner, but it can be easily understand by beginner because it is simple language to explain to us. That is why writing your requirements, providing the more detail what you need from AI is powerful. It will generate according to our can see I have just returned the simple prom. After that, I analyze it. I analyze it. So this output is better, but I cannot understand. And then I identify the gap. What is the gap? So I didn't understand these two words repleted and inaccessible because I don't know, because I am a beginner. What I get the AI, so exper renewable energy. Use simple language because I have to learn renewable energy in simple language. I have to understand because I'm a beginner. Simple. When I got this idea after analyzing the first output. That is what the prompt engineering is. There is a first and most step you have to follow is in iterative prompting, you have to write any first instruction after that, analyze output and change your next prompt according to your need and more detail as you can. Possible. After that, it will generate some AI, which is more effective than previous one. That's simple. So that's why iterative prompting is very most and effective method to get the best output according to your need. We will move to next lesson of this model number three in which we learn some context management that is how we have to provide right background information in our prompts. Let's dive into that. 12. 3.3.1 Context Management - Part 1: Welcome back to our next lesson of this model number three in which we are going to learn what is context management. So we will see some context techniques and some tips all example of like this. We'll see what is the role of context in prompts, which is very important and some tips, and we will see some examples. Let's see. So context management. Context management means providing background or additional information to AI in prompt to guide AI to generate a output that we want. What is the role of context in prompts? Remember this, if you provide too little context, too little additional information that can lead to some irrelevant or unclear output or response from AI. On other hand, if you provide too much context can lead the model and reduce the output quality. How we can write the best or right amount of additional information to AI in which we can get the best output from AI. Let's discuss in detail. Model right now. The key is to include just enough information to guide the AI without overloading it. Yeah. You have to include just enough information, what you need. That is enough information to guide the AI without overlaiding it. Because some people will just write the additional information which is not required generate output, which is not required. We have to delete that we have to write what we need exactly. That can lead to a better output from the AI. We will see some tips for managing context. Let's see. Be specific, include details that help the model understand your needs. Just include details what you need. That is simple. You do not need write or additional information which is not required that topic in that output. Just write the specific details that help the model understand your needs. Next is use examples. If your task is complex, you can include sample outputs to set expectations to guide the module to generate output like this only. Already we have earlier discussed that is few short prompting in which we have provided some examples how the output should look like, exactly what we have in this context manner. Context means providing additional information or examples helps AI to generate a output that we want. Okay, that's simple. Use examples and third one is avoid redundancy. Redundancy means, keep the prompt concise and to the point. Okay? 13. 3.3.2 Context Management - Part 2: You can see the example, best example here. So these two proms are very well structured, but it is more overloaded. You can see this. You are an expert in climate science role setup, it is very good prompt, and this is also very good prompt. But it is overloaded, right? So write a detailed essay about the causes effects and potential solutions, it is more detail, rather than this, but it is optimizer. It is overloaded. Why? It can be defined by seeing the output only. Because we will see in the ChatGPT. The optimizer prompt keeps the task focus while being informative. First, we understand these two proms. This is a well structure, it is also a well structure, but it is all ded by more additional information. But why? The more detail can be guide the AI to generate the best output. But why it is overloaded. When you try AI, you are an expert in climate science. You are an expert in climate science. You do not need to write all these subtopics because in climate science itself, it already know these topics and you already understood. Your task is write a detailed essay about the causes effects, and all these things. It is all these things already known by AI because it is export in climate science. Instead of this giving more information, you can just write like this, write a 400 essay about the main causes of climate change and three potential solutions. Use examples and data to support your points. Told the AI, use examples and data to support your points. What is the data to support AI points means that this is carbon emissions, deforestation, industrial pollution, renewable energy sources. All these things comes under this example data support your points, which is already known by AI that is climate science. You do not need to write all these subtopics. Okay, because if you don't give this additional information, already AI can know what are some causes and effects done by the climate change touching, all those things. This is overloaded because we have given so much additional information, this is not required because already AI know AI is no expert in climate science. But when compared to here, it is well optimized because why you are assigned the role? That is you are a climate scenge expert in which the expert know all these topics. You are intent, write a 500 word essay about the main causes of climate change and three potential solutions. We'll go to the ChatGPT and I will paste this overloaded prom first, then we will go to the optimasure prompt. Let's go. I'll take the new chat. Let's paste it. This is a overloaded prom that I have directly copied from my PPT and will paste here. Let's see what AIs output is. Here's some output that climate change causes effects and potential solutions. Causes of climate change has some points explain about causes of climate change here. Let's see deforestation, industrial pollution, effects of climate change. Okay. Extreme weather winds, bidiversity loss. It's good. It's some detail as detailed because in overloaded prompt, if you give the O Edit prompt, the output also be overloaded, simple, right? So AI is enside by us, the AI will only generate according to our needs and prompt. Let's see this overlaid prompt we have written here. So the output also same that we have given here that is quite long and more detailed. Let's see what happens with up to measured prompt. I will copy from here and go shared GPT and past. And it is some optimizer prompt. Let's see what is outputting for you. The main cause is, you can see it is explained G deforestation, greenhouse gas emission, initial agric activities, three potential solutions transition day renewable energy, reforestation conversation efforts, policy reforms and global agreements. Conclusion. It is best output from previous then. Why? So it is not about output. It is all about prompting. Okay? So you can see here you can see the output. There is something causes of climate change, global warming, carbon emission, deforestation, industrial pollution. You can see the exact output from other otro measure prompt like main causes of climate change, greenhouse, gas emission, deforestation, industry, and agricultural activities. So it doesn't give additional information in this prompt here, explain like this. So when you compare this, write a detailed essay about the causes and potential solutions to climate change, touching, global hermi carbon emission, hinder shell pollution. I never provide additional information in this optimizer prompt, but AI know about that topic. Okay, because it is expert at climate science. Okay? So it will automatically generate about what is the main cause of climate change, like greenhouse gas emission, deforestation. We don't need to provide an additional information here. That is simple. That is why the context management is very important. Okay. 14. 4.1 Prompt Optimization: Come back to master prompt engineering model number four, in which we are going to see some advanced prom patterns. Let's dive into that. Before going to discuss advance prompt patterns, we see some prompt optimization tips techniques. Okay, we already discussed earlier some best practices to write prompts. Okay, don't confuse, recall them. What is actual is prompt optimization. You can see the optimization is the art of fine tuning, fine tuning, it is similarly refining prompt, training AI with your prompting. Okay, simple that. So optimization is the art of fine tuning your prompts to ensure clarity, reduce ambiguity, and improve engagement. This is three is very important. You have to keep in your mind while writing the prompt. Okay? The best prompt will reduce ambiguity and any irrelevance response. Okay? It's about asking the right question in the right way to get the best response. That is, prompt optimization means simply asking a right question in the right way to get the best response. So there is some key points we have to keep in our mind while writing the prompts. First, we have already discussed that is clarity. Okay. Clarity means using simple and precise language, avoiding the confusion or unclear words or sentences that AI cannot understand our intent to generate a relevant response to our query or task. Okay? Instead of C, you can see the example here, we have to recall, again, tell me about history. You can see that there is no clear in that. Tell me about history. History means it is a broad thing. So the AI will think, Okay, I will have to explain history. So it will just generate a random information relative to history. Instead of that, if you use clarity, that is, can you provide a summary of World War tools, causes and outcomes? It is a specific topic in the field of history. Now, AI has think clearly, Okay, this question have some clarity that the AI will think, I have to provide a summary of World War to causes and outcomes. It is a specific topic that AI can generate best output related to this prompt. That's why we have to make sure the prompts should be clear and specific as possible to get the best output from the AI. Let's see second point that is roll off formatting. Formatting is nothing but using headers, bullet points, and small headings. That is all these things. It is a best practice if you use formatting in the prompts. If looking to become a professional prompt engineer, your writing skills should be very effective. Okay. The more effective you are at writing, the better output you can get from the AI. You can use some format like using bullet points in your prompt, number last or headers in your prompts to get structure response. Like an example, you can see here the list of following in order, advantages of solar energy didvantages and future potential. Just to guide the AI to get the output in this format. It is simple. Third one is engagement techniques. What is the engagement techniques? When you are looking for content creation or article writing or anything that is. So for that, we have to frame your questions to invite curiosity or provide context. Context means here, background information, right? Additional information regarding your topic. So example, you can see. Imagine you are a scientist in 2050. What breakthroughs in AI might you describe. Imagine you are a scientist. We assign some role in 2050 that is a future. Okay? So how AI think like that, AI think that I'm a scientist in 2050, what the breakthroughs in AI might you describe. So that AI will generate a best output regarding that AI is thinking I'm a scientist in 2050 because you build a connection. The AI is connected with scientists in 2050, in which it can generate a best output. Imagine what it takes to engagement techniques. If you are not use this type of prompting here, imagine you are a scientist. You just use what breakthroughs in AI might you describe in 2050. That is a simple thing. That is a simple question. If you use simple, imagine you are a scientist. That you are doing you are getting a response to engage in their thoughts. Okay? They engage in their data that will connect with their knowledge base, and it will think, and it will generate engaged content rather than asking simple questions in which there is no engagement in the Okay. So this is why you have to use some techniques, words that AI can think that AI can imagine and which connects their knowledge base and words that can describe your output very easily. Okay? Key ones that we have to keep our mind while writing the prompts for AI. That is to get the best response to improve engagement, reduce ambiguity, that is unclear response and clarity. Okay. Let's go to our main part of this model that is advance prompt pattern P one. Which we are going to see and most important best practices, prompt patterns as a prompt engineer, you need and you have to use for solving complex tasks. Let's start. 15. 4.1.1 Ask for Input Pattern (Advanced Prompt - Part1): So in this advanced prom patterns, part one, we're going to see the five different prom patterns which are most important to get the best and effective output from KI. Okay. Let's start one by one, and let's dive into our first prom pattern which is Ask for input prom pattern. Okay? In the next part two of this particular model um or four, we will discuss another five most important prom patterns that will blow your mind. Let's dive into our first prompt that is Ask for input prom pattern. So what is an ask for input prompt pattern is? You can see to use this pattern, your prom should make the following fundamental contextual statement. So what is an contextual statement? As we earlier discussed what is a context, giving the background information of particular tasks that you are looking to do with the AI. Okay, that is simple, providing your own details to AI in order to get the personalized output for, like that. Okay? So by using these fundamental contextual statements to AI, in the prompt itself, it will take your information and it will it will do that task according to your requirements, in which we can get the personalized and effective output for our task and what we're looking from AI, like that. Okay? So to do that, you can use this simple statement that is Ask me for input X. So what we are writing in this year, we are just telling to AI, ask me for input x. X, it can be any task, question, ingredient or any goal. So these words are refers to you to your requirements to your background information. When you define that task and you will ask to AI. Ask me further. So I'm so tedious that when I give that particular details, you need to do the task like that. So you can use this Ask for Input Pattern by giving information to AI in order to complete the task and in order to give you effective output for your requirements. Okay? So let's jump into ChatGPT, and we will see some examples how this actually ask for input prom pattern works. Okay, let's dive into our ChatGPT. So as we know, the ChatGPT has newly given the versions of GPT five, so you can get reversion like the ha HIPA as usual, you can upgrade for the most smartest model as well. But remember one thing, the prompting never changes, but according to the model versions, the output will change. Okay? There is no change in the prompting, but there is a change in output according to your models you select in the ChatGPT or other AI models. Okay. I hope you understand this point. So I will just go with a free version of this ha chiPT. And I will past some prompt, which includes the Ask me for Input prompt pattern. You can see from now on, I will provide fitness cos and other relevant details about my routine. So you are just describing your task. You can see I will provide fitness cos and other relevant details about my routine. You will create a weekly for code plan tailored to my input. So this is most important thing. So what is an input? It is a input. That means you are telling to AI according to my requirements, you need to provide a fitness code. Okay. Like that. For each day, include exercise, it is a main task. For each day, include exercise sets and reps. At the end suggest a recovery activity for the week. My name is Alex, and this is the Ask me for Input prom pattern. You can see here. Ask me for my fitness scores and current fitness level. It is a prompt pattern, fundamental contextual statement. Ask me. So what we are telling to AI. So there are two conditions while using this one. So we need to focus. I will provide fitness goals. Okay, you need to describe this first condition, and you need to write this fundamental contextual statement at the end of the prompt that is Ask me for my fitness goal. Okay, let's see how it will works. So you can see got it, Alex. First, let's start with the basics so I can build the best weeklp workout plan for you. So what are your fitness goals? So when I give the answers for this particular questions, okay, I will just ask some follow up questions and then when it get the required details of me, then it will generate the partialized fitness routine for me. That is how this Ask me for input prompt pattern will work. That is very powerful. Let's see, let's give the answer. I will just take this muscle gain. And I'll take the beginner. Let's hear this one. Thanks for S. You can see. After gathering my information, my personalized information, it will generate the tailored fitness plan for me according to my beginner level and my fitness goal like that. You can see day one is a push a day to pull back, biceps day three and day four legs like that. This is how you can use this particular prom pattern. This is simple prompt pattern that I have just shown you as example, you can do anything you want in order to get the best output which is personalized to you. Not only the fitness, you can go for the learning something new skill or you can generate the content for specific purpose. There is so much things you can do with this particular prom pattern. You need to use two conditions. That's it. That is, you need to tell AI, I will provide. So what it may be that your details should be mentioned here, tell I will provide fitnesses. It can be anything. I will provide my personal data. I will provide my content details or specific details like that and write the task and just add the ask me for input prompt pattern statement at the end of the prompt, you just go with that. You will see the output which is very effective when compared to just writing the simple prompt. I hope you understand this prompt very well. For more example, let's take another one like a copy and paste this prompt, we can see. From now, I will tell which speaking language you should use to translate the given text, how are you? Now, ask me for which language I need to use. So you can see here from now. From now, I will tell which I have just described. I will tell which language, which is speaking language. This is the most important condition you need to describe. Should This is a task here, right? And this is the fundamental context statement. We need to include at the end of the prompt. You can see you should use to translate the given text, how are you? Okay? When I give this particular prompt to AI, it will start giving the asking the input for me. You can see. Got it. Alex, which language would you like me to translate? How are you? So it is asking me to provide you a partialized input. When I give the input, which I am required the output is, then it will generate the same output that I'm looking for. That is simple. I will just give the Spanish. Let's see the output. Now you can see in Spanish, how are you? Is this one? Do you want me to also show the formal version user respite polar situations or like that? You can go forward like this. That is simple. You can go with Indi French and you can do the German. You can do so much things by using this prom pattern. This is simple to examples that I have shared with you, so you can go for them more and more for more understanding of these prom patterns. So I always tell you because you need to practice this particular prom pattern for different aspects for different requirements for different applications, then you will get the more knowledge about how to craft this particular prom pattern for doing the things to get the specific output for your specific preferences, like that. I hope you understand these points very well. Let's dive into our second prom pattern that is persona prom pattern. 16. 4.1.2 Persona Prompt Pattern: What is Persona Prompt Pattern is? We need to assign a specific role to AI in which the AI thing in that loop and it will give the best and specific output according to our requirements. Okay? You can see here, you can use this act as a high school math teacher. So the thing is here, you need to assign a specific role. In this case, you can see act as a high school math teacher. So we are assigning the specific role with some experience like that. Okay, you can add anything you want in the specific manner to get the specific output from AI. Okay, you can see in this case, we have just taken the act as a high school math teacher. We are just assigned a role to EI in order to think like expert math teacher, high school math teacher, and it will look in that particular field. I will just give the output like high school math teacher will give. It will just go in that deep explanation, it will explain also topic to the student. I hope you understand these points. You can assign a specific anything that act as a expert math teacher, each and everything if you want. Okay? You can see this is a simple prompt pattern, assigning the specific role to the personal prom pattern. It is the most important and very effective prompt pattern for every industry. You can not only in the education, you can take anywhere you want to get the best output from the AI in the specific field in which you can get the effective insights from EI for the specific topic. This is a most popular and highly effective prompt pattern. You should use for every prompt that you are looking to write in any EI tool. I hope you understand this one. You can use this one. So let's jump into our ChatGPT and we will see the difference between the simple prompt and with the personal prompt pattern. Let's dive into that. I'm just going to the ChatGPT you can see this is a simple interface. I'll just paste this prompt pattern here. Now, before going to this particular prompt pattern, I will just remove this one, act as a high school math teacher, and I will just take this one as explain Pythograph soem to a 15-year-old student. Let's give this one. No, I will give the answers like simple giving the answer for according to our prompt. You can explain Pythograh theorem to a 15-year-old student. Now, you can see this is the output. Okay. When we assign the role to particular application, you can act as a high school math teacher and explain Pyogras theorem to a 15-year-old student. In this case, the AI will think, I am a high school math teacher, I need to explain the Pythagraom to 15-year-old student. In that case, the AIT in that loop, and it will give the answer like how high school math teacher explained to the student. You can see your response. You can see all right class, let's tap into Pygm. You can see first remember this only works for the right angle triangles. You can see this output here. You can see this one. But in the previous prompt, you can see the output. Okay, let's break it down in simple and one way Pythograph serum is a rule in mathematics that works only for the right angled triangles. So this is also a good one, but it is simple. There is no student engagement in this particular output. But when you see this one, when we assign the role, act as a high school math teacher, the AI now just explaining the topic as a teacher. Math teacher. You can see this is one, all right class. Let's dive into Paddan. First, remember this only works for the right hand triangles, and with angle equal to 90 degrees, the two shorter sides, longest side, simply you can see, suppose one leg is six centimeter, you can see the example is is taken, and you can see this one. Imagine a trangle with three sides, three sits. So this is engaging way. So when you add the assigning a specific role, the EI output will turn into the human written output. Not only that if you copy this one, you can check the AI tool, it is a AI written or human writing. Of course, it will come with the AI written. But the way of writing way of explaining the things will be different from the AI output. I hope you understand these ones. That's why you need to add this assigning a role to this particular any prompt that you are looking to get the answer from AI, you can get in engaging way and you can get the accurate some accurate output and effective output that you can use for your personalized experience. I hope you understand these points very well. And let's see another example for more understanding purpose. Just copy and paste this particular prompt you can see here, act as a travel recommender. Now you can see it is act as a travel recommender. Okay, even the word is wrong, there is no problem in that, the AI will automatically rectify it. The AI is smart enough to understand your intent, what we're looking from AI. I should automatically give the best output. There is no problem in that. You can see act as a travel recommender. It is a simple ascender AI specific role that is travel recommender. Okay? Now I have just include some. I will tell you which city you need to give recommendation. So this is the Ask me for input prompt pattern, first condition, right? So I have just described the AI. I will tell which city you need to give the recommendation to visit such beautiful places in that city. Okay? This is a task. Now, ask me, it is a second condition, we need to include at the end of the prompt that is fundamental contextual statement that is Ask me for input X. You can see here X is looking to visit city you are looking to visit. I hope you understand these points. So what we are doing here, so we are just combining the a current prom that is persona prom Pattern and a previous prom pattern that is Ask me prom pattern to do some particular task in which we can get the best output from the AI. Okay, let's give this particular prom to AI. You can see great. Which city are you planning to visit, so I can recommend some of the most beautiful places there. Okay, now, I will just give the city name like Tokyo. Let's give this one. Perfect choice. Tokyo is the city where modern skyscrapers meet ancient temples. So this is a simple plan, travel recommendations that I has generated. Now, if you see this one. Okay. Try by yourself, remove this one act as a travel recommender. You can get the exact output, but there is a difference for more understanding by yourself, you can try by yourself just to copy this prompt by yourself and you can try for the different applications. Try by yourself, just go write the personal prompt pattern for your requirements, check that output and another output which is not included the personal prompt pattern, you will see the magic of this personal prom pattern. I understand these points very well. Now we will jump into another third most important prom pattern. That is quotienRfinement prom pattern. Let's dive into that. 17. 4.1.3 Question Refinement Prompt Pattern: Now what is an quotient refinement prom pattern is? So it is the most important and very useful prompt pattern that you can use everywhere in order to improve your output. Okay, let's see how it works. You can see the template for this pattern can be expressed as whenever I ask a question. It is a simple prom pattern, you can just focus here. Whenever I ask a question, suggest a better quotien and ask me if I would like to use it instead. That is simple, right. So not only the question, you can take any text, paragraph or story like that. When you think writing skill is not very well, you can ask AI whenever I ask A or whenever I provide this text. Okay, suggest a better text and ask me if I would like to use instead. There is so much things you can edit Bself we can use this prompt pattern in so many ways in order to improve your output writing skill or prompt writing skill as well. Let's see this one how it works. That is Question Refinement Prompt Pattern E. Okay, now, we will jump into our this chatbt will go with the new chat and we'll try with simple giving the quotien first, right? I will take this quotien first in order to see the difference between that. I will just return the quotien or prompt to AI. Please generate a story which I need more engaging words and fun for ten year bond. So there is a lot of grammatical mistakes and all those things here. So because my writing skill is proof, know what I will tell. Let's see first output from AI. Got it. Let's create a fun. This is a simple story that AI has generated. Okay, that is good. Now what I will tell. Okay? This prompt is having some grammatical mistakes and some statement, right? So know what I will do. I will just write this particular prompt here. What is? Please suggest me the better version of my prompt. You can see I have just given the prompt. In your case, you can take any passage, text or essay like that. That is up to you. Now, we are running the prompt engineering here, right? So we can improve the prompt writing skill by using ChatGPT. Otherwise, we can generate the prompt by using HAGPD by using this simple prompt button, that is, please suggest me better version of my prompt. That is, please generate a story which I'm more engaging words, and there is a lot of mistakes in this prompt. Now, I will just give this the AI. No, it will see you can see that here's a clearer, more polished version of your prom that will guide AI to produce the best kind of story. You can see this is a better prom. Write a fun, exciting, age appropriate advanced story for 10-year-old boy, use simple but engaging words, lots of imagination and playful tone to keep it entertaining and easy to follow. So even though if I don't have the a background information for this particular story. The AI is already trained by so many prom patterns and the output, the AI has everything. The AI has the knowledge and writing skills, right? So in that case, if you don't know how to write the a best prompt, you can use this prompt pattern as well, get the best engaging prompt effect to prompt from AI itself. You can see there is a lot of difference between this prompt and this prompt. You can check these two prompts output are different, and I will definitely tell you this output will be amazing because it has some specific ness, DanierOldboy, and it has also just given some engaging words, lots of imagination, contextual context in the prompt, all those things. You can see this playful tone which is in the specific manner the output, it can be a specific way, right? You can use this prom pattern. This is a simple sentence prom pattern. When you write this 325526 lines of prom, you can use this. Please suggest a better version of my prom. It will give the amazing results for you. It will give the amazing prom in order to see what time, okay? Let's see. You can say, if you want, I can also give the variations prop like bedstem story. You can go forward as the AI suggests you. We can do with all those things. Now we will see another statement we have just learned in the previous one. L whenever I ask a question. Okay? No, this is the most important thing here. Now what we do? I will just write what I will do. I will just use this prom pattern. That is refinement prompt pattern here. That is, whenever I ask a question, suggest a better question and ask me if I would like to use it instead. Okay? You can go with the question or the prompt. Now, in my case, I will take this prompt test. Okay. Now I will just replace the question with the prompt because my case is here, learning the prompt engineering to improve my prompt writing skill. Now you see whenever I ask a prompt, suggest a better prom and ask me if I would like to use it instead. Let's give this one. Now that I will just take this got it from now on. Whenever you ask for a prom, I will suggest a better version and the check if you would like to use instead. Now I will write any prompt here. I will just take this one, write to Hader words article and global warming. Okay, what happens here? It is a simple question or prompt. Let's give this one. Now, AI will just giving the output here. If you see here, sometimes the AI will not follow our instructions very well. No can see. If you see here, what have told to AI. Whenever I ask a prompt, suggest a better prompt and ask me if I would like to use instead. Okay? The AI is when I give the question or prompt to this particular, this is a write 200 words article 0N global warming. The AI has generated the global warming output, not the prompt here. Okay? So now the EIs task is to write a better version of this prompt, but the EI has generated the output. Okay? Now, last one, you can see that you can see, do you want me to suggest a better prompt for your request before we move forward? No, it has taking this one. No, to solve this problem, sometimes AI will do the mistakes, right? So you can do. You can just give the warning prompt like that. I have told you you can give the order to AI, just like to come back AI into the specific pattern like this, you can take this one. I have told I have told you to suggest a better version of my prompt, not the output. Sometimes the AI will do the mistakes. In that case, what you need to do, you need to tell to AI. I have told you to suggest a better version of my prompt, not the output. What happens here? The EI will automatically write the best prompt here because AI can do the mistakes here. You can see you are absolutely right. Thanks for pointing it out. Let me fix that. You original prompt, write 200 voice article 0N global warming. Here is a better version of your prom. You can see write a clear engaging. This is a simple, better version of my previous prompt that is this. Okay. This is how you can use this one, right? Let's combine three different prompt patterns here. Write the prompt, which contains three different prompt patterns like Ask for input prom pattern, persona prom pattern, and quotient refinement prom pattern here in order to see how the AI will do the task. Okay? Let's just copy and paste this prompt here. You can see act as expert prompt engineering with a ten years of experience in writing effective prompts for AI. So it is assigning a role, which comes under the persona prompt pattern here. I've just described, I have just assigned the role to AI that is expert prompt engineering. I have just given background information that is context. You can see ten years of experience in writing effective prompts for EI for that specific one, writing proms. Okay? Now you can see, I will provide my prompt. You can see this is the first condition of Ask for my input prompt pattern. Okay, you can see this Ask I will provide my prompt you to suggest better prompt from my prom. You can see suggest better prompt from my prom. It is a task. Okay, you can see. Now ask me for prompt to suggest a better version. It seems like something confusion. But if you see the output of this one, you will amaze. You can see now ask me if this is the second condition of Ask me for Input prom pattern. Now, ask me for prom to suggest a better version. Now, I will ask me the prom. When I give the prompt, you can take the basic prompt. The AI will automatically just provide the effective prompt because we have just trained AI like using the personal prom batter, we have assigned the role to AI in which it will just work under conditions of prompt engineering. Let's do this how we will check the output. You can see perfect. I will act as expert prompt engineering. You can see, please share your prompt and I will suggest a clearer, stronger and more effective version of it. Now, I will just take the example like you can see that we block post on AI in detail. Let's check the output. Now you can see. When we have in the previous one, when we write this one without the persona prom pattern, you can see this is a direct task here. When we use the previous one without assigning the role, it will just take in this one. It has provided the output of the particular prompt, not the better version of prompt. Now when we use this one, act as a person of prompt pattern. From now this interaction with AI, it will do the task as a prompt engineering only. I hope, understand this. Good. The great capability of char gibt, it will follow the prompt. It will follow the previous chat, previous assigning role like that. Okay, that's why the chargb is very amazing tool that we will use for the engaging content. Okay. You can see. You can see the difference between that. When I ascend the role act as expert prompt engineering, when I provide the basic prompt, it will just return a better version of prompt you can see here instead of giving the output for that particular prompt. That is why the act as a personal prompt pattern is so most important thing. You need to use to get the specific output form. So in this prompt, we have combined three different prompt patterns. That is, ask me for input prom pattern, persona prom pattern, and the Question Refinement Prompt Pattern. Okay, what is the ion refinement prompt pattern here? Whenever I ask a prompt that is write block post on IDtail you need to suggest a better version of the prompt. You can see this is the output, the better version of the prompt. I hope you understand this prompt very well, right? This is a portion refinement prompt pattern here. So it can seem some confusion, but it can be easily understand by practicing by yourself. You can go with a specific different application right by yourself because of prom writing skill, it can be the learned by writing only. When you try by yourself, you will get the writing skill. Not only that, even you can go the specific application in specific topic like that. I will show another example, if you can see, act as an expert prompt engineering with ten years of experience in writing prompts for AI in educational content creation. Okay, I hope to understand this one. So in the previous one, I have just taken for the effective prompts for AI. So you can write you can get any prompt, you can write for the image prompt, video prompt, or text to prompt, you can write anything. I will give the better version of prompt. But even you can go in the specific application like education sector, content creation sector like that to get the best prompt for it. Okay? Not only the prompt, you can go with the passage I content creation like that, you can go as you want. Now you can see this prompt here. I'll just return the same prom, but I just included specific application for AI in educational content creation. Now what happens here? Let's see. No, not that you can see here. I will provide my prompt you to suggest better prom from my prom. No, ask me for prom to suggest better version. Let's skew this one. Now you can see, got it. I will act as expert prompt engineering for educational content creation. Now, it is a specific one with a specific educational content creation. Not only that you can go in the deeper sector for the physics, for the mathematics, for the English, you can go as you want for the specific one to get the more clear, effective and optimized version of the output from AI. Now, I will just write the prompt, write a full lesson about photosynthesis. Okay, let's give this prompt here. Now you can see, perfect. Let's define your prompt here. You original prompt, write a full lesson about photosynthesis. It is a better version of prompt. You can see create a complete engaging lesson plan on photosynthesis for middle school students, include a clear explanation of the process, key terms. You can see this is the thing we just AI as adding the background information to AI. To give the best output here. It is following the context. It is following all those things here that we have earlier discussed in the and PBT. You can see. This is simple one statement of prom. In that case, it has generated the three to four lines of prompt. This is how EI is doing the work in this world right now. You can see this is one. You can go as you want. You can go for the specific here, even not, you can go for the specific topic to get the best output from here. And remember one thing, the output of AI is dependent on your prompt, the how you all prom is well structured, well defined, and well specific, you will get that output in the specific manner without any inaccuracies and hallucination words. I understand these points very well. Prime of yourself, you will learn the art of writing the prom and you will get the best output from ChatGPT or other AI models as well. Let's jump into our fourth prom pattern that is cognitive verifier prompt pattern. Let's dive into that. 18. 4.1.4 Cognitive Verifier Prompt Pattern: What is a cognitive verifier pattern. You can see the template for this pattern can be expressed as this is a task. How did World War two impact global politics? It is a simple quotien here, right? Now, the real cognitive verifier prom pattern will start from the here. Ask me subdivided quotis relative to in this main topic, which helps you to generate best overall output. After I provide answers to your subdivided questions, now ask me subdivided questions. So if you focus here, the way of context here, the way of writing the prom, you will get some reasoning style in this prom pattern here. That is called an cognitive verifier. Now you can see it is a simple task. How did World War two impact global politics? It is a simple specific task. You can see World War two, right? And the main task is impact global politics. Now, A will just give the answer for this one. How did World War two impact global politics? Now, it is simple question, Dia will just give the answer. Right? Now, when you write this Ask me subdivided is related to this main topic, what is the main topic here, World War two impact global politics. I understand these points. So what is it? Impact World War two impact global politics is the main topic here. What I'm telling to AI, ask me subdivided quotiens related to this main topic which helps you to generate best overall output. Now what we are describing to AI here, ask me the subdivided quotients related to the main topic that I have given to you. So what is the main topic here Worldw impact global politics? Then AI will ask the subdivided quotiens related to this World War to impact global politics. Instead of getting the bunch of output, all the information, you can gather the specific and required output from AI by using this prom pattern here. Okay, don't get confused by just comparing the Ask for input prone pattern and cognitive verifier pattern. There is a difference in that and we will explain in the few seconds. Okay? Again, see, ask me subdivided questions related to this main topic here. Now, which helps you to generate the best overall output? Okay? When I give the answers for this particular subdivided quotiens that are asked by EI for the main topic that is World War two impact global politics, then EI will give the answers according to my preferences requirements and the specific information that I'm looking to get from this main topic. Okay, you guys here, which helps you to generate the best orall output after I provide the answers. When I provide answers to this particular subdivided is that is asked by AI. Okay? Now, ask me subdivided quotiens. The summarization of prompt is, when you describe a specific topic to AI, specific task to AI, now AI should ask the subdivided questions of the main task. Okay, I will ask your preferences like what is your goal, or what is the information you're looking to from this particular main topic or task like that. When you drew the answers to that particular questions, then according to my answers required ns and my goal, it will generate the output that I'm looking further. Now, let's see what is the actual output of this prompt here. I'll just directly copy this prompt. I'll just go to PT and just past here. The scene. Got it to give the best oral output on how would World War two impact global politics? I will first break the topic into subdivided quotients. Once you answer them, I can create a detailed and structured response tailored to your input. Now you can see. Now here are the subdivided quotis. You can see there are seven quotiens When I give the answers to these particular questions, then according to my requirements, the EI will give the answers for me. Now, what are the difference between the asking me for input prom pattern here and this one? There is a similar thing here. You can see we have used the Ask me subdivided the condition as well. We are provided the last condition that is no ask me subdivided questions. You can see I provide answers to your. I also called some Ask me for input prompt pattern as well and the cognitive verifier prompt pattern as well. You can use the both, but you need to check the difference between that. Okay? What we are telling here, the questions to the main topic here. We're going to specific one. Right? Don't get confused, you will see the difference of these 21. Now, when you check the difference between the ask me for prom pattern and cognitive verifier prompt pattern, ask me for Improve prom pattern, the basic one, the cognitive verifier prompt pattern is advance one. You can see the detailed thing here. Questions related to the main topic which helps you to generate the best overall output here. And then ask for input prom pattern, it will asking my preferences, same like this one, but it is going in depth recording rather than the ask me for input prom pattern. The county to verify prompt pattern is going in depth of the specific topic. You can see main topic here. It will asking the ions related to the main topic in depth manner. It will asking my preference each and everything for this particular topic. In the previous one, it will just asking some simple things like what is the task about? Here, what is your requirement? The two things are same, but it is the county to verify prompt pattern is in depth when compared to the Ask me for Input prompt pattern. If you have the confusion regarding that, you can check by yourself. You can use the two different prom patterns in the same way in the same chat and you will get the difference in that. I hope understand this one. Focus on this cool to verify prom pattern first. You can see when I give the answers for this particular questions, it will just provide the output which is styled to my prefer SS. You can see power sheets. Which aspects intres you more the rise of the USA or USSR superpowers or the decline of urine coronal empress. I'll take the specific. I will take this one as a specific question like this. Take this one portion and next second question. I will give the answers and you can take World Bank. Second one question is, answer World Bank. Let's take the third one. Dynamics should include political let's take this political one. One is political one. What one D colonization. Do you want me to explain, let's take number five tional boundaries, Asia, let's take Asia. Number six, should I highlight how the use of forms. Let's take this ones. The seventh one, long term impact, you want me to connects. So when I give the answers here for this particular questions, it will generate the tailored and effective output of World War two impacted global politics because it is a history one, I will just combine all the things which happens in the World war two impacted global politics. It will just throw stones, it will explain all the things that is headed to me, right? So in order to get the specific one, what I am need in that particular history, I will just provide my preferences. I'm looking for this specific topic, and I'm looking for the specific preferences. You need to give the answer in that particular condition. Like that you can use count two verified prompt pattern, which you can use for the large broad amount of data. I hope understand these points. When you use this by yourself, you will get the idea how well it is important. This prompt pattern is. You can see when somebody World War two transform global politics, you can see the output here. Okay. I understand. It is a simple thing. You need to use this whole prompt, ask me subdivided questions relative to the main topic. You can take instead of the main topic, you can write the prompt. You can write the specific skill. You are looking to learn from the specific topic. Like that you can change by yourself, you can try for yourself for this, then you will get the best specific output according to your requirements like that. Okay. I hope you understand this prompt pattern very well. I'm giving an assignment for you. Just combine the four different prompt patterns like Ask me for Input prom pattern. So as a, you doesn't need to include the Ask me for input prompt pattern because it is already included in there and you need to use the personal prom pattern, question refinement prom pattern and call me to verify prom pattern. Combine these three different prompt patterns in one prompt and take one specific task and write the prompt and see how the EI will do the things which will amaze your mind. And we will improve your prompt writing skill according to your requirements. Let's dive into our fifth prompt that is Outline Expansion Prompt Pattern. Let's dive into that. 19. 4.1.5 Outline Expansion Prompt Pattern: Now what is Outline Expansion Prompt Pattern is? As the names suggest, you can see outline expansion. So what is that outline? It can be anything like eBook outline, content outline, or course outline or curriculum outline. All those things comes under this out and expansion. So what is an expansion? Expanding the particular outline according to our needs and requirements. That is simple. Okay? So the structured pattern contains like this one, intial prom setup, generative bullet point outline, interactive expansion, iterative exploration, and final output. What is actually this points refers to that? What is that entire prom setup? We'll just write simple prom. You can see the example here. Act as an outline expander, right? So it is a si personal prom pattern. You can see this one. Okay, you can see the real prom pattern statement. Generate a bullet point outline based on the input that I give you. So what happening here? We're just providing this particular instill prompt to AI. This is the first step, Isel prom setup, right? Now after that. Now when I provide, you can see here based on the input that I give you. What is the input? You can take any topic or you can take any skill that you are looking to learn, or even you can take anything that you are looking from the AI in the specific manner. I will give and then ask me for which bullet point you should expand on. I will comes under this one interactive expansion. Right? So first, we'll write the prom setup when I give the topic, then the AI will generate a bullet point outline. In that particular outline, we will just select any of that. You can see. You should expand on. Each bullet can have at most three to five sub bullets. Okay. You can see first, you can check the different books. You will see some different content section. In the content sys you can go the topic and there are some sub topics listed under the main topic. Okay, that are the sub bullets, right? Can see. You should expand on each bullet can have at most three to five sub bullets. The bullet should be numbered using the pattern, create a new outline. So what happening here? This is the third one that is interactive iterative exploration, fourth month. I can see, create a new Oline for the bullet point that I select. Okay, it comes under this particular fourth point that is iterative exploration. In the next one that you can set at the end, ask me for what bullet point you to expand next. Ask me for what to outline. Okay, don't be confused. It is some advanced prompt pattern. If you check this one very clearly and peacefully, you will get the idea about how this particular prompt will work. Okay. And we will just come to the habit and see how this will work in clear manner. Okay. So remember these 35 steps in order to write this outline expansion prompt. You need to set up a initial prompt setup that is this is one at a out and expander generative bullet point, that is all the This is the instructions which is very most important in order to use this Outline Expansion Prompt Pattern. After that, you need to provide the topic or anything task you are looking to get from the e. It will generate the bullet point outline. In that outline, you will just see which bullet point you are looking to take in order to generate some more subtopic of that particular bullet point. In that, you can go with the interactive expansion. You can expand the outline based pano the outline you selected previous, which is generated. After that, you will create a new outline, you can see, create a new outline for the bullet point that I select. It will comes under the ratio exploration. You will generate the more subtopics of the main topic again and again. I will become in the Iterative exploration. It will explore them more in depth knowledge of the particular topic by generating the bullet points. And final output, which particular bullet point you will just select it. You will tell just I need this quantent it will generate the final output. That is simple. Let's jump into ChatGPT and we will understand the real use case of this one. I'm the ChatGPT I will paste this prom pattern here directly here. You can see act as an outline expander, generate bullet point, all those things here. Now, I have just AI, I have just included something you can see. Follow below structure to generate outline. That is main topic. Even though you can just remove this one, let's check first the output is. Let's see this one. Now, got it. What topic would you like me to create the first outline for? You can see this one. No, as I said, you can generate a bullet point based on the input. This is the input asking for this. When I give the topic or anything here, then it will start creating the outline to me. Let's do the input for this. Prompt here. I will give simple topic that is advertising and marketing. Let's give this one. This is a topic, right. Now, it will just generate the bullet point. You can see this is a simple main topic, this all the subtopic, right? This is the main topic, this all the subtopics. You can see there are up to seven thing, but the subtopics have the bullets, right, not the number format. So what I will just tell to I, let's take this one. What I will tell here, I will just instead of taking bullet point number systems. Let's say I Apo or ten structure. Let's say, here's advertising. No you can see it is a good. No it is a structure we have generated the outline using the full number system that is introduction to advertising and marketing definition, purpose of marketing, difference and relationship between advertising and marketing. These are the main topic and these are the subtopics. Now it is well structured right now with the number system. So what is a major advantage of these particular outline expanderes? You can see the instruction here. Generate a bullet point outline based on the input that I give you. So we have given the advertising and marketing. Okay, then ask me for which bullet point you should expand on. No Cs ask which bullet point you would like me to expand first. You can see which number bullet would you like me to expand on first? No it has simple follow my instructions, you can see here. Now, each bullet point can have atmost three to five sub bullets. So we have seen here three to five subblatsO, three, four, four, four, five, like that. You can see. Now, the bullet should be numbered using the pattern, you can take the number pattern or all those things here. Create a new outline. This is the task here. Okay, create a new hotline for the bullet point that I select. Now we will select one point that is, let's take a method of market research. You can take the 4.2. You can take the whole sentence that is whole subtopic like methods of market research or even you can just tell to I 4.2. Now let's see whether it works or not. Now, great expanding 4.2 methods of market research, surveys and focus groups. No, it has taken that, you can see, 4.21, now, you can see, create a new outline for the bullet point that I select. Now I have selected just 4.2 in the previous outline. What happens here, you can see the task here. At the end, ask me for what bullet point to expand next, so it is done. Ask me for what to outline. This is the second task here. Now, you can see here. I have just selected the subtopic, 4.2. I have just written the number. Now it has again generated the sub bullets of 4.2. I hope you understand these points. These are the 4.2 methods of risk that you can take 4.2 0.1, 4.2 0.2, 4.2 0.3, 4.2 0.4, even, you can go like that. Even, you can go. You can see which numberul would you like me to expand on next. If I use this 4.2 0.4, in that particular 4.2 0.4, it will just create the 4.2 0.4 0.1 like that Eve like that you can go in the depth in depth information of the particular topic. Let's give this 14.2 0.3. I'll just take this one. Now you can see this one. It is just taking the in depth of that. 4.2 0.312. Again, it will ask which numbered blood point would you like to miss expand on next. If I give this 14.2 0.3 0.4, it will just take with the depth subtopics of this particular in person vases remote interviews like there. You will give the instruction. I will generate the so many subtopics of the main topic that you will select here. There is this prompt pattern will work, right? And not only that, even you can just tell to here just to break the chain. Now, what we can do? Now, just write the content. Four. This one. 42.3 0.5. What happens here. You can see. It is a simple content we have just get from this particular one here. That is you can see conduct interviews effect requires skillful preparation. This is how you can do a large amount of information. You can divide into so many subtopics in the subtopics. Even you can go the subtopics like that you like that, you can go in the in depth in depth knowledge of the particular topic that you can get some deep insights from EI for the specific topic by generating the subtopics, and when you want to get the particular content from for a specific topic, you can just tell to I now just write content and give the number. That's it. You can get the content from here. You can see which number bullet would you like. It will ask again because we are just writing the prompt pattern light, you need to ask for every output you need to ask, which bullet point you are looking to expand on next like that. That is simple. No, we need to use this particular prompt here. So when you are looking to write some E book or a book, you can use this prom pattern in order to get some idea how to write the contents, how to choose the main topic, and what are the subtopics, and you can research for research purposes, you can use this book in order to create the course and so much things, you can get the idea about all those things by yourself by using this simple outline expansion prom pattern. I hope you understand these points. Try by yourself and for the specific application. I'm giving an assignment for you just use the different prompt patterns up to we have learned right now like ask me for input prom pattern, question refinement prompt pattern, call me to verify and ask me for input prompt pattern and Outline Expansion Prompt Pattern. Combine these five prompt patterns in one specific prompt for the specific task or anything you want. Write it, try for yourself in the ha gibt or other AI two and see the output that will amaze you by using these five different prompt patterns, even you can create any web app or AI agent by simple writing the system prom in the playground of OpenAI. You can do so much things by combining this whole prompt pattern into one to create the amazing solution for any task to solve it. Okay? I understand this point. So please if you are looking to get better writing of the prompt unit to try by yourself, right? I hope you understand this prompt very well. Let's dive into our second part two, that is Advanced prom patterns part two, in which we will see the different fi another amazing prom patterns in the next session. Let's dive into that. 20. 4.2.1 Tail Generation Prompt Pattern (Advanced Prompts - Part 2): We have already learned that five different prom patterns in the Part one. Now in this session, we are going to see the advanced prom patterns, part two, in which we are going to discuss the first prom pattern that is tail generation. Let's dip into that. Now what is that tail generation prom patterns. You can see here to use this pattern, your prom should make the following fundamental contextual statement. Okay, to focus here, this is a simple method of this prompt, so we need to use in our prompt according to our requirements, right? You can see at the end, repeat Y or ask me for X. So if you focus here, so we have already learned about the cognitive verifier prom pattern or ask me for input prom pattern. You can see at the end, we will add the ask me for input X in there, ask me for input prom pattern there. In the tail generation, what we are doing here, we are using just we are adding the fundamental contextual statement, at the end, like simple repeat Y, for example, repeat some particular task for every output. That is simple. I will explain each and everything in the charge bet in few seconds. So to get the better understanding you can see here, you will need to replace Y with what the model should repeat. For every output you will get from the AI, you will tell you should repeat this statement for every output. Okay, you need to add this statement for every output at the end or at the top or are you want. You can instruct this one. So as a tail generation, tail generation means at the last point. So in that case, we are using this at the end, repeat some task or some statement or some passage like that or ask me for input text. So we have already learned this in the previous session that is ask me for input prom pattern. Now we are going to see the tail generation at the last Okay, you can see, you will need to replace Y with what the model should repeat, such as repeat my list of options and X with what it should ask for or for the next action like that. This statements usually need to be at the end of the prompt or next to last. Okay? You can focus on this one. It is so easy to use this tail generation prompt. Let's dive into our hachPT and we will see how it works. Now I will go to hachPT and I'll just paste this prompt. You can see here. From now on, okay? At the end of your output. Add this disclaimer. This output was generated by large language model and may contain errors or inaccurate statements. All statements should be fact checked. So by writing this particular statement, okay? For every output you will ask a question to AI, it will follow this instruction. For every output at the end, it will add this following statement. That is an fundamental contextual statement. Okay? Even you can write this anything according to what task and output of AI. Okay, in this case, we are using AI. You can see this from now on. So for every interaction through this prompt chaining, for every interaction in this prompt pattern, it will just repeat this particular statement up to here for every time that the AI will generate a output. You can see. Now, you can see, ask me for the first thing to write about. But it is a second condition. Now, the AI will ask me what I need to generate. When I give the task or when I ask a quotient to AI to do something, it will generate the output at the last end of the output. I will add this particular statement. That is, it can be anything. So in this case, we are using some note statement. That is, this output was generated by a large language model and may contain errors or an accurate statement like that. Okay. We can go beyond this. You can write anything you want. Okay, I ask me for input prom pattern, we will use two different conditions like I will provide so and so details, you need to do the thing. But here we are what telling in this tail generation. At the end, you need to add this particular disclaimer. In your case, it can be anything. Okay? At the end of this output, you need to do this particular task. Like that you can go whatever you want. Okay. Let's see how it works. We can see. Got it. From now on, I will include the disclaimer at the end of every output that is. So what's the first thing you would like me to write about? When I do the question or when I do the task, it will generate the answer along with it will just add this fundamental context statement or note statement at the end of the output. That is how we can call that tail generation. Let's do some task right. Essay. How about global warming. In 50 words. Let's take in 100 watts. This is a task I will give to AI. Let's give this one. Now, the AI will generate the output. You can see global warming is one of the players. This is 100 watts. Essay. No, you can see it is simple written our statement at the end of the output. You can see this output was generated by a large language model and may contain errors or inaccurate statements. A statements should be fact check. So this is a simple tail generation prompt. We have taken only for returning the statement. Okay, you can go for a different type of things like ask the simple is about the previous output or gather the feedback like that, at how the EI has helping you. So at the end of every output, you need to ask the feedback from user like that. You can do so much things by using this tail generation prompt pattern, you can use in any way in different ways. You will get the best output. And even you can build some amazing feedback system by using this particular tail generation from pattern. Not only that, you can change the difference between. For example, I will write I will ask you to AI for another task like, now, right, write a story about nature work? 10-year-old. Y. So this is simple task I will just provide to AI. Let's see this. Once upon a time, it will start generating the story, and you can see this is the fundamental context statement that we have Alien just told to AI to follow it. So for the every output, it will generate in this context, right? In this context window of this chair, it will just provide this one note statement at the end of every output. Now, if you want to break this particular step or you can use the simple forgot above no. Right. What happens here? The hagiby have the great capability that is memory update. Up to here, if you write any quotient, if you write any prom, it will give the output as well as it will just add the notes statement this disclaimer at every end of the output. To break this one, you can use this one as a forgot about or like that. Now what happens here? The chargib will automatically remove the memory. In the previous one, now it will just add current version like this forgot about no write. Let's try whether it will work or not. Forgot about no write. Five. Let's take 50 words. 50 words, essay on nature. Let's give this one. What happens? Let's check. Nature is a foundation of life. Ping this is a simple. No agency. It has generated same note statement. In order to remove this one from the context window of this particular chart, you need to specify the GPT not to add the disclaimer at the end of the prom for that just come here, forgot. From now onwards. Don't add disclaimer. We need to use this particular one. Let's use this one. What happens, the atbive will automatically stop to add in the disclaimer at the end of the tab. That is simple how you can write the instructions according to your requirements like that. I hope you understand these points. Up to now, we have seen just adding the statement, but you can do so much task. At the end of every output, you need to start the new task like that. I'm giving an assignment for you drive by your so it is a easy one. Center of writing, instead of using this statement, go with that task, just give any task to EI. For every output, it will ask a new task to start like that. Okay, I hope you understand this one. So like that you can use dich tail generation. For every end of the output, it will start the new task, like that. Okay. I hope you understand this prompt very well, and let's jump into another prompt pattern. That is semantic filter. Let's dive into that. 21. 4.2.2 Semantic Filter Prompt Pattern: What is a semantic filter? So it is a simple prom pattern you can use to remove any filters from the output or like that. You can see this use this pattern, your prom should make the following context or statement that is filter this information to remove X. That is simple. You can remove any duplicates, repetitive words or any other inaccurate information from the output as well. How you can do, you can replace X with an appropriate definition of what you want to remove such as names, dates or cost rather than like that, okay? So what is FA in this filter means you need to filter out the words which are not related to a topic or repetitive words or the effective words like that you can filter any of them by using simple prom pattern that is filter this information to remove so and so word or definition like that. Okay, let's hop into ChatbT we will see some examples. Come to ChatBT I will just take this prop, you can see, remove the Dilly expenses cost rather than $10. In the following my tell expenses. These are the expenses, the DL expenses. What I just tell you AI? Just remove the daily expenses which cost rather than $10 in my following daily expenses. Let's see the output. It will simply remove the cost daily expenses which are greater than $10. In this case, it will just remove the lunch $13. I will just give the breakfast $8 and dinner $7. You can see I remove the lunch $13 since it was greater than $10. So would you like me to also give the total after filtering? There is not one. You can go like that. Okay, let's see another example which help us to understand this one. No, pum, I will just take this prom. You can see here. Filter the following text to remove any personal identifying information or information that could potentially be used to reidentify the person. So what I am just to enjoy, you can see these are the name. Now it will work on this and company and it is mail of that particular person. So what I'm just telling to EI, filter the following test. So I'm just telling to EI, filter this following statement and remove any personal identifying information. It can be the phone number. It can be email, it can be the name of that person. It can be any personal identifying information, or information that could potentially be used to reidentify the person like that. What happens here, the EI will automatically remove any person's name or personal data regarding in the passage or anything. Let's give this what happens here, you can see here the text with all personal identifying information is remote. You can see an individual live ser address, it has removed redactor city, the work at reactor Company and can reach totally. It is simple just remove all the personal identifying information from his name, right, his address, and spin fees or tech crop like that. Even you can just tell to EI. Instead of personal identifying information, you can take this like any name, name of a person like that. So what appens here? Let's give this one. But if you see here, he leaves at a address, it is removed the name as well, but it also removed the address, city name and company of six because we have l to AI or information that could potentially be used to identify the person. In that case, to remove this one. Remove this one, then just give this one. What happens here, the AI will automatically remove the name of a person and in rest up there, it will just keep it as it is. You can see. Lives at 123 Maple Street, Springfield. I work for the Taco Company, and it is a mail. That is simple. You can see you also want me to remove personal identifier like email address, we can go with that as well. So it is simple filtering what's like that. Let's say another example. KensiFlter the following sentence to remove the repetitive information. Hi, how are you? I'm fine now, how are you? So if you focus here, there are some repetitive words that you can see here. Filter the following sentence to remove the repetitive information. Hi, how are you? I'm fine now. How are you? You can see here the filter version of the repetitive information remove. Hi, how are you? I am fine now you. Okay, that is simple. Would you like me to make it more natural and polished as well? It is simple sentences. It can be easy for you to just edit these particular sentences, small sentences. But what if I tell, there is a hundreds of pages of the whole passage, and hundreds of pages, thousands of pages have the content, which is have some repetitive words or inaccuracies, gramatal mistakes, filters, all those things. In that case, what you need to do, you need to use this simple filter prompt pattern by just adding the text. Now, for the hundred page of the content, how we can upload this one. You have the great option here. Come here, las button, add any photo and file from here, add just your document which have the content which need to be edited and need to be removed some filters like that. It can be 100 page 50 pages content like that. Use the semantic filter option and just tell TEI, remove the repetive information from the given PDF. Okay, and remove the personal identification or name, email address like that. You can do all those things by using chargeb itself instead of doing by yourself, and you can save a lot of time in proof reading or in editing the document as well. Okay, you have to understand these points. And another very important point is, before you give your own data to AI, make sure you need to set up some settings, come to the lefts corner side here. Okay, just click on your profile, just go settings and see for the data control, which is very most important. Okay, click on the data control and improve. You can see the option here. Improve the model for everyone. If your option is on, just click here, just make it off here. Just keep it off. Why? Because for every interaction you do with EI with your own data, the data is trined by the chargebty. There is a chance of showing your data to other people. So to avoid this one before using your own data, before giving your own information or any data to EI, just come here just off the data control option that is improve the model for everyone. It will avoid using your own data to try their model like Cha chit. I hope understand this point very well. Like that you can proof content pages website like that. You can paste the link as well, and you can do so much things by using this semantic filter option from pattern. I hope I understand this point prompt pattern very well. Let's jump it our third prompt pattern that is menu actions. Let's dive into that. 22. 4.2.3 Menu Actions Prompt Pattern: What is a menu action prom pattern? It is the most important prom pattern when you are looking to create an app by using Cha JBT or by just creating the prompt by converting the prompt into app. You can use this manuacion prompt pattern very well because it is like some programming language. You will set some rules. I will works under this condition only. Okay? This is the manuacionPm pattern which looks like, which works the same operation which will write in our programming field, right? So what happens when what we need to click? When I click SOO option, what it should be happen like that. It is a simple straightforward sentences, side forward instructions that I will follow in order to complete the task. Okay? It will work like simpletlist app like that. I will show in a few seconds. You can use this manu action prompt pattern to build some amazing apps, but just writing the prompt. Okay. Let's see how it works. To use this pattern, your prompt should make the following fundamental contextual statement that is whenever I type X, you will do Y. Okay. Like, for example, when we click some button in the laptop or phone, it will do some task, right? It is simple as that. It will just work like that. You can see option provide additional menu items. Whenever I type Z, you will do Q. So set of instructions. Whenever I do these things, you need to do these things like that. And you need to add this one very most important thing that is at the end, you ask me for the next action. Okay? These prompt patterns are nearly are similar when compared to other prom patterns like ask me for input prom pattern tail generation, they are all work similarly like that, but there is something small differences in all the function of prom patterns. Okay. Now we'll see this one. Let's jump it over Cha gibt in order to understand the MntansPm pattern very well, IhaiVT and I will just use this prompt. You can see. So I'm just telling to. Whenever I type add task, you will add task to my Tudleist whenever I type remove task, you will remove task from my Tudor list that is simple. Whenever I type list all, you will display all the task currently in my Tudor list. That is simple. At the end of every interaction, you will ask me for the next action. Okay? For every task that we done, after you instruct AI, it will again ask you for the next action. Let's understand by just giving this to AI. You can see, got it. I will treat your input as commands for managing your to list. If you use your task management app in daily life, similarity works like that only. Even you can use this CarGBt as a task listing app by using this simple prom pattern. Let's see how it works. You can see got it. I will treat your input as commands for managing your Todo list. Here's the current list we have stole from earlier. It is simple example it has taken. Booking a meeting with US my US clients at 5:00 A.M. Morning. These are the s task. Okay. Now, what is your next action? Add, remove or least. Let's take this one. Add some I will take this one add take meeting. Which you keep client at 5:00 P.M. Today. What happens? Now, when I use this ad here command, it will automatically update my Tut list section. Let's skew this one. You can see here is up to date to list options, booking a meeting with my US client going to office at 11:00 A.M. Meeting with UK client at 5:00 P.M. Today. Like that you can remove this one anything. I will just take this remove remove the fastest task. I will take this one. Number. I'll just tell to AR. What happens? It will just remove the first task like that. Now, I will just tell to A and list all the task. You can see here are the current task. So what I will tell to list all remove and current Task. What happens here, it will just give all the tasks that it has just remote and add previously. You can see current tasks are going to office, all those things. It is remote task. So as I said, the chargeba have the memory, which is a great cability of hagibt store your previous conversations, your ideas, your data, and it will automatically remember those things like it is working like a web app or mobile app like that. Okay? So you can do so much things by yourself, which is simple one application. Specific applications, you can do so much things by writing the prompt here. Okay? Now, you can change for different application. Now for the task, you can do for a grocery list app or expensive tracking app, but just writing the prompt. Before AI has just before the hagBT, we are writing some code in order to create app. But now with the help of EI, we are just creating the apps but just describing in the text format. With the help of prompting, we can build the amazing solutions by just writing the prompt. Now you can do these things all by yourself. This is a simple app. It is just working like task management app. Let's say another example. I'll just copy paste here. You can see whenever I type add expense amount, you will add expense to my budget tracker and update the total spent. Okay. Whenever I type remove expense amount, you will remove expense from my budget tracker and adjust the total spent. When I type summary, you will display the total spent and remaining budget. At the end of every interaction, you will ask me for the next action. Let's give this one. Got it. I have no track or expenses with the budget system. To start, please tell me what is your total budget amount. Now I will take example hundred dollars. What appens you can see perfect. Your total budget is thousand dollars. Currently, your spent is $0. A remaining budget is 1,000. What happens? I will just add this $250 spent. Let me soil. Spent on shoes. What happens here? Now, you can see, add it. Shoes 250 spent, total is spent to $50. The remaining budget is $750. Okay? That is simple. That is here, spent shoes, how much amount you have spent it is added. Okay? Now you can use the remove option you can remove the total spent to $50, add extra to $50 like that. Or you can ask the summary of the budget. You can take all those things by yourself. Okay, now you can remove the previous spent and Alpha hundred dollars like that. What I can say, remove previous spent amount and add extra $500 to my budget. What happens here, you can see here. Done. Remove $250 spent on shoes, increase your budget by $500. So what is the total cost, you can see? Remaining budget is $1,500. Like that you can just play with the Cha GPT by your instructions by creating the amazing app, you can do all those things by this menu action prom pattern. You can do all those things by simple using this prom pattern. Okay? Now, what you can do, you can tell to AI just a summary, summarize all my expenses. What happens, you can see here is your full expense summary. Total budget, $1,500, total spend, remaining budget, $1,500, expenses history, choose to 50 removed, currently no active expenses like that. Like that you can train AI to do some particular task. No, I'm giving an assignment for you. So please just go use all the prompt patterns up to we have learned. Okay, along with this men options prompt pattern, combine all the prompt patterns together in one prompt for the specific application to solve a particular task. Just try yourself, see the output. If the output is very amazing or some problem solving game changer, remember, you can build WebApp by using the same prompt here. This is a menu actions prom pattern. You will just tell to AI you will give the command to do the things. Okay? It will automatically done you will task according to your requirements and commands like this one. Okay, I hope to understand these points very well. Now let's jump into our fourth prom pattern that is Pat checklist prom pattern. Let's dive into that. 23. 4.2.4 Fact Check List Prompt Pattern: So what is the fact checklist from Pats? We can see what are the fact check here? So as we earlier discussed, the EI models like a GPeana other EI models are trained by a large amount of data, right? So in that returning the response, according to the user question or prompt, the output can have some inaccurate data, inaccurate data or some hallucination words or inaccurate information like that, which have some mistakes, all those things, right? So in that case, how we can check the generated output is correct or not. Okay? By using this fact checklist prompt pattern, we can see the correct whether the output have correct information or not by seeing the facts in the output. By analyzing the facts in the output, we can easily tell that the generated output is the accurate output. We cannot get the 100% accurate information from AI, but we can get some nearly approximately 90% or 95% accurate information from AI according to type of model you are using. It is a GPT, 4.5 or five mini like that. Okay. Now, you can see. Let's see how this will fracture stromPattern works. You can see this is a statement we need to use in order to use this pattern, you can see, whenever you output text, generate a set of facts that are contained in the output. The set of facts should be inserted at the end of the output. The set of facts should be the fundamental facts that could undermine the veracity of the output if any of them are incorrect. So what is the meaning of that? Whenever I just ask a question or prompt to AI, it will generate some output, right? So along with that output, it should separate the facts, separate facts, set of facts from the output, and it should be inserted at the end of the output. That is simple. Now, what are the facts? The fact should be fundamental facts that are the roots of that particular information. That is the fundamental knowledge of that particular information. Based upon that facts only, the AI has generated the output. Okay, if we analyze that facts whether these facts are correct or not, now we can rate the AI's output. It is correct or not. Simple. Okay? Let's jump into harGPT how it works. How come the ChVTJt his prompt here. No, Cs it is a simple question. See, write a brief summary of the causes of climate change. Okay? It is an of mine. You can see at the end of the output, generate a set of fundamental facts contained in the output. Okay? This facts should be fundamental to the summary and the inserted at the end of the text. Ensure accuracy as incorrect facts would undermine the validity of the output, whether the output is valid or not, so we set we need to check the facts of the output. Then we can rate the AI's output is valid or not. Let's give the question to we can see. It is simply summarize the climate change causes of climate change like that. You can see this is a simple summarization of the climate change. You can see these are the fundamental facts. Okay? You can see these are the climate change largely caused by human activities, greenhouse gases strike, carbon dx, SH war, two trap heat in the atmosphere, burning fossil foils, deforestation, natural factors. These are some fundamental facts. These are the facts. Which help EI to generate this output. Okay, I hope you understand these points. So what I need to do instead of checking, ins of analyzing this whole paragraph whether it is correct or not, I can simply verify these facts, okay? When I verify these facts, whether these facts are correct or not, if I verify these facts are correct, then I can say this output is similarly correct. Okay? If the facts are not correct, then I can say this output has some mistakes like that. I understand these ones. So this is a simple thing. When you go for the large amount of output, you can simply separate you can use this prom pattern in order to analyze your output, whether the EI output is correct or not, or is there any mistakes in this particular output? You can check by saying the fundamental facts. When you check these facts, if they are correct, then this output can have chances to be the correct. These facts are wrong, then you can say the output of this AI can have some mistakes. Like that, you can verify all those things. You can see if you start analyzing all these sentences, it is a simple paragraph, but if you try to create the content more data, right, more text, you will waste a lot of time, right? By just using this fundamental facts prom pattern, we can easily remove you can separate the facts. You can just verify each and every fact and you can get there and you can analyze the output whether it is correct or not from AI. I hope you understand this point very well. So this is how we can use a fact checklist prom pattern in order to verify the generated output is correct or not based upon the fact. So when it comes to verifying the facts, you can just copy here. You can search it in the Google or see the videos in order to analyze the fact which is correct or not. No. So that's why. Now, the prompt engineering is not only the writing, the prompt, getting the output. Okay? No, the real prompt engineer is the one who can analyze the output and refine the prom again to get the best output from the EI again and again up to, we can get the output, which is 100% correct. Cannot see the AI will generate the 100% correct information, but we need to try by ourselves. So how you can analyze this prom, how you can analyze the output of the particular prom that is correct or not. So in this case, you need to have the specific knowledge. That's why the prompt engineering not helps for everyone. The prompt engineering helps for the specific people who have the specific knowledge on the specific topic. For example, if you are the math teacher. Okay? Now, you have all knowledge, basic knowledge, some advanced knowledge of mathematics. Now, you can use ajibTT save the time. You can ask any question. Now what happens here? For that particular math problem, the hajb will give the answer. Now, the answer have some mistakes, and you will analyze it and you will rectify it because you know the mathematics. Then you can easily write the prom, the chargeb will give the answer. No, you will rectify it. Okay? When you rectify the output by using this fact ulsh prom pattern, you will get the idea. Okay, the EI is doing the wrong thing here. Then you will use the hagibt as motley, not just giving the question and getting the answer. It can be have some misleading information. It can have any accuracies in that, lot of mistakes in that. In that case, the usage of AI is wasted. Okay? So this Hagibi other AI tools you need to use as a time saving tools, but not as a replace of your Dai task. I hope you understand these points very well. These prompt patterns are well defined for the LLM, not a specific tool. Okay. It is not only for the Chachi Bit, it is for all the LLMs. It can be the Cloud, Chachi Bit or other AI tools as well. Okay. I hope you understand these points very well fact checklist prom pattern. Let's dive into our fifth prom that is chain of thought prom pattern, which is very most important. Let's dive into that. 24. 4.2.5 Chain of Thought Prompt Pattern: Now, what is a chain of thought from patternes? You can see a prompt designed to guide the AI through a step by step reasoning process before arriving at the final answer. This is similar works like achieving a big goal by dividing the goal into smaller chunks in which we can easily complete that task in small time of internal and we can like that we can complete all the task and we can achieve the major goal like that. Okay? This chain of thought prom pattern is very most important and very useful prom pattern in order to solve the complex problems. You can see here, a prom designed to guide EI through a step by step reasoning process. Okay, what is what happening here, step by step reasoning process. We're just trying we are guiding AI to solve a particular task step by step by analyzing each and every step, before arriving at the final answer. Why use it Idle for complex problems requiring logical thinking or multi step solutions. By prompting the AI to think out loud, you can often get more accurate and insightful responses. Okay, that is simple. So let's see the example how it works in the chargeabi. Now, I just take the sum problem. You can see here. You are solving a math problem. A train travels at 60 kilometer per h for 2 hours and then at the 80 kilometer/hour for 3 hours. So what are the total distance traveled? Breakdown your reasoning step by step before providing the final answer. That is simple question here. So what happens here? Instead of giving this one, let's give this one here. Instead of using the chain of thought prom pattern, I will just take this as simple task. You can see, let's solve this step by step. Even if I don't tell AI, go in the step by step, the chargebra is automatically taking and thinking as a reasoning purpose. You can see, let's solve this step by step. Step one is used a distance formula, you can see this is system formula, car the distance for each part of the journey, all those things. Okay. Now what I will do, I will just paste this one and I will change the values in order to get the 110 and it's taking the 118. Let's skew this one. Now, what happens, sir? In the previous adjust, I don't use the chain of thought prom patent statement here. In this case, in this prompt, I will use this branden your reasoning step by step before providing the final answer. I'll just include the chain of thought prom pattern in this prompt here. You can see sure let's solve this step by step. Step one, recall the distance formula. You can see distance equal to speed into time. Step two collet the distance for each part of the journey at the distance together. You can see the difference between here. Right? There is no difference in the previous one and second one. Why? Because the JAGPt is getting smart enough, okay? Now what happens? It is a simple question. It is a simple problem. But it is where you go for the complex problems. Okay? When you write the task, when you write the for complex task, you need to use this reasoning because in the reasoning part, you can also check the process of solving Okay. That reasoning can help you to cross check the output of the AI. In the step by step, you will get the best idea how the AI is callating the solution, how it is solving all those things in order to save what time. Okay? So I've just taken for the simple example, we can go for the complex problem as well. It is up to which type of Jagi model you are using if you are using latest version of Jagt model, so it is automatically smart enough it will just when you give any problem, any physics or mathematics, any problem even if you're not write the breakdown your reasoning step by step before this chain of thought prom pattern, it will go in that way only because it is trained well trained enough. Okay? But the specific for solving the mathematic problems, all those things, I will just go in the step by step reasoning purpose only. If you are using the very old mon of the ha GPT 3.5 or like that, you need to specify this one separately like a breakdown your reasoning step by step before providing the answer final answer. Just try this same thing. I just try this one in the other AI models like cloud, all those things. They will give the answer. If your output is not in the step by step process format, you can use this breakdown your reasoning step by step before pouring the final answer like this one. Okay, that is up to you how you can use this prompt pattern in different ways when you require. Okay? Now, this prompt pattern is not only help in the solving the mathematic problems, physics problems, not only like that, even it can solve so much very business problems as well. So where there is a reasoning step required, you need to use the chain of thought prompting. Okay? Where some sort of thinking logical thinking is required or some effort, which should be done or when we have some complex situation, you can use this chain of thought prom pattern in order to reason in order to see all those things where the deal works in step by step to understand the problem as well and how EI is solving the problem as well. By using this chain of thought prom pattern, you will develop amazing skill that is art of analyzing the problem and art of solving the problem. When you learn this chain of thought by using Cha GPT to solve a particular problem, you will start learning the solving a problem solver. You can become a problem solver. You can use a ha GPT to become a problem solver as well. I understand these points very well. So this is now up to we have seen the ten different, most important prompt patterns which saves you a lot of time. And not only that, it can help you to become problem solver, much more things by using these different prom patterns. Now I will giving an assignment for you now up to now, we have seen so many prom patterns, recall and try to combine all the prompt patterns in one prompt for the specific problem solving, write the prompt and try to solve any problem. If this done, now you have the great skill to write any prompt for your requirements. When you try to learn this writing the prompting skill by using the different prompt patterns for different applications, you will realize there are a lot more app ideas you can generate. You can create, you can develop and you can launch in the market in order to make some value by using this prompt engineering skill. So up to now, we have just completed the prompt different prompt methods using HGPT. Okay, let's jump into our fifth model that is specialized methods of prompting. Let's dive into it. 25. 5.1 Prompt Chaining Technique: Welcome back to model number five that is specialized techniques in prompt engineering. So in the previous session, we have just completed the different prompt patterns, and we have learned so much things in writing the different prompt patterns for the specific applications and how we can use it for different ways. Right? Now in this model number five, we're going to see the specialized techniques in prompt engineering. So we will discuss in this mod number five that is what is prompt chaining, how the prompt chaining works and some applications, we will see how we can write for the different use cases by using harGPT by writing the prompt chaining method. Okay? We will see all those things in model number five. Okay. Before going to charBT and we will just look at what actually prompt chaining is. Okay? You can see why you use prompt chaining. You can see some tasks are too complex for a single prom. Sometimes the prompt engineering is nothing but writing the prompt and getting the output in one time. That is not the prompt engineering do. The prompt engineering is nothing but writing the prompts for specific application and refining the prompt according to the output from AI, and we need to rectify the output, which is correct or not by using fact check list. And often that if the output is satisfied, right? Then we can go with that prompt. Otherwise, we need to write the again prom that is called follow up prompt. These follow up prompts are called prompt chaining. So this prompt is linked to the previous prompt, right? So for the specific application, that we can change the output of previous one in order to get the oft and required output from AI. That is called an prompt chaining. Prom tening is nothing but connecting the prompt, writing the follow prompt, refining the prompt according to the output to solve some specific problem. It can be the complex or it can be the small problem. Okay, you can see some of the applications as well. So tasks are too complex for single prompt, for example, writing a research paper outline, developing a marketing campaign, solving multi step math problems as well. So as we are discussed in the chain of thought prompting, so it also works like a prompt chaining as well. What is this one? Multi step path problems, what is this one? Not only the mathematics problem, even you can take any business problems or anything. So instead of writing the whole prom at one time in the hagibt or other AI modulus, we can write the step by step prompt in order to get the best output from AI in deep insights. That is called prompt chaining. Okay, I hope you understand these minds. This can get understand by writing the prompt in the hag Bit. We will look all those things in a few minutes. But before going to the Chachi Bit, you can see by breaking the task into smaller parts, you get more precise and coherent results. Instead of writing the one prompt one time, the AI will generate the output. But that one particular prompt, you can divide into smaller parts. You can write the prompt after the checking each output for the specific task. That is smaller task. You can refine the prompt according to the output and you can get the best deeper insights from AI. That is called and prompt chaining is. We have some steps you need to follow in order to use this prompt chaining. You can see start with a general prompt. It can be anything. Okay, we will see all the applications, how we can write the prompt, all those things in a few minutes. Start with the general prompt, refine each idea. So what is called a refine each idea. So when you write the initial prompt, you will get the output from AI. So you need to refine that particular output. Okay, what it may be the prompt, I will generate the output. You need to refine it, whether it is correct or not, whether the output is have some inaccurate information or not link that. You can do with the fact chat list prompt pattern as well. After that, you need to iterate based on feedback. So what is this one? According to the feedback of the output from the previous prompt, you need to write the follow prompt. Which helps AI, even to understand the better insights, and it will generate the best output for you according to your requirements. You understand these points very well. And let's jump into hagibty and let's see how will the prom chaning works according to our requirements. Let's jump into that. Here in the hagibty you can see it is a dark mode of the hagibty. Let's provide this prompt. You can see this is a simple prom that I have taken. You can see you are a experience, you are an Experience in marketing, especially in running social media campaigns. Now your task is to generate hard copy, social media post, video ad content, target audience, budget recommendation for Facebook ads for selling digital for mens only. I will tell you. So first, remember these things. So I am just without using the prompt chaining method, I have just given the task to EI in order to complete. What is the main task of AI? You can see to generate a ad copy, social media post, video content. So these are the different tasks that have combined in one prompt to do the task. So it is only one time prompt here. It is not called prompt chaining. Okay, let's see the difference between that. After that, you can see. I will tell you which task should be done first. So before we will just take this one. You can see. So this whole task that I'm looking to do with the AI, let's give the Prompt to the AI. Let's see the output. You can see got it. Let's put together a complete Facebook cats campaign for selling your men's detail watch lg everything step by step. Now, you can see it is add Copy, social media post, we add content, target audience. So now if you focus here, right, if you observe. So I have got the answer from AI for all task. What if I just click on edit this one. What if I tell you AI? Just know your task is to generate ad copy only. Okay, I will just remove this one. Let's see. No previous prompt, I have just given all the task in one prompt. Now, the AI generated all the answers according to my instructions. You can see. Now I am just giving one task one task to AI. Let's see the answer. Got it to generate the best performing ad copy, I will need a few details from your first. Now, when the previous one I have just asked to AI to generate campaigns, AdCpy, social media post, all those things, the AI has not asked me the questions. Okay, He just given the output, you can see this is the output. Now, when I go for the specific man, then AI has just asked me my details. It is asking my preferences. When I give the answers for these particular questions, then it will generate the add a copy, according to my product, target audience, all those things. I hope to understand these points. So this is why you need to go with the specific M. This is how the prompt chaining works. Instead of giving you all the task to AI at one time, Okay, you can go. You can give the task per one task only. You can give the only one task to AI in order to complete that very effectively by deeper insights instead of taking instead of throwing the stones like that. He understand these ones. When I give the answers for these particular questions, it will just give the best ad copy according to my requirements. You can see, would you like to provide these details so I can craft added R copy. So I create a few general example across different tones as well. So you can do the answers like one equals to my product or services or selling watches for men's only target Audience, a gender and men like that you can go, then you will get the answer. That is simple. Now, if you focus here, that is how you can go for the specific man. Now we just click here. Now, you can go for the second task. You can add social media post. Let's see, two tasks what happens here. Got it to make the best AR coop en should made a post, I will need a few details first. So when I give the answers for this again, it will generate the output according to my details. So this is how you can check all the prompt patterns, prompt chaining here. The prompt chaining is nothing but instead of giving all prompts in the one prompt, you can divide that whole prompt, whole bunch of task into specific task, and you can use that prompt chaining in order to get this solve a specific task with deeper insights instead of getting all the answers from AI like it looks like a summarization. But we need the deeper and effective output. Further, you need to go the specific manner. You can use this prompt chaining as well, right? Now we have seen how we can go for the specific one. Now, let's see the main prompt here. So I'll just paste it again. No, you can see. Now, you are experiencing market especially in the running social media. So what I have tell you AI, it is all the task to AI. Now, you can see this is a prompt pattern we are going to use. You can see I will tell you which task should be done first, then you need to proceed. Ask me which task you want to generate. So what is this one? So we have already learned this one for the Ask me for Input prompt pattern. Please recall it if you forgot it. You can see I will tell you this is a first condition of Ask me for input prompt pattern, and this is the last condition of input prompt pattern. Now what happens here, let's say? Got it, you want me to act as expensive marketer for selling messil watches through Facebook ads, and I will generate exactly what it was whether that ads copy, social media, video content or like that. So which task people would like me to generate fast. So in the previous one, we have just given the given to AI, no we task to generate dad copy, the next one we have seen, and we have added another task that is social media post. In the fourth one, we have just given all those things here, right? We have just given all the task into one prompt. No, we have just added the Ask before input prompt pattern in which we have just retained. I will tell you which task should be done first. Then you to proceed that particular task and ask me for which task you want to generate first. Okay? Like that, when I give any task to complete here, let's take this like video air content, right? I will take this one. I'll pass this particular task here. Let's give this one. Now you can see, here's a puppet. Let's build the video add content for your Men's ditail watch campus. It is a video ad Script. You can see this is a whole AD Script, you can see this one. So here, you can go with the video ad content or even you can go with the social media campaigns, add copy like that instead of here. It is a task. We have just given the task here. Then it will have just generated all those things here. This is how the prompt chaining works, right? Even you can change the different task, it will just done this task. Now, let's focus here, first one. Now you can see. In the first one, we have just given all the tasks to AIT in order to compute. Now you can see this is a simple answer for every task. No, you can see video content is small, slow right, very small when compared to this one when we go for the specific one. You can see this is a video ad script and mens digital watch. Now you can see this is the whole video ad content, and this is the first one we have just assign the whole task for AI at one time. You can see the video content for the first prompt. Compare those things. You will see the power of going specific with the prompt sharing here. It's a simple giving all the task, you will get there. Output, when we go for the specific one, it will ask some is or even it can generate the output, and it will just ask the same. In the fourth one, we have just connected all the task in one time and have just used and asked me for input prompt pattern in which the different smaller tasks are combined here, and we can access any task by just giving the task here. Or in the task, it will generate the answer. I hope you understand these points very well. This is called prompt chaining here. You need to assign your roles and you need to assign a task. After that, you need to check whether the output is correct or not. If it is correct, you can use it. Otherwise, you can we please convert above this add content into engaging or like that, you can go. You can write the follow up questions. You will get the best output. Instead of getting the all output, you can go specific output with deeper insights from AI. This is how the prom channing works. Okay? No, not only that, even you can write here, Ask me, I will tell you which task should be done first, then you need to proceed. Ask me for which task you want to generate. Now you can write this another prompt pattern as well. That is C to verify prompt pattern. Let's give this one. Ask me subdivided. According to task that I have. Which helps you to complete that task. Let's see what happens here. Got it. We will move step by step. You can see this simple six to seven line prompt is working like a web app. Okay, you can see, got it well moved step by step first, you will tell me which task you want to generate. When I select any of these particular task like let's say social media post. Now I will just fit this task here. What happens here, the EIA will ask me some questions according to this particular task. Because to generate the personalized social media post according to my requirements, according to my preferences, instead of just giving the output according to my task, it will asking my preferences. This is how the prompt patterns all work well. You can see, we have just as we have using the small sentence. You can see ask me subdivided is. Now, you can see this whole prompt chaining is working like an webp which helps you to generate the different content according to our requirements. When I give you the answers for this particular requirements, it will generate the best social media post according to my requirements, which is the effective one, which is the personalized one and which makes some value in my content. That is simple. This is how you can write the different prompt patterns combining for the specific application with the prompt chaining and different ways, you can build amazing solutions as well. You can change your content with the AI as well. Okay, let's try by yourself, go with your own task and try different prompt patterns for the specific one with the prompt chaining works with the prompt chaining method, all those things, try by yourself, and this is how you can learn the prompt writing skill with the AR. I understand these points very well. Now, let's jump into our prompt engineering for the different applications, how we can use Chagbt to done the different use cases and for the different applications like prompt writing, image prompt writing or other. Okay, let's dive into that. 26. 5.2 Prompt Chaining for Different Application Use Cases: So to, we have learned what is a prompt chaining and we have seen the use cases using the HGPTH we can write the prompt by using prompt chaining method. Okay. Not only that, in this session, we are going to see some different prompt engineering applications and we are implementing the practical in the HGPT and how I will approach to AI for the specific application like creative writing, coding, marketing content, or customer support, IEQ Generation, or image creation, prompt creation, how will approach to AI? I will show all those things practically in the hagPty in this session. Okay. Let's dip into that. You can see as we already know, we have discussed some applications of prompt engineering in the previous classes as well. You can see you can use the Char GBT or other AI models for any type of task, but the real problem is you need to have that great or effective skill or writing the prompt. I will show all my secret approach or I will approach to AI from the starting point to ending point for the specific task to complete it and I will try to cover all those things in this session and as upcoming sessions as well. Now, let's jump into Cha GVT and I will show how I will approach to AI while I'm looking to get some output from AI for the specific task. Let's start in the Chachi biting, and I will start with the hi. No. Hi, how is your day going? That is well. Now what I'm doing, I will just using this helpful assistant prompt at the intell stage. So you can see here, U a can say you are a helpful assistant, you will do what I will tell. Please focus here, this is a simple intial prompt we need to use in order to get the best output. In this whole this context window. I hope you understand these points. When you start with this prompt pattern or prompt, helpful assistant prompt, what happens here? It will start giving it will the upcoming all the prompts and task will work under this in SAL prompt. That is why the hagibt have the best capability. That is it will remember your previous chat, answers, all those things because it has some memory future in the hagibt. Okay? We will cover all those things in this session. You can see you are a helpful assistant, so we have just assigned a role to AI. You will do what I tell. So we have just given the Ask me for input prompt pattern condition album. You have experience in detecting unusual words. Inaccurate information, and you will generate best and effective output without any mistakes and hallucination inappropriate information, or you understand. So we have just giving that task to EI initial stage of the chatting interaction with AI. We have just telling to I, so you are helpful assistant, you will do what I tell. So even if we don't tell, so it will generate the output according to our input only. But by giving the extra instructions to AI, it will just work under these conditions only. In that case, we can try to mitigate or we can try to reduce the inaccuracy in the output. I understand this one. You can see. We have just giving the AI a simple assigning a role helpful assistant and you have the experience in detecting unusual words, inaccurate information, and it will generate best and effective output without any mistakes, hionization, improper information, or you understand. That is simple. This is an initial stage prom that I use. For every task that I will try in the chi or other AM models as well. If you don't use, there is no problem in that, but if you use to build, if you're looking to get the specific information from AI or to learn some skill, you can use this particular sure stage helpful assisted prompt. It is very helpful to you. No. C, I understand. I will carefully check for unusual words, inaccuracies, mistakes, hallucination, and only generate accurate clear and effective outputs for you. Would you like me to conform back and understanding each time before giving final output or just directly give the best is? You guys it is asking me to conform back my understanding each time before giving the final output or just directly give the best version. So based upon that, you can set a condition. So before giving the final output, you need to show me which type of output you are going to give to me. That is simple. That is how you can control ChatGPT for your requirements and for your task to get the best output without any honization inaccuracies in that. I understand these points. Let's start with some use cases. Okay. So I will show some one, two, three different use cases, how I will just use this ta gibt or other AI models to do the task. Now, on the top of that, you can do by yourself as well for better understanding all those things. Okay? No, I will just go for specific tasks like just I will do the task. I'm looking to sell my own digital watch for 20-years-old boys. Can you help me to go online? So I have some mistakes as well, and I'll just given my requirement. I'm looking to sell my own digital watch 40-years-old boys. Can you help me to go online? It's very simple I have just taken for the example purpose. Let's see what the output will be. Searching for, I can help you start sending your digital watches online targeting 30-year-old boys. I will outline step by step plan covering what you do before launching all those things you can see. It is showing simple, you need to define your product audience, early legal business setup, choose platforms to sell. It is all those things, you can see the output as well according to our requirements. It can be anything task, right? Now, what I will do instead of going all those things one time, I will go only for specific task in order to get more information in the particular one. I can go further defining your product audience, clearly, all those things. What I will tell you how can I define that's two? How can I define my audience for my product go in depth. I'll just go with this one. I'm just asking the follow up questions in order to get the best output for the specified one in the in depth information of the output. You can see I've just asked to AI, how can I define audience for my product go in depth? Great quien. Defining your audience deeply is one of the most important steps before selling your watch online since your product is digital watches for old boys, you can see demographics, basic. I need to see this particular one in order to sell my watches for old boys. You can see psychographics, lifestyle interest, I need to see their users psychology lifestyle, all those things, behaviors, pain points, ideal customer persona example, all those things, I need to see. You can see this is a great output from AI when I go in the specific manner, instead of getting all the stuff from AI, I'm just going into the specific manner, which is we have learned in the prompt chaining method. You can see if you want, I can create four different orange personas like College Jim Boy. No it is asking me to go in the specific one. Instead of the 20-years-old boy, you can go for the call it gym boy or specifically for the tech gamer boy or fashion focus or like that. Because in that market, we can easily sell to them because it is a personalized for personalized boys. That is simple. This is how you can go for this follow up questions. You can ask for this one, all those things. Now even I can go by taking this one of the task here, how can I solve the pain points? Let's take this one. I'll just tell to how in. Solve. Pain points. By selling how can I solve pain points? That helps me to sell mine watches. So I'm asking again, the follow up question in order to get more information for the specific topic. You can see that smart angle. Solving your orange pain points makes your product stand out on drives, all those things. You can see pain point is expensive watches are out of budget, pinpoint boring designs, low quality, limited futures, lack of trust in new sellers brands, all those things. When I see these particular points, when I try to reduce this one, when I feel all those things, their pain points, then I can show all the pain points in my ad copy or the social media to sell my watches online. It is simple, right? So this is how you can go for the specific one. According to the output. You can refine, you can see that each idea and the output, you can try it by yourself. So this is how you will approach all those things, okay? For better and effective output, even you can use the we have always in the previous session like you are experienced advanced marketer in which you have the ten years of experience in the selling goods in online, like that you can assign a specific role. You can use the different prompt patterns that we have earlier discussed in the previous sessions. It can be the cognitive verifier, it refinement like that. So you can all combine it in order to create one prompt we can assign a task to AI in order to complete that task effective manner and in the accurate manner like that. I have just taken some example I have shown how I just go with that. For every interaction to AI, you can just come here. You can try to write URA experience. You can use Ask me for Input prompt pattern as well, Persona Prompt Pattern or to verify prom pattern in order to get the best output from AI. Not only that, for example, I will take another use cases like that by using the person of prom pattern, all those things. I'll just take here. You can edit the prompt directly here. I'll take R and experiential digital marketer and your task to complete even task by me. Ask me task that you will do it in your field of marketing or digital marketing. And even you can give the background information like you have the ten years of experience like there, and you have so many you have achieved so many you have helped the businesses or individual freelancers or the business owners to grow their business to the next level. You can write all the background information to AI in which the AI will think in that form, and it will help you to generate the content in that particular way and in that particular field that you can see the output and you can compare and you can use it in your w tally life. Okay, I'll just going with this option. So you can try for yourself, add the specific attributes to AI like your experience and what you're looking for that. You can try that EI. Okay, if you are looking the best engaging content from EI, then you need to tell to AI you are experienced digital content writer in which you have the ten years of experience in writing, engaging content like that. So if you are looking from AI, you need to give that particular attribute to AI with some experience or with some background information in order to get the best output for your requirements. I understand these points, well well, and I'll just go with this one. Not just ask me the task to me, you can see, I'm ready to act as your experienced digital marketer. Which specific task digaal marketing would you like me to the first. What I'll do I'll take any of this particular thing you can see a copyrting that we have already seen in the previous one. This is what similar like that. I'll just go with the Xu optimization. I'll just give this task one. I can see SEO optimization, you can see, great choice. FO Optimization, I will need a few details first so I can create the most effective strategy for you. We can see it is asking me the preference quien requirement quotient that when I give the answers particular these questions when I provide the data that I is looking from me, when I provide that it will just go with this SUO Optimization task. Even though when I don't use the county verify prompt pattern that is ask me subdivided quotis, it will it will just asking the quotients. That is how the AI is changing, the AI is getting smarter enough. Even though if you don't use some prompt patterns, according to your task, it will just take your task. I will work like personalized to somebody. I have to understand this much. When I give the answer, I like business website type, it is a commerce blog, service spacer or portfl like that where the target audience. It can be professionals, local customers like that. You can go main goal is conversions, and you have a competed website, you can give the name as well. When you give the answers, it will just complete the SU Addit checklist or Kw plus content will do all these tasks according to the task you have given and the information you have provided to AI. That is simple. This is how you can use entging telling to, so just write the SUO optimization or complete the SU audit for my website. You can give the link. It will give the output, but it is not as very effective when you start interacting with this particular helpful assistant prom pattern and assigning your role with some background information, attributes, all those things, you will get the best output. When compared to the giving just your website link and just asking a please complete SU dit checklist. So there is a lot more make your output difference with the different prompting, right? Even though you can just write for the specific people, that is you are a helpful digital marketing assistant, like even you can go for the specific one. Okay, you will do what I tell in the digital marketing. You have experience in detecting unusual words, inaccurate information, and it will generate best and effect to output for digital marketing task without any mistakes, ionization, improper information, or you understand. It is a helpful assistant only. You can go for the digital marketing assistant or customer support assistant like that you can start. You can do for the specific context window for specific application by starting with the helpful assistant prom pattern and after that, you can start, you can continue with your task for the specific application to get the best specific output from AI. That is up to this particular thing. I'm giving assignment for you. You can just go with this prompt one, write the output, and after that, you need to write the specific helpful assistant prompt, try by yourself, and you will see the amazing output when compared to these things. This is how you will need to write the different prompt patterns, prompts. You need to check the output. Based on the output, you need to select the prompt, which is help you to get the best and effective output for your requirements. That is simple, right? So in the next session, we will see how we can use these prom patterns to generate the prompt and image prompts as well. Let's dive into that. 27. 5.3 Writing Advanced Text Prompts using Prompt Chaining Technique: The previous session, we have seen how we can write the prompt for different use cases by using prompt chaining. So we have seen for the digital marketing task. Now in this session, we are going to see how we can use the prompt, how we can write the prompt with the different prompt patterns to generate the prompt for text to prompt as well, and image prompt as well. Let's dive into this in this session. So now we will take the previous chat only in order to get the best output. I will just take this. Okay. Now it is our simply interacting with AI. You can see this if you are helpful assistant, we will keep the same as it is. No, we will change this one here. Come here. I will just go with one prompt. You can see. You can see her. You are an experienced AI prompt writer in which you have writing prompts for AI tools like ChatGPT. So don't worry. This prompt have so many mistakes in grammatical or those things because we have tried AI not to provide the harmonization or inaccuracies in the output. The AI is trained by some natural language processing words like this. The AI is well known of our intent by just seeing in the prompt. You don't turn in there. You can see you are an experienced AI prompt writer in which you have writing prompts. You can go with in which you have ten years of experience in writing prompts for AI tools like ChatGPT. No, ask me subdivided questions. You can see we have using the personal prom pattern, that is we have assigned a role to AI. That is you are an experienced AI prompt writer. And we have given the background information that is context. You can see in which you have the ten years of experience, even you can take this one, ten years of Expertise like that, you can take this all. Expertise of expertise in writing prompts for AI tools like chargeb. We can go with for the different AI tools like cloud, all those things. This is the background information. That is ten years of expertise in writing proms for Atols like ChargBD you can go for the specific one. That is simple. Now we have just use the no Ask me subdivided questions related to digital marketing, that information is required for you. Now, what happens here? If you see here, we are just provided the context, background information to AI. Now, we have given the task. That is Ask me subdivided ion. This is called cognitive verifier prompt pattern that we earlier discussed. That is it will ask the subdivided sedative to our main topic. When I give the answers for that particular questions, it will generate the partialized output according to my requirements. That is simple. Now, what I told to AI for that digital marketing task. Okay, so I need a prompt for what purpose? For what purpose you are looking for prompt. So I've just given for the digital marketing. Even you can go for the content marketing, you can go for the different use cases as well if you're looking to get the prompt for ChatGPT or other AI models as well. Okay, I'm looking I'm looking for the prompt for the digital marketing. Okay. Note that information is required for you to generate best effect best effect to and engaging prompt. After I provide answers for you for your questions, then proceed to generate prompt. This is the simple prompt. You can see here. You can see all those things. That is simple. Let's give this one. Then you can see, got it. Since you want me to act as an experienced AI prompt writer for digital marketing task, so I will first ask you set of subdivided questions. Once you provide your answers, I'll create the best effective and engaging prompt for your use case. You can see all the questions, right? You can see what is your primary objective? Who are you trying to reach? Which platforms do you want to focus on? It is all the information that AI is looking from me when I provide the information. Okay, to AI for this particular questions, the AI will generate the personalized and effect to prompt to me in which I can use that particular prompt in other AI modules to done the task. That is simple. Instead of just giving the ChatGPT write the prompt for the digital marketing. So and so, you can go like this one. You can go just by interacting with this entire prom set, writing this one, and giving the answers, you will see the difference in the proms in generating the best output. That is simple. I give the answers, let's give the answers like brand awareness. I will go with this one, first question, first answer is a brand awareness. Second one, I will take target audience or location interest for 20 years, let's take years mail that is M and platform channel, I will focus on Instagram. These are all my preferences and requirements. I'm looking to get the best output according to my requirements from AI. That's why I'm giving the answers for this one. You can see. Product and service detail, what is your selling points or competitive advantage, product or service? No, you have different ways you can dump your product details to the RAG GPT, and you can paste your document from here plus button. Even you can give the link of your website in order to get all those things. I'll just some example question like selling digital watches like that. I take this one. For 20 years. Boy. Okay. This is a simple fourth question answer. Let's take the fifth one. One style. What style of mistreating do you prefer professional, casual emotional storytelling like that? You can take according to your needs. Okay, in the case of a purpose, you can take all those things like casual persuasion like that or emotional aspect. I will take emotional. 61 that is content type. Do you want Ad copy, social media post, email campaigns, block prompts or receive keywords or a mix. So as I said, if you or mix all the things, it will give some output, which is not clear or effective. When you go for the specific like I want ADFpy social media first, like step by step, you will get the best output for each task that you have already learned in the pred session. Okay, let's take with I want ARCpy on. That is my requirements. You can see what the budget and scale. Optional, but I will give the answers for better understanding for better output. Do you want the AI to consider budget friendly strategies or large scale campaigning ideas? Yes, I will take this yes. Now, let's see the output of this particular prompt. No, this is our prompt. You can see you want pranaweness for old males on Instagram promoting digital watches for boys in an emotional tone focusing on ad pop. It is a specialized and specific prompt that I can use in any AI module to get the best output from AI. That is how you can write the prompt. You can see this it is used and assign it is assigned a role to AI. That is, you are an expert digital marketing copywriter with ten years of experience in creating a high converting ad copy for Instagram. This is all of my requirements. It is going for the specific one. This is called prompt engineering. Prompt engineering is nothing but writing the prompts for specific application to get the best output from AI. This is how you can generate the prompt with the help of AI. This is how you will treat with AI, interact with AI with best interaction strategies and prompt writing strategies with the different prompt patterns we have earlier discussed in the previous sessions. Right. You can see it is the Chap also assigned a role that you are an expert digital marketing comparator, and it also added the background information that is ten years of experience in creating high converting ad copy for Instagram, which is a specific platform. Your task is to write multiple versions of Instagram ad Copy that will emotionally connect with 20-year-old boys and build strong brand awareness for digital watches. Requirements target audience is all my requirements that I have given here. Okay, requirements and deliverables three to five versions of Instagram add copy. Each should highlight emotional triggers like style identify and include at least one CTA suited for instagram. So this is the whole prompt that we can use in any AI to get the best output. That is simple. Okay? Even you can engender the proms like this very easily by just using the prompt patterns that we have eared discussed in the previous session. Right, you can see all the prompt patterns we have learned here. The AI also just included the simple personal prompt pattern, all those things here. You can see this is a whole prom. You can just copy here. You can go for the simple new chat. You can paste here and just go with this one and you can see the output which is amazing according to your requirements that is short form, quick, scroll friendly, not just a watch it is wide, it's your story. You can see the ad copy for Instagram. You can see this is the output from AI. I hope I understand this point very well. Not only that, you can convert this particular prompt, whole prompt into prompt template in which you can share it with your friends in order to they can edit with their preferences and requirements they can use in anywhere like that. So fund that we need to just write this one here, this prompt. I'll write this. No, please convert about prompt into prompt template in which user can edit their preferences. When I give this one, the chargi will automatically convert that particular prompt into the prompt template in which the user can edit their preferences, like you can see here. The target audience or insert audience details age, gender, location, interest. So when you share this particular whole prompt with offends or anywhere, they can edit their preferences like what are their target audience product services like that, and they can get the best output from AI. In the next session, we will talk about how we can generate the image prompt. Let's dive into that. 28. 5.4 Writing Advanced Image Prompts using Prompt Chaining Technique: Directly change edit this particular whole prompt here. I will just go with you are an experienced AI prompt writer. We will write instead of AI prompt writer, I will just take this one like image prompt writer. Image prompt writer, which have ten years of expertise in writing AI prompts or image proms like image proms for AI tools, we have taken for the text. Even you can go for the image generation tools like Image journey and Lean on to AI like that. I will take Leona to AI. Near to AI and other as well. I'll just take the EI image proms for EI tools like Lena AI. You can take the AI image generation tools. Like Leonard ROI and other as well. No Ask me sub divided quis related to digital marketing. So now I will just take it out and I will take for the simple another image generation topic like cartoon animals. Can see this is Colin animals. Now what task is to get the image for cartoon animals according to my requirement, no you can see. That information is required for you to generate best efect to engaging image prompt. After I provide answers for your questions, the proceed to generate image prompt. That is simple. That is let's send it. No, Charge will just ask me some questions regarding my prompts. You can see Animal type, what's the type of animal you are looking to get the image prompt from me like that one. When I give the answers for this particular question, it will generate the prompt the prompt can I use in every AIimage tool that prompt according to the prompt, the settings in the AI generation tool, I will get the best output. Okay. Instead of writing Bsel instead of thinking by yourself how to write the image prom, you can use this ChatGPT in order to get the best image prom and you can edit by yourself according to your requirements, and you will get the best output from AI Image generation tools as well. Let's do the answer quickly in order to get the best output here. Take the lion. Style what cartoon style do you prefer Disney like? I take Disney like? And pose action, running or dancing like that I'll just running clothing or accessor. Do you want the animal to wear cloth or accessories example, hat glass a superhero consume? I will just take hat. I will just showing here the example. In your case, you can try by different requirements, not only here, it is showing. You can take any of the preferences you want instead of this one as well. Background setting what environmental should be in jungle, city like that I just take this as a general. And mood or expression, let's take funny or cute adventures. I'll just take the adventures. And the seventh one is correct details do you want a specific color patterns or like that? I'll just take as the blue. And I'll see let's do these preferences to ChagVT According to that my requirements will give the two outputs. You can see sometimes the new version of Chat will ask which responses you need, which responses you prefer. Sometimes you can go with what you best look for you. You can see. Perfect. Thanks for details. This is a simple eight disney style cartoon lined with blue for wearing a hat running image. This is a image prompt. Okay, even you can see this image prompt. So I'll just go with this one. I prefer this response. Now you can see it is our image prompt. Do you want me to take the multiple versions, variations of the prompt, one more magical, one more funny, one more realistic Disney? You can go for that one. S IN versions I take S IN. When I give this one IN, it will give the three different refined versions variations of the prompt. You can see it is a variation one, variation two, variation three. Even you can try this whole all prompts in image generation tool to get the best output. You can choose any of these prompt. I will go with this one first one, and you can do all those things by yourself as well. Change Bra also have their own AImaeGeneration tool that is SRA like that. You can use all those things by yourself. You can just page this particular prompt into the AimageGeneration tool, you can see the best output from AI as well. Not only that, you can just told to AI. No, I am look I need the prompt template of particular this image prompt. It will just create the image prompt template. Instead of getting further specific output, it will create the image prompt template, you can share with your ImRI. They can edit with their preferences. They can get the best output from image generation tool. How you can do this. It's just tell we have already seen in the previous way, you can use this particular prompt now please convert. I'll just go with that one. Based here. Now, please convert. Above prompt into the prompt template in which user can edit the preferences. So when I give this one, it will just give the prompt template here. You can see Air cartoon style of Disney like Chat style, all those things. Okay. When you share this whole prompt with you or any colleagues or students or anybody, they can just edit their preferences. They will just put their preferences here and they will get the best output from AI. You can see the example filled in version as well. So I'm giving an assignment for you. Please try this whole prompt for the video generation. Okay, video generation app generation like that, and you will get the best to prompt for it, you can just write the prom, you will get output from AA. Just try by yourself for the video generation app generation like that and see how the ajet will change your prompting skill as well. I hope you understand these points very well, and let's up into another topic that is understanding different capabilities of LLMs. Let's tap into that. 29. 5.5.1 Understanding Different LLM's Pros & Cons: Let's see our next topic that is understanding different LLMs, pros and cons. So in this topic, we are going to learn how different LLMs like Cha Gibt Gemini, Deep Seek, or other ARLs works in various tasks, and we will see some tips to match proms to strength our each model to get the best output in and we will learn why understanding LLMs matters, some techniques, all those things, and we will jump into ChargePty and other AI models like Cloud purples.ai to understand their capabilities by writing the exact same prompt on different models to compare outputs and identify the best fit for our needs and task requirements. Okay? We will see all those things in this session. Okay? So why we need to understand the different LLMs prompts and cons. So as we know, every AI model is trended by different large amounts of data. In that case, so different companies are trend their AI model in different ways in different prom patterns like that. For example, agit is trend by different prom patterns or data patterns like that Gemini Cloud or other various AI models or trend by their own patterns. Okay? So in that case, we need to understand the LLMs, pros and cons in which particular task this model will shoot. Okay? So how we can do this one. So by using that particular AI model for our task, we can check their output. If the output is shoot for our requirements, we can use that particular AI model for our task. At the same time, we need to check that prompt in different AI models as well because instead of just using one AI model like for example, chagBT when we get the output, and there is a chance of getting the best output rather than chargebty in Gemini or Cloud, according to our task and prompting skill. Like that, we need to use the different AI models. We need to check first the different AI modulus for our task. When we see that particular AI model is solving my problem very effectively, then we need to choose that one particular AI model main objective of this understanding different LLMs, pros and cons, how we can check all those things, right? So you can see why understanding RLMs matter. So each language model has its strengths and knowing them allows you to tailor your promps effectively. According to your task, you will write the prompt. According to your prompting and task, you need to choose the particular specific AI model to done your task to complete your task because the different LLMs have their own strengths and capabilities and weakness as well. So sometimes as a prompt engineer, you do not mastering the AI tool here, but you are mastering the EI prompt writing skin. You need to write the prom according to the AI module and your requirements, then you will get the best output from AI models. I hope you understand these points, so you can see. So what are the tip that we can easily check the different AI models, which is the great one for M task, which uses format task in order to get the best effective and to solve the complex problems. There is only one tip you can follow that is test the same prom on different models to compare outputs and identify the best fit for your needs. That is simple. Just take one prompt right in the Char GPT. Okay, see the output and use that same prompt in other AI models like Gemini dot A purplctEI. Check their output. When you just compare all the outputs of different AI models, then you see which output is best suit for your needs. Okay. When you see that particular A model, generating the best output, which fits for your needs, then you will go with that particular AI model for your task. That is simple. Okay, even you can use different AI prom patterns that you earlier discussed, like ask me for pro prom pattern, personal prom pattern, they will work in any AI model as well. But the problem is, you need to understand the pros and cons of elements like ha jet, all those things in order to choose the A moral for your specific task like that. I hope you understand these points very well. And in the next session, we will jump into the practical way, and we will just implement this understanding elements, different proms and cons, all those things. We will just write the same prom and we will check all the different AI models, and we will try to understand their pros and cons according to our task and we will choose one model for our specific task, all those things we will see in the next session. Okay, Let's dive into that. 30. 5.5.2 Capabilities of ChatGPT, Gemini, Claude, Perplexity & Grok A with Use Case: Session, we are going to see the pros and cons of different LLMs by adding the same prompt and task in every model, we will check all the output of TI models and we will see which one is great. Okay. Let's dive into that. So I will just start with the high conversation with AGPty and we will go from step by step. You can see it is just given the reading how it is going. No, I will start with the insure prompt that is helpful assistant that were earlier discussed, right? So you can see this is a simple helpful assistant prompt that we can use for every interaction which starts for better output without any mistakes in the output. Okay, you can see you are a helpful assistant. You will do what I tell you have experience in detecting unusual words, all those things we have earlier discussed in the previous session. Okay. Now we will take one simple task. We will give the simple task to CharPT and we will analyze the output. After that we will go to other lens like Cloud, Google Gemini, perplsy.ai, and Grog and even other AI models to check whether which particular AI module is best suit for my particular task that I'm going to show infuse against. Let's give the task to AI. No I am in the hagibt I will just give the simple task that is you are an experienced AA expert in the field of deep learning, right? Now, your task is to explain about deep learning in simple words or you understand. So it is I am looking to learn the deep learning from hagibt. Okay, in the simple words, then I will give just this task to ChagbT and we will analyze the output whether it is good or not. Can see, I have understood you want me to explain deep learning in simple words. Here is a clear explanation. Deep learning is a way for computers to learn from lots of examples similar to how our brain learn. It is a simple definition and simple explanation about deep learning in simple words. You can see the simple output here. Now what I will do. I will just do with the same step by step prompting in the other A models like step one that is Cloud. I will start with the high like we have done in the hag BD. You can see it's a high, it is reading from the cloud. Now I will just give the insure prompt that is helpful assistant prompt. Let's give this one. As you can see, it is a response from AI. I understand that you would like me to helpful and accurate in my response, that is great. Now I will just give the task that we have earlier done in the ChachPi you can see. No, yes, I understand you would like you to explain deep learning in simple, easy to understand terms. You can see what is the deep learning and the deep part, how it works, real world uses deep learning is an essential pattern. You can see the explanation what is deep learning. Think of deep learning like teaching a computer to recognize patterns the same way child learns. And it is how it works. It has the best output when compared to the habit because the habit just given the definition and simple terms. You can see D planning is a way of computers to learn from lots of examples similar how our brands learn imagine want a computer recognize scares in pictures like that. You can see it is the output is in the simple and when compared to the cloud, given some more examples like real world use cases, voice assistant, photo apps, recognize pass like that. The two both outputs are well and we'll check other as well. Now we just copy this simple assistant helpful assistant prompt, I will go to just Google Jen and I will start with the high here also. You can see I will just page this helpful assistant prompt. I understand I will do my best to provide helpful, accurate and effective response. Now we will give the task that we have done in the Cloud. You are experience re expert, you can see what the deep learning and imagine to teach a computer to recognize a cat in a picture. Now, very right. It is also generated the image when compared to other EMRs like cloud and ChargePD. You see, the Gemini has simply generated the image all. You can see license by Google, no I understand my task is to explain deep learning in simple words. Deep learning is a part of machine learning which itself a part of artificial intelligence. Think of it is a way to teach computers to learn from data, much like human brrain learns from experience that DPs deep learning refers to the use of deep neural networks, which are networks in many layers. You can see traditional programming, machine learning, deep learning. Along with it, it has generated the image which is great to understand what is actually deep learning is. Very good when I compare to these two cloud and hagibt I will just read this Geminis answer because it has generated the image in order to we can easily understand what is actually deep learning is. I understand these points. Now, let's hop into perplexity.ai as well to understand whether it works or not in the same thing. Now I will just copy the same thing here. You can follow the same process in other AI models as well if you want to check that particular capabilities of AI models like rock and Mistral AI, Deep Seek, and Quenchat. There are a lot of AI models you can drive in the same process, trying the chargeability other AI models as well, to understand whether you are prompt, which is looking to get the output from AI will shoot for you or not. There. In the purples data, I will just start with the high like we have done in the earlier in the previous EI models. Now I will just repeat the same helpful assistant prompt purples EI. GCs, I understand my role is to provide accurate clear, all those things. Now we will give the simple tasks that we have given in the cloud hebnhag provide the same thing here. For the purpose of knowing whether this particular A model will help me in this task or not in better way when compared to other A models. You can see deep planning is a part of artificial intelligence that helps computer to learn from the large amounts of data using models that is called neural networks, which are inspired by how human brain works and what natural neural networks learning and training, what makes deep learning special. We can see the find complex patterns can automatically learn so when seeing this output, you can get in the conclusion like this perplexit dot EI will just give the answer from the online sources. By researching in online sources, you can see, you can directly go here, right? When you click here, it will just redirect to the website that has taken some information to show you in the perplexit.ai. So it will just take the sources you can see here when you click. It all the sources that the purplesyt AI has taken the information about deplanning and it will summarize to give the information. You can see this is all over. So what happens here when you are looking for the best output or real world use cases or accurate information because you can use the perpls.ai or Google Gemini because it has a searching platforms or a research platforms in which you can easily research the online sources by just giving the prompt here, it will research and it will give the real information instead of just giving answers like JGB or Cloud. Even though the Cloud and Hagib also give the accurate information, but by using the purplesit dot a, you can easily check the resources that the purplsi.ai has taken the information. You can easily check here by clicking the link here and you redirect to the website, you can see the all information which we can trust that particular content from AI. That is simple. That is how purple.ai will work. You can see the steps it has taken to and all those things and answer. Okay. Now when compared to this one, so while if you are looking to get the engaging content, with all those things, you can go with the Cha chptGulGemini, Cloud or Grok as well. But in the Perplexia dot EI, you can use for the research purpose for real content purpose. But in that case, you cannot expect the engaging terms, but it will just generate the real terms. Okay? So to make this particular answer into engaging way, you can write another prompt, follow up prom in order to get the engaging content. I hope to understand these points very well. Now, jump into Crop and we will see whether it will works in the chachT or other likenu. That. Okay. We will see in the Cronk as well, whether it will just give the output, which is best when compared to other AI mols or not. Let's dive into this one. I mean the croc I will start with the i like we have done in all the previous I mols. Now it is gain I'll just give helpful assistant prompt. Let's see this one. Yes, I understand I'm here to provide accurate helpful effect to response and sce the task, same task we have done in the other AM models previously. Now, see this. Let's see. Yes, I understand as an AA with expertise in deep learning, I will explain simple words. Deep learning is a type of artificial lgence that help computers make decisions by mimicking how the human being works. It uses something called neural networks, which are likely layers of connected nodes. You can see the example here, right? You can see that it is also simple terms that I can understand with this one. You can see input or layers, output. Deleting is a powerful because it can handle complex tasks like recognizing phase, understanding, speech or like that. You can see when compared to the chaps, tent. Imagine want to come recognize scars in picture. So it is a good output, because it has a concise and direct to the point in the hat in the cloud. We have seen some similar output when compared to hagibt, but it has given some examples as well, real body usage, how it works, like simple terms, it lots of examples that it make mistakes, learn from errors, all those things. Okay. Then compared to Google Gemin it has generated the image as well, which is which increase engagement in that particular content because with the image, we can understand rather than the writing reading that text, okay? It is a great thing about Google Japanese. It has also generated the good image also for better understanding when compared to ha gibt and Cloud. And when compared to the purples dot EI, the purples.ai is given the answer, but it is not in the engaging way, but it is in the real terms way. But taking the resources from different EI or different online resources by websites or research papers, it has given the answer. So in that case, I can trust this particular content because it is a real one which comes from the top sources of the deep learning. Okay, I have to understand the ones. Now when cope or graph, it is a fast, accurate and given responses that we have seen in this platform. Can see it is also giving the best output in the engaging way. You can see de planning is a type of EI that helps computers lean and make decisions. I see it is given the simple terms input layers output. Okay. So all outputs are best, but in the Google Gemini, it has ended the image so that I can read this Google Gemini. It is the best way for taking the best output with the image or video or other things you can learn easily with the Google Gemini as well for the engaging way and for creative writing, you can use a GPT and Cloud Okay, for the research purpose and for real information, you can use the perplicity.ai and for the fast and engaging and effect to answer, you can use grog for a task. I hope you understand these points. This is how you can rate var task according to the AI models capabilities. And you can go for the different task. I hope you understand these points. To learning this deplaning, I can use whether the Google chem or purples dot EI because the two are such platforms in which I can expend the correct information about deep planning. For the creative and engaging platforms, I can use grog gibt and cloud because they are AI modules that are trained by different online sources data. But in the real time, we can use the Google Gemini and purples dot EI for getting the real terms and the updated version because they are tried by real time online sources. And remember one thing, we cannot say every AI module will just generate 100% accurate information because it is an AI will do the mistakes, right? So for that. We cannot rely on that, so we need to use the fact check list prompt we have earlier done in the previous sessions, so we can use. In the next session, we will see the same process for other AI models like Quin Chat and Mital AI deeps. Okay, let's dive into that. 31. 5.5.3 Capabilities of Deepseek Ai, Qwen Chat, Mistral Ai & Copilot with Use Case: In the previous session, we have seen the different AI models fros and cons like Char GBT, Cloud gemi dot EI, and Purples dot EI and Cro AI. Now in this session, we are going to see another four different AI models fros and cons by adding the same prompt we have done in the CharBT. Like, we are going to discuss about Deep Seek Microsoft Copile G chat AI and mistrl AI. Let's dive into that. I will just start with the high conversation and I will just copy this helpful assistant prom that we have done earlier in the previous A models, we will just follow the same process in order to check whether these modules will help me in this particular task or not better than other A models. I will just start with the hi give you the deep now you can see it is just giving the printings. Now we will just page helpful assistant prompt intial prompt. You can see we will see the output, what should be done. You can say, I understand completely, I am here to be helpful and reliable assistant. I will do my best to follow your instructions and provide with your accurate pack to an appropriate information. Okay, I'm ready to assist you. What would you like me to do? I will just copy the same task we have in the previous session with the other AI models. I will just copy and page the same task, you can see you are an experienced EI expert in the field of deep learning. I'll just given the task. Of course, understood. Let's explain deep learning in simple words. Imagine you want to teach a very young child what a cat is. The old way, traditional programming, you will try to write a gender rule book for the computer. It has a point ears. It has whiskers, it has a tail. This fails quickly. What about a hairless cat? A cat with its tail hidden, a lion role. The rules become impossible complex. Complex. But the deep learning way, you show the child thousands of pictures and say, This is a cat or this is not a cat, the child's brain starts to figure out the patterns on its own. It might notice that future text for texture, eye shape, and no structure. No, this is the same very engagable and very simple words information when compared to the previous As we have done. It is generating the output in the very simple words that even the child can understand because it has taken the example as a care and a very deep explanation of the simple words, you can see it has a point ears like the deep learning way and you can see the deep in deep learning. What is it deep refers? You can see all the points here. After that, it has given some real world analogy. In the simple explanation of making orange juice, we can see key takeaways in simple terms. But it is a great one because I can understand this when compared to other AI models we have seen in the previous session. So it is a very engaging way or even the child can understand this particular information. Okay? So this is a great output when compared to the previous five different AI models we have seen in the previous session. Okay. Let's do the same exact process in the Microsoft co palette, and we will check whether it works or not. I'll just copy the same intell prompt here. I will just go the Microsoft co palette. I will start with the high Hi, S, good morning, seas remember my nine because I have already used this. I will just give this particular helpful assistant prompt. You can see absolutely SIF I'm here to help you with the precision, clarity and reliability. I will just give the same prompt our task that we have done in the Deep Si or other A models as well. I will just give this one. Let's see. Crystal care SF, let's break down deep learning the simplest to be possible Li I'm explaining it to someone with zero tech background. So what is the deep learning? Simple analogy. You can see deep learning is a type of AI that teaches computers to learn from examples, just like when to. Imagine you are teaching to child recognized animals. You can see every AI module is taking the example of cat or child teaching a child like that because so this particular data is already in the online sources. So based upon that online sources, the AI is trend. If you recognize there is a similar thing in each AI models output that is it is taking the example as animals and it is imagine your teaching to child in every output we have seen in other AI models as well. Okay. It is given a simple analogy, you show them lots of pictures of cats and dogs. So you can see it uses lots of data. Was inside the deep learning system is, real life examples, would like me to explain how neural networks work mix or dive deep learning user in your favorite apps. Okay, it has simple given the example and definition and given the simple analogy, you can see showing them lots of pictures of cases. Over time, they learn to tell the difference even with the pictures they have never seen before. It uses lots of data like this. You can see the output here. You can try yourself with your own task and you can check all the AI models output, you can compare, you can analyze output. Then you can select only one AI model, which you I think this output will help me in a lot when compared to other AI models. Just go with that AI model in that particular task. As a prompt engineer, we are not here to master specific tool, specific AI model, but we are here to master the art of writing the prompt. You need to focus on the prompt. You need to check every AI models output, then you need to select only one AI model for reword task according to our requirements and complexity of your task. That is simple. That is the purpose of this analysing the different elements, proms and cons. I hope understand this one. It is also a good output when compared to the deep sick as well. Let's give the same process in the quin chat. I will just copy the same prompt, the quin chat. Let's start with the hi. It is again the same different output. Let's see the first helpful assistant prompt pattern here. Understand completely, I will follow your instructions precisely. I will detect and flag initial words, all those things. Let's give the same task we have done in the previous EI model, Microsoft Cp Okay. Yes, I understand completely. As an experienced expert in deep learning, I will explain to it in simple, clear accurate words. You can imagine you're teaching a child to recognize case. It is the same example that every I model has taken to explain the deep learning in the simple words to me. Let's take simple example that is how does it work? It has layers. Each layer learn something simple first. Real life examples. Why is it special in one sentence, let me know. No, I think that Deep Sik is generally the best output band compared to the enchat and Microsoft cop because Deep Sik has given the very engaging and very simple explanation that even child can understand. We have seen in the previous one, right? Okay. Now we have another AI model just for the last A model that is star AI. Let's start with Hi. There are a lot of AI models market right now you can dry by yourself and you can go with that. I will just give you the idea how you can compare all those things, how you can come with the conclusion with according to task, which type of AMR should be used according to requirements and creativity. I understand these points. Let's give the same prompt that is helpful assistant in the here, this one. Understood shakes. I'm here to assist you with accuracy clarity, that's good. I'll just take this one. Absent the shake, what is deep learning? Deep learning is a type of AI that teaches computers to learn from examples like just like Minstre is inspired by how our brain works using layers of artificial neurons. How does it work? Why it is special, real world examples, voice assistance, limitations. So when compared, this is a straightforward answer without any engaging or simple terms explanation, you can see here mista doesn't give it. There is no engaging in that. That means simple words. Okay? It has given the output directly with some good information. But Quencha also given the best output, Microsoft Copart has also given the simple redons but in the deep if we have seen, it is also simple explain that deep learning the simple words like even child can always understand. Get the old traditional programming with lots of examples. It feels quickly N see all those things. Then you can easily understand this type of information or skill by using the deeps because it has given some engaging words even the child can understand. By understanding these examples, you can understand the main topic in Depth. Okay, I hope you understand these points. When you use your deep or AI models for different tasks, you will get the idea of how these AI Ms are working, all those things, okay? Not on that, every model have their search button think like that, you can use this depth for better answer by reasoning all those things. For every model, it is there. You can use the search Depth or you can even can add some files, all those things. You can go for the Microsoft copilot. You can go. You can turn on the talk to copilot all so you can upload the file. Quin chat also have their search button, button or even you can add the upload document as well. For every A model, there are the tools you can research, but in all those things you can upload, you can learn, you can divert all those things by User and you can analyze the output, and you can come with the conclusion which type of AI model is best to complete your task according to your requirements and the output of AI models. Okay, that's This is how you can come with all those things, okay? In this si out of AI models, I will use the deep Okay, for learning this leap learning because it has given the best output when compared to these IMs. In that review session, we have seen other AI models as well. But in learning the engaging content, when looking for the engaging content or looking to learn the complex topics in simple words, I will use the Tips because it given the best output in my case. In your case, you can take other Als as well like hareptic loud as well. Okay. So up to now, we have seen the nine different AI models. We have compared all those things for the simple, similar and same task that is deep learning explanation. For the conclusion, I will use the deep seek for learning the deep learning because it is given the best output. Okay? For the real use cases for the accurate information, I will use a purplicity.ai for researching purpose, and I will use Google Jey as well because it is Aussie searching platform like that, and for the engaging content, and for ato writing, I will use a ha GBT Cloud and grow Deep Seek like that. And for other tasks, I can use a Quin Mistle A. But I'm not very happy with this queen chat and Mistle for this output. There are only the simple HGB Cloud, purpose dot A Google Gemini and Dips these are the best AI models that we can get the best output from AI. It is all up to your prompting skill, how you will talk with AI, how you will interact with EI and how you will use a prompt patterns in order to get the specific information from AI. I hope understand these points. So please try by yourself for the different tasks, take two to three tasks, different different tasks and do the same task in all AI models check whether which type of AI model will help you in that particular task, go with that and you will learn the art of writing the prompts for different AI models according to the capabilities of AI model and according to requirements and prompting skill. I hope I understand these points very well. In the next session, we will dive. Which type of AI tool is the best to generate the EI prompts for our requirements. Okay. Let's dive into that. 32. 5.5.4 How to Use ChatGPT, Claude, Gemini, Perplexity & Grok Ai to Write Prompts: Previous session, we have seen nine different AI models, how we can compare all the outputs and which type of tool is best shot for our requirements and task by comparing the outputs of simple and same tasks that we have done in the previous session, right? So in this session, we will see the exact same process for writing the effective EI prompts for our requirement. And we will check which EI model is best at writing the prom for our requirements. Okay, let's dive into that. So I will use first ha GPT. We have seen already the prompt writing and how we can generate the prompt by using charBT in different ways. We have already done in the previous session, but now for understanding the different LLMs, pros and cons and capabilities of AI models and writing the effective prompts for our requirements, so we will check all those things in this session. Okay. Let's start with the Cha GPT. Now instead of going with the high, I will just take directly to the helpful assistant prompt we have done in the previous session that we'll start with this one. And after that, I will just write simple prompt to generate the prompt. You can see here this is a prom. That is you are an experience things. You are an experienced prompt writer in the field of psychology of humans in marketing. Now your task is to generates two to three different versions of prompts for AI. So I'm just giving you are an experienced prompt writer in the field of psychology of humans. I'm using the personal prom pattern with adding the context background information to AI. Okay, I'm giving a task that is you need to generate the best two to three different versions of prompts for AI. Let's give this one. You can see, got it. Since you want to prompts in the psychology of women's in marketing, I'll be generate two to three strong and effective prompt versions. You can see three different prompt versions that I can use for my requirements to understand the behavior of a consumer or customer. To sell my product or to do all those things. You can see you are a marketing psychologist with expertise in consumer behavior and decision making, analyze a given product or services, you can see. You can see the second version that is act as a human psychology. Prompt version, you are a neuro marketing strategy skilled. If you see here, for every prom, it has assigned a role that Charge also know about the personal prom pattern, right? So that's why we are using this particular personal prompt pattern in which the EI can also write the prompt with the prompt patterns we have earlier discourse. You can see these three different prompt versions of my task. Task is generated two to three. It has simply generated three different versions of prompt. You can see in the field of psychology of marketing. Now I can use these three prompts in any Amdle to check the output. You know already about how we can generate output from different A models. That is simple. Now what we will do I will just take another example. Here you are an experienced prompt writer in the field of psychology human in the marketing. Now the task is to generate due to the different versions of prompt for EI. Ask me subdivided quis relative to the mendas that you're required to generate prompt. So we have already learned this ask me subdivided is all the verified prompt pattern in previous sessions very well, you know, and we'll just and we'll just compare the same process in other AMOLs as well, whether these words in the other AMLs or not like chargeabilty too. When I give the answers, it will just give the prompt two to three different versions of prompt. Let's give the answers quickly. Purpose of the prompt is add Second question target audience, or you can take the males. Let's take the third. Product or services, Luxury watch take this one luxury watch. Question is psychology angle should the prop focus on emotional triggers here, desire, trust belonging desire. Or to form it, I just need a R code. Let's give you the answers for these questions. You can see it is also simple details. These all the details we have given. No, it is three different prompt versions. Okay, you can see here. I can use these prompts to generate the ad copy according to my personalized details. Okay? When the previous session we have seen has given the three different versions of prompts, which are not related to my personalized requirements. But when I write this ask me related questions, it has simple ask the questions. When I give the answers for this question, it has generated three different prom sons up unto my requirements. So we have already done in the previous session, we have learned all those things. Now we will jump into other AI models like Cloud, Google Jem purpose, dot EI, and grow AI in this session. Whether the same process of this qualitative welfare prom pattern, all the generation prompts in the CharPT or AI models works or not, like others. Okay, I will just for the same process in other Air models as well. I'll just go with this prompt here. We'll start with this fast me and we'll check the output of each one. You get the best answer. Now just copy this first prompt here, that is us paste it here. You expenser you need to generate 23 different versions of prom. Now you can see version one, consumer behavior analysis Prom. You can see here it is a very long prompt. You can see our consumer psychoge expert and economic Analyzer author or insert product and services. You can see it is a Version two, version three. You can see why it's a great ring, right? So when compared to the hagiPt and the cloud, the Cloud has given in the type of prong template. You can see product and services, we need to just edit this product and services with our requirements directly. Instead of giving the just a prom, it has given the personalize. You can see here. These proms are very well crafted and it is very detailed when compared to the hagiPt. You can see these are the proms which are in the less format, you can see, right? But in the cloud, it has given them more in depth of the exact prom template, you can see. You can check all those outputs by yourself for better understanding and better analyzation. We have get the best and crafted in detail background included prompt in the cloud when competive algebra. Let's do this our other prompt here cloud. Let's see the output. Now it will ask me some questions. You can see I need to give the answers for these quotiens When I give these is, when I give the answers for this quotient, then it will generate the best output. Now, if you see here, it asking so many questions you can see your purpose and application, and I will just go with anything like, and I will just go with the same inputs here. Just copy this one. I'll just paste here. Let's see the output. Perfect. Based on your applications, add copy for luxury watches targeting miss focus, you can see it is a prompt with my requirements. You can see version one, version two, this is a version two, and this is the version three, right? Now, if you analyze the outputs of cloud and hagibt you can see these prompts are very small and it is also very good prompt. But in the cloud, it is given the very detailed and very effective prompt templates and prompt when compared to hagibty. You can check here. So it is best when compared to Cha chibi and Cloud, the cloud is best at writing the proms very well when compared to the ha chibi. I hope understand these points very well. Now let's see other Google Jey and other A models as well for better analyzation of A models. I'll just go with Google Gemini. I'll just pace this one. I'll start with the intial prompt and I will just copy the same first prompt. I'll just pase this one. Al gender 223. Now, you can see here. The Google Jemi also given the best prompt, create a marketing strategy, but it lacks one thing. I have just asked you to give the two to three different versions. Nowadays giving only two versions. Not only that, these two prompts doesn't include the ask me for Improve prompt pattern or background information very much when compared to the cloud. You can see it is a simple prompt creative marketing strategy for limited edition, sneaker release or draft marketing Bob software company, which is not related to my all those requirements. Now I will just give the same process we have done in the two BI models that is this one. Et's give this one. No it will ask some questions to me. Let's see if the ox or not. This is an excellent approach. These are subdivided questions. When I give answers for this particular question, it will give the answer. I will just copy the same requirements that have done in the previous EMRs. I'll copy this one, I will just paste in this chat of Pool Gemini. Based on your answers, it is a prompt version two. Only two versions or it is generative. So when I write another version will also generate. But in the one term, just have given just write a two to three. No, it is only taking the two versions. When I just study it and just to keep at the three, it will generate the three versions. That is simple. This is how the different AMLs work according to our requirements. That's why you need to check each and every AMl for your task. Then you can go with that particular AML to complete that particular task. Now you can see it as best prompt when compiled. But it is a good one, which is not effective as much when compared to the cloud. I understand these points very. That is simple. You can see you can try by yourself for better and depth analysis. Now I will just use the same thing in the Patric EI. Well let's dive to that. I'll just copy the first prom that is this one your helpful assistant. Copy the same first prom here. You are experiencing prompt right there. I can see it is a Version one version two, version three. Imagine you are taking the art as a personal prom pattern very well, personal prom pattern here, direct prompt. That is fine. Now, I will just give this third prom that is your experience prompt writer. Let's go with that. You guys to generate the best two to three A proms in the psychology of mens in marketing, I will need more clarity from you. So when I give the answers for these particular portions, then it will give that two to three different proms that we have already seen in the previous sessions, right? Now, let's give the same information that we have given in the previous AI model that is Gul Jamin. Let's page this one here. You can see it is a simple prompt version prompt two, prom three. It is simple prompt. You can see her. Three to two lines prompt. You can expect from plectEI. Okay. Let's jump into that same thing that is Croc. And we'll start with a simple first two prompt that is your helpful assistant. Start with this one. Let's go. And we'll just go with this second prompt that is your expense prompter. Okay. I understand. Let's give this 1 second form. It is analyzing, understand the request. Your task is to generate two to three different prop options for you in psychology. Marketing is taking time to give the answers. It is thinking in deep. Let's see. You can see it is three different prompts. You are in a marketing expert specializing human psychology with all those things. It is a prom template data given that format. You can see it is prompt template. You can see this one target audience. We need to edit this one. The three proms are well defined. You can see these three proms are using the prom patterns that is personal prom pattern act as and you are an AA consultant in the neuromarketing. And it also generated with prom template in which we can edit our preferences and we will give the answers best wanted. Okay. Let's give another prompt with our prom patterns that is asking me subdivided questions given the grok. Let's check this one. So I need the exactly we have done in the previous sessions, we need to give the answers. I will just copy the same things here. We ask the seven up to seven questions that I need to give the answer. I'll just give the same questions to Grog. Let's see the examples or prompt here. Based on your input, I will create the prompt. Now, you can see, these are the three different proms versions according to my requirements. You are a marketing specialist, act as a neuromarketing specialist, you are a psychologist with expertise in consumer behavior, right? You can see these are the three different prompt patterns. Okay. So up to these five different AI models like Cha JBT Cloud, Google Gemini, purples.ai, and Grob so the conclusion of the I models is when in the chargBT we have seen the best I prompts as well, according to the prompt patterns that is personal prom pattern, all those things it has added very well. In the cloud, we have seen even more advanced prompts when compared to the ChagPT, Google Jem purples dot, and grok as well because it is a well crafted and more detailed prompt we can use in the II Mule to get the best output. You can see the cloud has dented the best prompts when compared to the harbt and Google Jimi and purples dot. So in the writing of these proms, I rate the cloud, AIM is the best at writing the prompts for our requirements when computed the other EMOs like ha Gib, Gemini, purples, and Crop. Okay. In the next place, I'll rate the Croce to writing the best prompts. And the next in the Google Gemini, you can see the proms are well crafted and in detail. And third one is hagibt and fourth purples.ai, like that. Okay, I hope I understand these points very well. Like that you can see. So as we have earlier discussed in the previous a session by taking the simple deep learning task. Okay? In that case, we have seen for the creative writing and effective writing. In engaging words, we have chosen the three different Emols like JGBTCloud and grow. Okay? In that case also, it is the prompt writing AI tools or in the engaging words with effective prompt pattern techniques or hagibt cloud and group because these three platforms are using the personal prom pattern as well, right? But in the gemini you can see here. In the Gemini, there are no prom patterns that has used an act as a personal prom pattern like that. It is just writing the directive prom instead of taking the personal prom pattern like that. In this purpose also, we have got some exact answers like we compelling R copy, create, pursue, generate a series. It is not taking the N or adding the personal prom pattern techniques like that. So by conclusion, we get the same thing here also, as we have done in the illustration. So it is Google hemney and purpose.ai like searching platforms. But in the writing the effect to manure or engaging words or using the prom patterns, we can use the Chagp cloud and grow for writing the engaging and effective prompts for our use cases and requirements. I understand these points. So in the comparing these A models like chargebt, Cloud and grow, so I will rate the Cloud is the best AI module to write the best prompt for our requirements when compared to Chabty and group. Help understand these points well well. So you can try the same process for the image prompt as well, try yourself, and just you will learn all those things by yourself by writing, analyzing the different prompts as well. Yeah, I hope understand these points. So in the next session, we will also check the other AI modules like deep sequence chat and mistallEI and Microsoft copi to analyze the prompts, which are the best at writing the prompts when compared to Cloud, let's dive into that. 33. 5.5.5 How to Use Deepseek, Qwen Chat, Copilot & Mistral Ai to Write Advanced Prompts: The previous session, we have analyzed the different AI models output for writing the AI proms. In that we have chosen the cloud for writing the best AI proms for our requirements. Okay. In this session, we will also analyze the output of different A models like Deep C Quenchat, Microsoft Copalt and Mistral AI with the Cloud. Okay. And we will see which type of AI models are best to writing for writing the A proms according to our requirement. So we'll compare all the four different A models with Cloud, which is the best one for it. I hope you understand these ones. Okay, let's start with the simple task in the Dip sk. I'll just replace this helpful assistant prompt and you will see the output. Yes, I understood completely. Let's go with our prompt. That is second prom that is your expense prompt writer like that. Replace the exact thing here also. Of course, expend prompt writer specializing the cycle of human behavior in the marketing. Here are the three different type of prompts. You can see Version one, the deep dive principle baser prompt. The prompt is designed for comprehensive analysis. The prompt is this one. Best for marketers Version two, let's wait for a few seconds and the prompt is That's great. Version three, computed with emotional framing prompt. I No, if you see here, the Deep Sik is also the best tool for writing the prompts for requirements. You can see it is also using the personal prom pattern that you are an expert marketing psychologist. W task is to analyze the marketing potential. It comes with the prompt template. You can see the background information as well, and the version to also grade you can see it also uses the personal prom pattern I the thar it is using the personal prom pattern. It is following that prom pattern, it is giving the identification or background information as well in the three different prompts. You can see these prompts are also similar to the cloud you can see here. You can see this cloud also generate a similar type of prompts in the deep Sk or in the cloud like that. Okay? Now, we will exactly follow the other prompt like using the Ask me for input prom pattern like that. We'll see check the deep also generate the best output or. It will ask some questions to me, you can see. Of course, as an expert in this field, I will understand the most powerful proms or tell for all those things. So I need to give the answers for this particular thing. So I will just take the same details we have done in the earlier. Kens it is asking quotients to me. I need to give the answer. Example, high this ws. You're also giving the extra information like instructions, all those things. This is a great one when compared to Cloud as well. Let's give these answers. Of course, based on your details you provided AbcrafT distinct. Ens it is also generating the three different prompts, Kens. You are an expert cooperator, write three versions of short, powerful add copy, incorporate all those things prompt to the legacian narrative angle. It is giving the background formation as well. Prompt three exclusive access and scarcity. No, if you see here, this is also well crafted prompt according to my requirements, you are an expert cooperator for the height and actually watch brat. No doubt Audenc is a successful man h 35 to 55. That is best output. You can see it is a whole prompt. It is similarly working like cloud. You can see the cloud. You can compute this with the cloud as well, right? The Cloud is also given the best output and deepsk also given the best prompt with the background information in detail information and using the personal prom pattern, all those things. Okay. So the Deep Sik and Cloud are best models to write the best effective prompts for my requirements as well. Okay, let's see three models, whether it will generate the best output or not. Let's jump it to quin chat AI and it will follow the same process that we have done in the Deep Sik or AM models as well. Just copy the same thing here.'s place this here. Let's give this one. Okay. Let's give the first prompt. That is your experience prompt title. Best this one here. We'll see this one. You can see the prompt version one is act as a consumer psycho expert in the triggers. You can see this also comes with the same prom template. The prompt version two is also same prom template, and prom three version is also called prompt template. You can easily editable. So these three prompts are very crafted, but in the simple tons or instructions when compared to other things. Okay? And let's do other things. Let's third prompt that is with ask me input prompt patterns like that. PSE prompt here. Let's do this. Thank you. I'm two to three versions in the field of cycle humans in the marketing. But to ensure maximum I need to give these answers. Let's copy the exact answers that we have given in the previous hat. Let's e the answers. It is asking the questions. I will just the answers here directly. Thank you. Input is clear concise and let's queue, based on your answers, it is answers. Now you can see the propson one is act as a luxury brand cycalespecial in the male identity. But if you see here, the proms are very small, and very well crafted, but it is a small. Even it is following the personal prom patterns like act as a U or a behavior as a consumer psychologist like that going in the specific matter, that is well, right? But you can see these prompts are very small and very less effective when compared to deep sick and Cloud. Okay? That is very great. Okay? Now, let's do the same things in the mixofcpalt. We'll just copy this first intell prom. This one. To save time, I will just go in the midsta paste the same exact initial prompt here or so. Let's come into the Microsoft Copalt which is the first one, and let's give the first task to misft co palette I will just give this one. Let's give this one. When as a prom writer in the field of psychology, you need to generate 223 pest. You can see these are three prompts which are in the lest sentences. Okay, these are the proms. You can see prompt one, prompt two prom three. These are the very low and there is no Bgunfmation. There is no following of prom patterns very well. There is no asking like you are an experienced or prompter like that when compared to the other A models. Okay, let's see exactly the mistal A as well. Absolutely crafting effective prompts for A, you can see it is using the prompt patterns, following the prom pattern that is as a behavioral scientist as you are a master operator, understanding of psychology triggers. You can see these prom patterns are also very well compared to when compared to the Google Microsoft copilot. Okay. Because it is also taking it is returning with the prompt update like that. Okay, let's give our third prompt, that is some prom patterns. Let's give this here. The That onions contest, I need to give the answers for this. Let's follow in the MitalE as well. He's asking some questions. It is very fast when compared to Microsoft Copale. I will give the answers that we have earlier done in the quinchet same answers, right? Same problems, same answers, same process. So we are analysing the different AA models for our requirements and task. I'll just go in the Mistral same things. Let's analyze first Microsoft cobalt. Now you can see when I give the answers and requirements, you can see it is simple three lines, I say, two of lines prompt, which is not very clear or not using any prompt pattern or any background information. You can see these are the three simple proms. There is no effect in writing prompts. In the Mistral, you can see. So according to my requirements, I have given the mistrl. You can see the prompt is very defined and very well, you can see here. A as a master copywriter and example, add a copy. It is also generated the visual all those things. So it is only generated one prompt according to my requirements, but I need the two to three pros that I have done in the things. So by this, we can just conclude that. So we have in the previous session, we have seen the five different AM models like hachtCloud all those things. In the things, we have analyzed the Google Gemini, Google Gemini, which is not followed the ask or URN personal prom patterns. Like that Microsoft copilot has also a search like Bing search, you understand. So it is As such platform. No, in the Google as such platform and Bing As such platform. In these two chat platforms, there are no the EI is not good at writing the proms because it is not fawning any prom patterns like personal prom pattern like that. It is only for research purpose or getting the real information from online. That is simple, right? In this case, so instead of or out of searching platforms like Google purples.ai or Microsoft Copint, so we can use the Cloud Deepsik Quenchat mistll AI, Cloud ChargePT and grog for writing the effective prompts for our requirements because we have done all the things, we have checked the proms of each Air model in the previous session and now as well. You can see that you can compare all the outputs, right? So in this particular platforms out of Microsoft copilot purples.ai and Google Gemini, okay? Out of that out of that three models, we have six models like Cha GPT, Cloud, Deep SI, Mistral AI, Groce, and Quin Chat. So in these six platforms, I will choose only two AI models that is Cloud and deeps for writing the AI proms. That is simple. So if I need only one AI tool, I will just compare the two outputs of two AI models, that is Cloud and deeps. I understand these points. So what we have done simply, we have just taken one task for all AI modules, we have unless the output. So in that particular nine modulus, I come with two AI models that are cloud and deep s for writing the best effective proms. So what I will do, instead of taking the two AI models, I will only choose only one AI for writing the best prom. For that, what I will do, I will just take the two proms, which is generated by Cloud and deep si. I will compare the output of that two generated prompts from Cloud and Deep Si I will compare the two outputs of Cloud and Deep Si. And according to the output, I will choose only one AI model for writing the ff two prompt, whether it is a Cloud or deep sik like that. This is how you can just come with the testing, analyzing the output with the similar tasks or other IMRs. You can rate according to requirements and analyzing the output according to our requirements. I hope I understand these points. So it is very easy. You can try to you can do for the same image prompts as well. I'm giving an assignment for you. So generate the image prompt for special pun for carton animal or like that. Take only one image prompt example, test it out on other AI models like ha chip, all other nine EI mods or other AI models out of nine AI models we have discussed in previous session and now. Check all the AI models and take only one test, analyze, and choose best AI model according to your requirements. You can use for lifetime that you can use generate different proms by using different models, according to requirements, all those things. This is how you can compare and choose only one AI module according to requirements and the output of that particular AI model. I hope you understand this topic very well. What are the benefits of using different evidence to write effective prompts? We have already discussed each and everything. Now, what are the benefits? So I can improve you on accuracy and precision in writing the prompts for different AI modulus as well. You can adaptable to use cases. According to our requirements, we can choose the AI model for our specific task by analyzing each output of EI models and we will check and we will come with that conclude, this is the particular AI model best shoot for mya task to complete it. So you can go with that particular AI model in which you can get the adaptability to use cases. Okay, I hope to understand these points very well, and third point is iterative optimization. So by analyzing the each and every output of writing the prompts, you can optimize your prompt very easy by combining different AI models. We have seen all those things in the previous sessions, all those things. And the major benefit of this using LLMs to write effective prompt is non experts can leverage LLMs to create high quality prompts without deep knowledge of AI or NLP techniques. Okay? So you do not require to learn all the prompt engineering from basic to advance all the things, but you can use LLMs to write the proms very effectively without learning all the things. I'm not talking about learning the prom patterns, but these prom patterns are very important to generate the best output, but you're not required to master each specific topic of writing the prompt, all those things, but you need to have some idea how this EI works. Okay? By using these different adelems to write afecto prompts, the non experts can leverage this capability to generate the best EI proms even though better than human beings. I understand these points. Okay. And another benefit is testing and evaluation. Okay? By analyzing the different AI models, effective proms, you will evaluate which type of particular EI prompt is well crafted when compared to the EI different models or well crafted when compared to different EI models. We have seen all these things in the earlier writing for the specific AI models like Charge Cloud, deep Ck, and other I models as well. So in that we have come with the Cloud or deepZk, ok? You can do by yourself, try by yourself. In the next session, we will jump into another prompt engineering tools which can save you a lot of time in writing the prompts as well. Okay. And we also explore open AI playground for a better understanding of AI, how it works according to our prompting and some characteristics of that. Okay. Let's dive into that. 34. 5.6.1 Basics of OpenAI Playground Prompt Engineering Parameters: Oh, in the previous session, we have learned how to use different LLMs to write affective prompts. Okay? Now in this session, we are going to discuss some prompt engineering tools like open A playground and other tools as well to generate the best prompt for our requirements and use cases. Okay? Now, in this session, we are going to see what is an open A playground, so what it is, and what is the benefit of using open A playground for our use cases to build applications or assistance API like that. Okay? Now, we have some parameters we need to understand, we need to have the knowledge about the temperature and all other parameters because within these parameters, we can control the assistant output or any output AI Ms output with our requirements according to our task and preferences. Okay, I hope, understand these ones. So we have different parameters like temperature, maximum tokens, so we will talk all over these things, top value and put. Now in the latest version of open A playground, so we have two parameters we can do the practical in the open A playground that is temperature and top. So rest of that, we have frequency penalty, presence penalty, and stop sequence. So these three parameters are removed from the open A playground right now. So we'll just talk about all these parameters and practically we will just receive temperature maximum topens and top V. Okay. Let's start with the temperature. So what are the temperature? You can see just remember the values and the what it will done, what it is done if the range of temperature is low and high. You can see low values example 0.256 up to like that, the values of temperature if low, the model generates more predictable and focused response making it ideal for tasks like math problems or fact based requires. This is best for getting the fact based or problem solving output, for example, math problems or. Okay. So with these slow values of temperature, we can get the output more predictable and focus response instead of getting the stuff. That is simple. It will best for problem solving prompt task or fact based queries. I have to understand this one. But the high values of temperature, if you put rather than more than one, 0.8 like that, the model becomes more creative and diverse in its output, but may produce less courent answers. It is quite opposite to the low values. When if you put this one, the output is have some stuff and all the things like creative writing, which may produce some less coherent answers, which is not related to our topic like that. The output is good, but it may produce less coherent answers as well. I understand these points. Okay. Now let's go for the maximum tokens, and what is the maximum tokens Tokens are chunks of text that the moral processes including words, punctuation and spaces. Okay. The maximum token setting limits the total length of response. The tokens are nothing but is the words, the type of characters that we have in the output. You can see the one token is equal to the four characters that is most important. For example, you can take that THAT that that word is equal to one token because it has four characters that is simple. What is the benefit of learning these tokens? We just learn input and output tokens in which we'll get all those things, knowledge about it. We'll see what is the tokenizer we have the great open A tool that is tokenizer in which we can easily analyze how much tokens are there in my output. Let's see this one. The maximum limit depends on the model. GPT four can process more tokens than GPT three p five. It all depends upon our type of model we are using to generate the output. We'll talk all those things in a few minutes. Let's come to top V. What is the top V? Nucleus sampling. It is also known as cool nucleus sampling. It can control the diversity of the response by considering the cumulative probability of token options. It comes from the zero to one. When set to one, the M will consider all possible word options. When the range comes to lower value, the mol focuses on the top few most likely words, reducing the randomness. That is simple. When you put the top value in the higher the model consider all possible word options, all possible word options to generate the output. If you put the value to the lower 0.30 0.5 like there, the marvel focuses on the top few most likely words, reading the randoms. Instead of putting all the words into output, it will just take some most important and few most words, it will present the output in that selectable words only. Instead of just putting all the possible words options. I understand these points very well, the fourth one is frequenty What is a frequenty penalty? Right now, we don't the parameters like frequenty penalty, presence penalty, and stop sequence in the open A playground. But you can find this type of parameters in the Google Gemini like that. The playground. I understand you can find these options in that. Now in this session, we are going to see only open A playground. Now, what is the frequency penalty? The frequency penalty discourage the model from repeating the same words too often within the response. What happens here by by changing the values range of the frequency penalty, you can control the model not to repeat the same words in the response. That is simple. If the range is two, there is no repetitive word in the output. If you put to one, there are some repeated words in the output. That is simple. That is frequency penalty. You can see higher values reduce word repetition. That is simple. Now comes to the presence penalty. What is that present finality? The present penalty encourages the model to introduce new concepts that haven't been mentioned in the text yet. What happens here it is it's a presence finalte which encourage the model to introduce new concepts, new words, new text or new statement that haven't been mentioned in the text yet like that. It will think by itself and it will generate the new statement, which is not mentioned in the textt or trained by AI or trained by data like that. Higher values make the output more diverse by reducing the repetitive themes as well. With that a simple presence finalities. What is stop sequence? Stop sequence tell the model when to stop generating out can define one or more phrases that signal the molt and its response. Example, setting a stop sequence as triple as you can see the and output when the sequence is encontrad. Encontt. What does this mean? So if you are looking to stop the output at particular word or number or any symbols, you can do with the stop sequence. Okay, to show the practical way, right now, the open a playground is just to remove these parameters, so we cannot say we cannot see all those things, but you need to have some knowledge about these parameters as well. Okay. There is some input field. We can get this hash or any word when that when the model will see the word in the output, it will stop generating the response at that word appear that we have just written in that box. That is simple. You can see the example setting a stop sequence as triples the output when the sequence is encontrol. Now let's see practical how this temperature, top value, and maximum topens will work. 35. 5.6.2 Overview of Temperature Parameter: No, I'm in the OpenAI Playground. You can just get the account for free. After that, you can see, you will get this simple interface. You can see the dashboard doc, EPA preferences, your profile. In the left side, you can see chat, you can create the chat, audio, images, assistance. Okay, how much usage, how much EBAs you have. Just use. You can see what is the amount you can spend, how much amount you spend. What are the IPA keys you are created? You can see all the IPA keys logs, you can see storage, batches, evaluations, and fine tuning. So the fine tuning, we will talk all this about some basic knowledge of fine tuning, all those things in upcoming model. Now, let's focus on this playground. Come to here, assistance. So for more information, you can check out the talks. You can learn more about what is the models, what type of models we have GBPi. Mini What is the pricing of these particular models. You can see GPT for the latest one, the input tokens are 1.25, output of $10 per 1 million tokens. So we have just nine words a token. One token is equal to the four characters. For example, TH 80 is equal to the one word. I is equal to one token. Per ten lats token, we call as one tokens. So when you model will generate the 10 million tokens, it will cost you $10 for the GPT five model. I hope you understand this point. Okay? This is how you can see all the models different models pricing here. According to that, you can choose the model to create applications or assistance like that. Now, come to embeddings, you can see the embeddings, moderation, what are the legal legacy models that moderation. You can check all those things from here directly, not only that you can see the badge vile pricing as well. IGP five, you can take the batch different models. With that, you will get all those things. For more information, you can check on YouTube or their dogs. They have the community, you can join, you can check the libraries. You can go here. There is a one section that is prompting. You can see here. You will get in the run and scale tab, you can see come with the prompting here and you can go with the prompt engineering. You can easily learn this prompt engineering by Chargb by itself here. You can learn in the developer format. You can change this Python to curl. You can learn from this here. If you are a developer, you know how to co you can learn this prompt engineering with the coding simple. In the Chargbt we have just write the proms and we'll just get the output from chargbt. That is a simple text format. Now if you're looking to build an application like ChagBT or specific AI assistant or AI agent, you need to learn some basics of APA calling to open EIS models, PTI or another thing. You can change the models number here and you will see the difference in the output according to the model you have selected and the prompting you have user. I hope I understand these points well. You can easily learn all the details of these here directly in the OpenAI docs prompt engineering section. You will learn all the things here. You can see the different techniques have their different formation, different things, different example prompt APA request, how we can just request the API that is according to the model selection, GPTfi instructions, all those things you can learn from here directly to enhance your prompt writing skill as a developer or to build applications by using coding like that. You can see prompting reasoning models, build a prompting Playground directly, you can come here. You can just click here and build a prompt. You can directly click write the prompt here. Can generate the image generator, code debugger, research assistant, you can directly create here. For the audio purpose, you can create the real time prom. You can go with this text to speech. You can write here, instructions and voice. According to that, it will generate the voice. Images. You need to verify the organization and the assistant. So we will talk about the assistant in this session for these three features, you can search in the YouTube or online resources, you can get the information from them. We will just go to AI assistant. We will create some new AI assistant in which we will get understand about temperature and top the well. Now, we are in the assistant, we will just name it like a copywriter. This is a App operator. This is a system of instructions that we will give to AI. Let's take the person of prompt patent. You are an experience or expert at writing RFP, which generates more sales and conversions and your task is to generate your task is to generate a fee according to user requirements. Ask. User subdivided quotients. Which helps you generate. Air copy. That is simple, that we have using simple personal prom pattern and ask me subdivided quotiPPattern. You can select any of them model according to requirements, all those things. I'll just go with the GPT-4 0.1. You can enable the file search code interpreter, so we are not using the code interpreter. It is best for code task, file search, you can add this one here, you can just go for the numb results files. You can add the files directly here. Even you can just disable it. Okay. And text format, you can go with three different type of text format, response format that is text JCN Object or JCN schema, as you can see this is JN schema. We'll just go with the text as well. No, this is a temperature and top Val. Okay? No, let's see. Now, according to my temperature and top value, we will get the output. We can control the model output very easily. First, let's give the instructions. Let's start with this one. I will just write the latest margin first. Now I will give the high. What happens here let's see? I'll just run this one here from here directly. Now it will take in the high it will ask some questions to me. You can see it is a operated staking you can see, hello, I am ready to help you craft winning ad copy to get started. Could you tell me a bit about what is a product and service is asking the question to me. We have seen this one already prompt pattern technique in the previous session. So we can try here directly here to get all those things here. You can see when I give the answers for these particular questions, it will just write the perfect and add copy for me. Now, let's give the answers directly, directly in the one. What product or services are you advertising Watches? Let's give the same example here next two. What is your target audience? Who is your target audience males? Let's give the third one. Where will this ad be shown given Facebook with spot mo Was main goal generate leads increase sales. Let's go with the fifth one. An special offers a unique selling points to highlight brand new watch. Let's give this simple my requirements, then we will see the output of it. That's fat, is giving your answer is well. Not is asking another pions. Great. Thanks for sharing here a few more quick questions to help me make your add even more effective. What is your age range? Let's take 18 to 25. Let's take the second question. That is, what is the style of watch luxury? The question is, do you want to mention the price of discon that is 20, 25% of? Let's for the question. Is there any specific brand name or model included that you can take anything like white. Let's take this as an example. And the fifth one. Any special offer features you would like to example waterflow. Let's take Smart features. Let's give this one. What is asking the questions by according to my answers, my requirements. That is how you can build the AI assistant for the specific application. Now, you can see, perfect. This is my whole ad develop your style, new luxury VT smartwatches and set up your game. This is the ad copy that have this generated by the AI assistant, but just simple giving the instructions here. Okay? Now you can see, you can see here. It is taken the 91 tousen 90 tokens. You can see here. When I click Q here, you can see the in and out. So for the input, you have just taken the 7111 tokens for the output it has generated, it has taken only 479 tokens. So what is the token here, actually. Now, before token, we will talk about temperature and top the value. If I just go to the two. If I put this temperature value for the two, I will just run again, answer like high. I will clear the thirds here. But is clear the thirds. Now I'll just go with the high. I put the temperature value to two. Now what happens here? This resistant film generate the response is taking time. While the temperature we have just learned about this? We can do for low values, the model generates more predictable and focused response instead of just giving all the possible words like that. You can see this by giving the two higher value, the model becomes more creative and diverse output that may produce less coherent answers. It will generate the output which have some coherent answers, which is not effective like that. You can change this temperature and top value as well, like that. No. When I give the high value, it is taking time. There is something failed. I will just give this temperature value to the one point ph that I will just give the gain. Just run again, see whether it works or not. I can see generating. For the temperature I kept the most high value, it is just taking to write all the predictable words, all the coherent reasons. You can see high there and ex to help you to crop winning a copy to get started with you. Could you tell me what the product, all those things? No. You can see this is the output from AI, right? Now, if I change this one again, right? I will run this again. I will just clear this one. I will write this one again. Let's see. It is taking time to produce output, you can see. You can see we get more information here. In the previous case, we have just again the two sentences. I'll write the high converting copy, add copy for it. Just give the answers for this. Now, if I increase the temperature, you can see we have got three to four sentences at the starting. I hope you understand you have observed correct clear? This is how we can change the temperature. If I put this one for just now, please observe this one. At the starting, we have four to five sentences. If I put this one to one or let's take 21, let's take 20.5. Now I will just clear these tokens. Again, I will just write the hi. I'll run this one. Sorry, just taking the whoo. Now you can see, we have got the high ICU interested in creating R copy to get started and make sure your R copy is effective, and could you please answer a few questions? No, we have got two sentences. When I increase the temperature nearly 1.7 to one there. So we have got the more sentences here at the starting time. Now what we can do? We'll just keep this one as balance and we'll just remove this top value to the 10.5. Let's say what happens here. 36. 5.6.3 Overview of Top-p Parameter and Tokens: Just clear this one. So what are the top value just to do? We will learn this thing. You can see top E value is for zero to one. When you set to one, the model consider all possible words options. Okay? Lower values, the mol focuses on the top most likely words, reducing randomness. So for the lowest value. Okay, the model focus on the few most words, it will select only few most words and it will just complete the output in that particular selectable words. When you set the value to maximum range that is one, the model consider all the possible words. It will select and it will just produce the output with the select possible words options like that. We'll check this thing here. I'll just take this 10.5 as like that. I'll just put in a high option. Now, you can see the answer. I'm here to help you create high converting ad copy to get started. Could you tell me a bit about these things. You can control the output of this particular assistant or AML by just changing the top value high or low. If I just keep this one as the same, if I just put the top value to the low, I'll just clear this one right high again. Let's see the output. You can see it is right. So in the previous one, we have seen some possible words. Now it has just taking simple answers where this ad appears like that. When I keep this too high, nearly to one, let's give this one like this one. It is saving, you need to wait up to saving. I'll just write, clear this one. I'll just start with the height. Let's give this one. In the previous one, we have got the very small quiens sentence, few words, sentence. Now, if you see here, for every quotien it has chosen some more words like what is your main goal, who is your target audience? Which platform will you be running this ad on? In the previous one, we have seen which ad platform you're looking or like that? Okay, you can see, do you have any special offers or futures would you like to highlight? So this is how you can control the output words, sentences or all the tokens by changing by controlling temperature and top V. So if you have some confusion, please try by yourself, just go and create assistant in the open air playground. Do by yourself, you will understand the real magic of temperature and top. You can control the output words, tokens, all those things by simply changing the temperature and top V value. I understand these points very well. Now, what we'll check? We'll just copy this one. We will check the how much tokens it has to do that, come to here, just search for open a tokenizer. Open A token, nyer. You will fight this one. You can just click on the open a tokenizer here. Now just paste your output here according to the you have selected the model. I will just go with the GPT 3.5 GBT four. I'll just paste this one, how much my output from the AI istent. Now you can see there are one night nine tokens, 462 characters. You can see how much characters what are the characters it will take for the one token. We can say hello for that, it has taken one token, and this is one token. This is one. You can see there are different tokens. You can see here. Token ID here, you can input contain one or more unique code caters that map to multiple tokens. Output visuation may display the bytes in each token in the standard way. So we can change all those things according to the models, you can see that tokens difference. That is simple. You can come here, you can paste any output from AMR, just keep here. Just paste here, you will see how much tokens it has taken, how much characters it have. It has. That is simple. Okay? You have the tokenizer, you can do with that. Now, other thing is, you can build the A assistance for specific work. You can write the instructions for the specific applications. You can build this assist. You can add this particular assistant to your website, or you can build the apps by using this assistance, ID, all those things. You can learn in the YouTube as well. Okay, I understand these points very well. Try by yourself, then you will learn all those things. So up to now, we have learned some basics of open A, creating A assistant with instructions and some parameters. 37. 5.6.4 Overview of Prompt Optimization Tool: Have another tool which is extension that we help to write the prompts very well. If you forget how to write the proms by using prom patterns like that, you can use this extension that is AI Prom tech. You can get this extension from directly the Chrome extension bar just to go with the Chrome extension web developer. Let's go with the Chrome extension Chrome web store and search for the AI Prom tech, you will get this one. Just write for the EI prom tech. You can find this one first option. That is a prom tech. This is a prompt optimizer. This click here, just act to Chrome if you get this one, act to Chrome. After that, you get this click here again. No this is the prompt optimizer. What happens here when I write any prompt, image, prompt, or video prompt or text to prom, it will enhance the prom and it will suggest a better verse as you can see. Okay. If you are not happy to use this AI prom tech optimizer Chrome extension, we have already learned how to write the proms by using different AI models in previous sessions, you can go with that. To save you a lot of time, you can do with these things as well. Okay, I will show how it works. I will just will need to mention the open E APA key here. Okay? To do that, we will just go to our playground, go for the APKs, create a new secret APK, I'll just take text purpose. Okay, create new secret APAK and I'll just copy and I will just close it and just come here. I'll page this particular EP. Now, save settings. Click on that. API successfully copy it. Now I'll write a simple prompt here, write A full OpticaO AR. Let's do this one. What happens here? Let's check. Let's start thinking here. No, you can see here. We have got the best A prompt here. As an expert A content writer, please write a comprehensive well stuctured article and artificial inigence covering its definition, historic key technologies current applications. You can copy directly here, come here, you can paste this one. You will get the best output when compared to just writing the input by yourself. You can see we have got the example, output here, best. Brief history of key technologies like that. Not only that you can go with the image generate an image like that. Let's see. Generate an image Wheat. Drinking milk. Let's keep this one. There is no problem if there are mistakes, the AI will automatically cover up all those things. No, you can see, we have got the image prompt as well. You can see generate high quality detailed image of an adorable cat wearing a soft light colored per pattern, gently drinking fresh milk from a small, all those things, you can just copy, you can go the image generation prompt image generation tool, you can paste it, you will get the best output. Not on that you can create the app prompt as well, even video. You need to write the basic prompt that is simple. You can see generate video, which shows ion is smiling. So I am looking to get the video prompt. When I give this video prompt, that is basic prom, the AI will automatically enhance my video prom. You can see creative vivid engaging video prompt, produce high quality realistic video of the majestic lion generally smiling the natural savanna, like that, you will copy and just go with the video generation tool paste, you can see all those things. Even you can create an app, you can create the prompt for white pudding tools like lovable.ai bolt to create the AI apps by using just text prom. Just you can see this, I'll show this one, create a app. He helps for students to solve their math problems. There is H output like that. When I give this one, themedically enhance my this prom that is app prom that you can use inward white pit. You can see it is a very big prom. As an experienced A prom writer for educational tools, craft a clear and detailed prom that instructs AI system to help the develop app assign. Now what you can do. You can use this particular prompt in the HGIPT. I will give you the punch hole you can see, I'll just copy this one. I will show the example how it works for specifically AirPm. I will just copy that pass this one. Whole prom. What happens here? The HAGIPT will give the whole prompt. That is white coding prom. You can use this particular prom in the white coding tools. I hope I understand these pawns. But the image prom and the Text to prompt video prompt, it will directly give the prompt you can use in the image generation video generation tools. But for the app prompt, you need to use that particular prompt and come to the cha gibt or other AI models to generate the full app prom. You can see. No, it is refined version. Now, you can use this whole prom in the bitcoding tools like lab.ai, bool dot like that you can use to create anything like that. Okay, you can use this extension. It is a free forever for ti en task, okay? 38. 5.6.5 AI Ethical Considerations: Have learned so many prompt patterns and specialized techniques, prompt changing, all those things in the previous session. Now we will talk about some ethical considerations. Okay, we need to follow these three steps or very most important factors that is avoid bias. So when you talk with the AI, use neutral language that you talk with the colleagues like that, you can talk with the AI, like the friend like that in order to get the best output. Okay, don't need to put a professional text, all those things. Okay? Just go with how the things like talk with each other. Use neutral language and avoid stereotypes. There is no need to put the professional English, all those things. Even if you have some mistakes in the prompt, it will automatically correct and it will generate the best output, and it is very easy to write all those things. Okay. Now, second one is ensure inclusivity. So consider divers perspective. Is the most important thing. We have learned so many prom patterns that is we have seen different methods, prom patterns to get the best output for according to our task. We need to get the output according to requirements by writing the prom which includes the different perspectives in the prompt. I hope I understand these points. We have learned how we can generate the three different versions of prompt right. So in that case, you will get three different proms for the same task. You can see that the three different proms have the different perspective of writing the style of prompt. In that case, you can use that particular three different versions of prom. You can put in the JGB or other EI model. Check each output. You can see then the output is different from the other prompts output. Like that you can write by this, you will get the idea in how many ways you can write the similar prompt in different perspectives. I understand these points. Try by yourself, you will get the idea about it. Okay? Now, th respect privacy. It is a very most important factor in this EI wall because we need to avoid that prompts which have the sensitive or personal information. Don't use your own personal data, but just give as a if your name is John. Okay, don't give your name, address, all those things. But instead of that give a fake or duplicate name or address in order to get the best output when you are required. Okay. When you get the output, change your change with your personal information like that. Okay? Because we need to respect privacy of each one, don't include your personal information or anything. It can be a credit card, or bank account number, all those things will comes under the sensitive or personal information because the AI is learning day by day with our interaction and data also. In that case, there is a chance of putting your data to another persons in which they can misuse, right? In that case, please don't give your personal information or sensitive information while using AI models. I understand these points. There are a lot more ethical considerations you can follow, but these are the main factors you need to keep in mind while just using AI chat words in daily life. I hope you understand these points very well. Let's jump into our model number six in which we will see some opportunities and the future of prompt engineering. Let's dive into that. 39. 6.1 The Future of Prompt Engineering: In this model number six, we are going to discuss some prompt engineering future trends and opportunities. Along with that, we will also cover overview of GenAI and role of prompt engineering in JI is the start with the future of prompt engineering. Okay, as I said, this prompt engineering is a most important skill in which we can direct interact with AI to get the best output from AI. Okay, this prompt engineering is required for every industry in which they can interact with char jibeor or other AI modules to get the best output from AI at the potential level. Okay? But when the char gibter or other AI models are used, there should be prompt engineering is needed to interact with AI to get the best output. I understand the points. So the prompt engineering will help you only when you have the specific skill, like if you are a electrical engineer or a mechanical engineer, if you have the expertise in the marketing or if you have the expertise in the healthcare so and so topic or so and so skill, then this prompt engineering will help you to exhane in that particular skill to save you a lot of time in that particular skill. Okay? I understand these points. The prompt engineering is nothing but writing the prompts for specific applications. Okay, like that, you need to have the expertise in specific skill, then you need to add this prompt engineering to that particular specific skill, then it will the game changer for you. Now, that is most important thing in this world right now. If you have any specific skills expertise, then we need to add the prompt engineering, then you are skill and this prompt engineering combine and it will make your position very high in your company or organization, or if you're looking to get the job in any company right now or industry. I hope understand this ones. You can see there are emerging trends you can follow multi multi model model. So many AI systems are moving behind the text to include images. It is already done in this right now. You can see the Google gm.ai and hag every AI module is integrated with the audio, video and images as well. The prompt engineering will soon involve creating inputs for those mediums. Prompt engineering already is updated right now. So to get the video or images or text or even app, we need a prompt engineering because in the text only we will get the output. You can take image prompt, image prompt is written in the text. If you're looking to get the video, the video prompt is also written in the text. If you are looking to create an app by using coding tools like lovable.ai, bolt AI, all those things are written in the text as a like that, you need to have the specific techniques in your mind and the context. All those things we have learned in the previous sessions, then you combine with the specific knowledge, you can build something new in this world or even you can sell your skills in your organization or company to make money in this AIS world. I hope understand the points. Again, engineering future is fine tune models. So these multi model modules, fine tune models are already, and automation also already is emerged in this year right now. Okay? But the multimodal models, you can check it the Google Gemini. We have all the image generation, text generation, video generation, app creation, all those things in the fine tune model, so business are trying custom models for specific industries. This is most important thing. If you focus on the ChatGPT or Google Gemini or other AI models, they are for the people. Okay, it is not trying on specific application, but it is trying on all the data of the Internet, all the world. Every person can ask any question to chargeability will give the answer. But the fine tune models are different from the multimodel models. Okay, what are the fine tune models? You can see the business or trying custom models for specific industries. Okay? When you, for example, if you are in the marketing industry, for a particular company, the company will take one customer model, trying with their datasets, trying with their marketing strategies, trying with their sales strategies, and that chat board is trained on only that particular company's data, not like ChatGPT or others. When you ask the question related to the marketing or particular details of that particular company, it will give the answers in that business data only, not like ChatGPT it will give the answer for every question. Then requiring prompt tailor to this specialized system. Then here that particular specific person who have the prompt engineering skill as well as the marketing fit. That person is required for this particular businesses to train their customer model. In that case, the prompt engineering with the specific niche or specific skill is needed. Next, integrate with automation. So you can see we have a lot of automation tools like NTN, me.com, and ZapiO and there are so many automation tools we have right now in this market right now. So in that case, we build the AI agents. That AI agents, we need to write the prompt system prompt. In that when you write the system prompt effectively, the AI agent will do the best work. Like that. They also the prompt engineering is required. So how do stay updated in this field? This is how you can stay updated in this field. So just simple. The basics of prompt engineering never changed, right? But the prompt engineering according to them type of AI model can change. So we have learned in the previous understanding the different LLMs, pros and cons like that. For every AI mode, there is something other techniques we need to learn, to get the best output from it. Okay? We have learned the prompt patterns there. With that prompt patterns, you can leverage any AI model. It can be charting Google Gemini or any AI model to g to get the best output. Okay? How you can stay updated with this fit, follow some prompt engineering online websites, forums, OpenAI forums, Google De Mind from Google Gemini forums, and you can follow so many pages as well in the social media, all those things. You can just get the update in this field with new models and tools coming in the future, all those things. Okay? You can follow the AI communities, and you can keep learning. Okay? That is the field will continue to grow, and so you should your expertise as well, because AI is day by day is evolving with the capabilities and advanced versions of their model. So just try these same prompt patterns we have just learned before, add by yourself, okay, and try to add another technique by to get the best output from them. Okay. Then this field is best for you to grow, and you should your expertise, you can improve your expertise in writing prompt as well. Okay. 40. 6.2 Prompt Engineering Opportunities: What are the prompt engineering opportunities we have? The growing demand for prompt engineering, as I said, so every industry want a prompt engineering because every industry will leverage the power of EI as soon as possible because in education, if you take education, health care, marketing, entertainment, where the EI can save a lot of time in writing or generating the code, all those things, their AI is used. I understand these pins. AI is everywhere. So the demand of prompt engineering also very high because they have the specific skill and they can use the leverage. They can use the power of AI combined and this is a game changer for the particular companies, they are looking for the prompt engineering. Okay? Just this is how we can find the opportunities. Okay, for finding opportunities, you can just go for the PLACE platform, start with the FLAC platform like five Upwork, people paraverg.com, like that. Place place your bid or place your services online dot. And even you can just talk with the companies, organizations like that. Start with the consulting, okay? I will explain my prompt engineering techniques to your company team member, they can leverage the full power of AI like that. We can start with LT, or even you can just you can go for the education. You can follow any or you can give your service for particular industry like education, health care, marketing, go with the specific one, start with the specific one, and it will grow as you want. And the next thing. Okay, now, what are the career opportunities in prompt engineering? So now up to now, we have learned how we can just write the prompt to get the output. Now, you will see there some job roles, we can follow. The prompt engineering we have already learned this previous that is writing the prompt for specific application, getting the information from the AI, for the specific role by combining your specific skill and prompt engineering techniques. It is a game changer for any company or organization that looking for, you can go with the prompt engineering field. For other prompt engineering roles require some coding language as well, because by using coding language and prompt engineering techniques, they can build the applications by writing the code and all those things. So some companies will also look for not only the prompt engineering, not only the writing the prompts, but with that some coding language like Python, JavaScript like that. Okay? You need to have to look all those things. Now, next thing is conversational EI designer that is EI trainer. As I said, the company's businesses are looking to train the EI models with their own data. Okay, to build a specific AI hartbod for their companies. So in that case, the prompt engineering is required to write the prompt and response as well. Okay, another name of that conversation AI designer is AI trainer. You can find so many job roles in the outlier.ai, and you can find the AI trainer jobs in directly online, you will get all those things because in this AI training, you need to have the advanced English writing and speaking skills. Okay? Because we are training AI model right? While we are trying Air model, we are trying the data with the English language. In that case, you need to have the advance command in writing and speaking in English. Because if you have the grammatical mistakes and lot of mistakes in writing the English statements, the AI can also generate the output in with the grammatical mistakes. The output contains some mistakes, haloiation words or all those things because we have trained with our mistakes. That is simple. That's AI trainer should have the high command in writing and speaking in English. So in that case, you need to have the great command in English. That is simple. You can become the conversational EIT designer as well. AI Chat booard developer. Okay? You can become chatbod developer by if you have the coding, if you don't have the coding by using the prompt. By using the prompt, you can generate the code from Jatibter or other Cloud, or even you can use the Childbod developer tools like the lovable.ai, all those things, no good platforms. By writing the prompt, you will get directly the app, Chatbot developer, all those things. You can do a lot more things by prompting by using different no good tools. A content specialist, what is an A content specialist? If you have the prompt engineering skill and the content writing skill, if you have expertise in the content writing space, you can when you combine this prompt engineering and content generation skill, it is a gain changing because if you have the skills in the content, you will generate the basic content from AI with your prompt writing skill. Okay. And after that, you will analyze that output, and you will change with your creativity in the content. Will update that particular AI content with your creativity. That is how the content can be from Chat can help you to become AI content specialist. Not only the hajbty can use any AI model, you can get output, and you need to add your creativity. Then this prompt engineering will help in your career group. I understand these points. And the next one is GenAI consultant. Okay? With the prompt engineering skill, you can just go with the organization just reach out them, so I will tell and tell them and tell your strategies, how you can leverage the power of EIN or organization in their organization by automation or by prompting skills or by just developing app or chatboard for their customer support like that, you can go start with the GenAI consultant by just running the prompt engineering and other no good tools as well. Okay, now, how you can find these opportunities. That's I have already discussed with you can go for the fiber and link it in, all those things, you can build a profile with your poofil website in which you will post some testimonials and projects you have done and some introduction video about you, all those things you can do with your freelancing platform, start with the freelancing platform. Their views, and you can go by yourself independently by adding by placing in the social media platforms, you can consult with the organization people or the companies HR. Like that, you can start with this. Many businesses outsource from creation for specific projects, offering lucrative fleancing gigs or jobs. You can go with the freelancing or job opportunities or entrepreneurship. The right. By using this prompt engineering skill only, you can build the AI power tools, apps and services using generative AI and prompt engineering. Okay, you can build the apps by using lovable.ai bol dot new agent three, and we have so many app builders, no good app builders, but just writing the text, you can create the app. By using the prompt digital skill, you can build any app. You just have the one great idea. Okay, one great idea can be done by solving the simple problem right now. So you can become the entrepreneur. Online entrepreneur, just create your own EAP tools, app services, post in the online, and you will also become the best entrepreneur online as well. You can see we have seen all those things. Okay, now, what are the tip? Building a portfolio of proms, showcasing your ability to work with different LLMs, can you give a competitive edge? Okay. So as I said, the prompt engineering is nothing but writing the prompt for AI models, not for the specific model. Not for ich AGB, not only for the cloud, but we are learning the LLM prompt engineering here, not for the specific prompt engineering for specific model. Okay, we need to have the specific skill. It can be the out of EI. Okay. With that, we need to integrate EI in that particular specific skill in case marketing or coding. Then you will save the time. Okay? That is how you can build a portfolio of prompts and showcase your ability to work with different LLMs can give you a competitive edge. That is simple and how to prepare for these opportunities. So these are simple steps you can follow Stay update with the new LLMs and tools. Always look where there some advanced tools or coming day by day like that and develop a specialization, healthcare, marketing, creative writing. Now, it is a very most important skill because you need to have the specific skill in order to add this prompt engineering. Then only you will get the best written from this skill. Everybody understands. Develop a specialization. If you have all day specialization, Congratulations, Steve, add this prompt engineering, combine those things and look for the solution that you can help with organization or companies, or even you can build your own solution. You can offer two companies or individual people and you can make money online like that. Develop a scalization if you don't have and build a network in AI computers to find projects and collaborations. I need to build a network, you can share your expertise in YouTube, Facebook, Instagram, or in the Liner in Aswll. Show you ado and you can find projects and collaborations directly in the social media platforms like that. Now, in the next session, we will talk about what is a fine tuning Andrad. Let's dive into that. 41. 6.3 Basics of Fine-Tuning and RAG: In this session, we are going to see what is a fine tuning and RAG. RAG means retrieval augmented generation. So what are the difference between them? And we will see some examples related to fine tuning and RAG. Okay. Let's dive into that. So what is a fine tuning is actually? We can see definition the fine tuning involves training a pre trained AI model on a specific da set to specialize it for a particular task. Okay? So it is simple we have already discussed about fine tuned models. The fine tuning model is for the Business As or EI for specific application. For example, instead of ChatGPT or other AMRs like Cloud, it is for everyone, right? So instead of that particular ChatGPT or AI models. Instead of that, the businesses will try their own AI models with their own data and techniques, strategies to save the lot of time. Okay. So for that, the companies will use the custom model or base model that is you can see the GPT three here. Okay? We have the suprise section Fine-Tuning models in the OpenAI Playground or OpenAI docs you can follow, you can see the fine tune models or those things from the particular documents in the open AI. We'll get the information from that. Okay? So basically, the idea is that if you want, for example, the company is looking to try an A specific model for that specific dataset. In this case, let's take marketing strategies. Now, businesses or company will take a base model from the Open EA or other AI models company as well. Take the base model. We'll start trying the pre trandEI model, you can see. They will start trying pre trained AI model on a specific data sets with the specific data set to specialize it for a particular task. Okay, in this case, we have taken the marketing strategies. Okay. Now, after training the AI model with the marketing strategies, the base model will become the specific AI model for the marketing strategies. Now this AI model have the knowledge of marketing strategies, and it will give the answers strategies according to the user prompt. Okay? Now, the chargeability for everyone. Now, this Fine-Tuning model is for specific one. That is simple. They will try the AI model on the specific data sets to specialize it for a particular task that is called Fine-Tuning. So for the fine tuning, we required the prompt and patterns, all those things. We need to write the prompt as well. We need to write the output as well. Then from the two, so we required a prompt the user question, as well as we required specific dataset from company that is marketing strategies. When we combine the prompt and the response, when the both are combined, then the AI model will learn from the prompting as well as the response. Then the AI will just predict the user question and I will generate a related response to user. That is how the fine tuning works. For more information, for practical implementation, you can see can check the YouTube videos for more understanding. I hope to understand this model how it will works. Start with the base model, the company or other people will just select a simple base model and it will provide a domain specific task specific data. Example, marketing strategies, legal log documents or their own data, and it will just provide this specific task to base model and they will train with their dataset and train the model to improve its performance on that task. The automatically this model will improve its performance on the task. That is simple. This is what is fine tuning is. Now, how it is a relate to prompt engineering. As we earlier discussed the fine tune models require simpler prompts since they are pre tiled for specific task. For fine tuning models, we require a prompt engineering in which they can train the model with their own dataset. So in that case, a prompt tering plays a major role in these fine tuning models. A. Okay, you can see the example here, general model, summarize this news articles for a teenager. Fine tune model summarize the model is already tuned in to create summarizes for teens. So for example, you can take. So we have just trying AI model base model. We have just a fine A model for the specific task that is summarized this news article for a teenager. It is a specific task. Okay, now the fine tune model is what happens here. So we do not require to write the prompt like this in the fine tune models. We will just write summarize because this fine tune model is already trying to create summarizes for teens. There is a small difference in that, you can check by yourself here in the PPT. Let's see what is Retrieval Augmented Generation. 42. 6.4 What is Retrieval Augmented Generation (RAG): Now, you can see the definition of RAG here. RAG combines a retrieval system, example, database or search engine with a generative model to provide accurate up to date information. Now, how it even works, okay? No, up to now, we have seen prompt engineering, fine tuning, and what is a RAG here. RAG is nothing but it is also like fine tuning, but in the fine tuning, we will try an AI model. But in the RAG, we will provide data. Instead of just training with our data, we will just provide our retrieval system, our accurate and up to date information. In the fine fill model, we need to write the prompt and response. In the RAG, what we will do we will just add a retrieval system in which the AI model will automatically take the information already is stored in the Sans source like databases or search engine like that. It will take the sources from real time database or search engine, and it will provide the output to user. Simple. In the fine tune, we will just try with our dataset. But in the RAG, we will retrieve the information from different sources. It can be the database or search engine e. So RAG combines a retrieval system, example, database or search engine with a generative model to provide accurate up to date information. That is how RAG will work. You can see how it works. The retrieval system fetches lament document space around query. The generative model uses that retrieve information to generate a response. That is simple. Okay. Now, what is the relation to prompt engineering? You can see here also the prompt engineering is required because the prompts guide both the retrieval process and the generation process. Okay, so we have the generative model, and other hand, we have the retrial system. So we need to write the prompt which the model will use the retrieval system when it is required. Like that we need to write the prompt in which we can just guide AI model to take the retrieval system in work and provide the best output here. Which works according to our prompt writing skill. So you need to have the knowledge of how the retrieval system works and how generative model work. In this case, we just learn with the prompt engineering only. Okay? Let's see the example here, real prompt. Search for the latest research and climate change, you can see here. For the generative prompt, we will summarize the retrieval documents in three sentences. It is a simple work slight example, workflow. So for the retrieval prompt, what we will just write to the AI model? Search for the latest research on climate change. Now, the AI model will first search the latest research on climate change. Okay, after that. The AI model have the latest research climate change data now. Now what happens, we will write the generation prompt, summarize the retrieval document in three sentence. Now, the AI model have the latest research data on climate change. According to our Generation prompt here, the AI will summarize in the three sentence. This is how with the retrieval prompt and Generation prompt will combine both works together to achieve a specific goal. I hope understand this ones, how it is related to the prompt engineering. No. What's the difference between fine tuning and RAG? You can see the XPEtFne Fine-Tuning RAG purpose is specialized model for a task. Here we are just integrating the external knowledge to provide a response according to the user query. You can see. Okay, we will fine tune a specialized skill or data set, but we are integrating the external knowledge to AI model here. Now, what is the data dependence here? Requires quated datasets. We need to build a dataset in order to fine tune a AI model. But in the RAG we do not require. We will just require the database or external APIs like that. You can see here and prompt usage. We need to write a simple prompts and response to fine tuning the AI model. In the RAG, we need to write the enhanced prompt flexibility because we need to write the whole system prompt or when to use the external database external source to provide a user response. We need to enhance the prompt flexibility as well. The real time updates, it is a static knowledge and it is a dynamic and up to date information because why it is static knowledge because we have Fine-Tuning AI model with our dataset which is a fixed dataset. Okay? So if you want to provide the updated dataset to fine tuning models, again, we need to try a fine AI model. With our updated information. But in the rack, we do not require to try AI model. We will we will just update the databases or external sources which are real time automatically, and it will take the real and update information from the external sources, and it will provide the best output according to the user query. I understand this points. Let's see the examples for better understanding. Fine-Tuning is, for example, you can take domain as a legal contract. But the prompt, we will generate a model that is summarize this contract in plain English for a client. In the Fin tune model, summarize. It will summarize. I because the Fine-Tuning model is already trying to summarize this contract plain English for a client. We need to write the prompt like summarize. In the final model, they will automatically summarize the plain English for a client. That is simple because we have trying to model with the legal contracts. I understand these points. For the RAG example that is medical research, they will write the prompt as retrieve the recent articles on alzheimer's treatments and summarize these findings. It is a retrieval prompt and it is an generation Generation prompt. So we have just write instructions to retrieve the answers from the external sources. After that, we were just told to AI summarize the findings here. Understand these points very well. So let's summarize the learnings up to we have learned that is fine tuning customized models for specific use cases, and a RAG combines retrieval and generation to produce dynamic accurate and context sech outputs, and both approaches complement prompt engineering offering a new ways to enhance AI performance. So this fine tuning and RAG are best according to our requirements, and both approaches will complement the prompt engineering offering a new ways to enhance AI performance. I hope I understand these points very well. Now let's jump into overview of generative AI. Let's dive into that. 43. 6.5 Overview of GenAI: In this session, we are going to discuss overview of generative AI in which we are going to see some basics of GenAI. How does GenAI work, real world applications of GenAI, future of GenAI, and the roles and responsibility of prompt engineering in GenAI. Let's dive into that. Now, first one that is basics of GenAI. So what is GenAI? Generative AI refers to mods that create a new content text, images, code or other based on inputs or prompts. It is simple like HHIPT Cloud or other AI models which create the text, images, code, all those things. Okay? The examples of GenAI you can take the HCIPter or other AI models. You can see the examples here, HCIPT generating essays and answers, mid journey, creating art, co palette helping developers with write code. We have different AI modules right now like lovable.ai and bolt new N to create the automations. All those things will comes under the GenAI which generates the text, images code, all those things. Now what the difference of GenAI here. Unlike traditional AI, which focuses on recognition or prediction, GenAI focus on creation. You will create new product with our inputs and proms. Unlike traditional AI, we have so many human clicker photos, human written text already. But with the help of GenAI, we can create new things by using just prompts where we can create any type of photo, image, any type of content, any type of code, but just writing and requirements in the form of proms in the GenAI applications, ter, for example, habt Image N and app development AI tools like db.ai. Okay, it comes under the basic or AI applications. Let's focus on the next one, that is, how does GenAI work? As we already learn about Charge or other AI models, these models are trained by large amounts of data, for every vast dataset we will just talk about, you can see. It uses large scale machine learning models trained on vast datasets. Vast data sets means very large datasets to protect and generate content. For example, you can take the ChatGPT cloud like that. It will generate the output based upon our requirements or input. Okay? The question can be anything, but it will generate the response because the ChatGPT and other A models are trended by large amounts of data sets, by using Internet sources and all those things. It uses large scale machine learning models, trend and vases to predict and generate the content like that. Now, what are the key models in GenAI? You can see text based on GPT-4 Cloud, image based tali stable dificien or Sra. You can take this one. Multi moodel, Gemini and GP division. Gemini and chargeable, also the multi model Applications, you can see, we can create the image, audio, video, and text as well in the same chat you can find in the ChatGPT Cloud, even you can find in the Google Gemini, Microsoft Co Palette, all this comes under the GenAI. Now, what are the real world applications of GenAI? You can take any industry. There is a use cases of GenAI. For the specific tasks, you can take automated content creation course, quizzes, and summarizes and other tasks in education sector, even you can take in the business like drafting emails, generating reports, creating marketing copy, all those things in the business sector. In the creative fields, we can write scripts, content scripts, design art and creative content for our use cases. In the healthcare sector, you can summarize patient data or generating treatment plans, and there is a lot more industries there where the GenAI will help to save a lot of time in di task as well. No ethical reminder. As we know, while GenAI is powerful, it must be used responsibility to avoid generating misinformation or biases, which is the most important thing here. We have already learned in the previous model that is ethical consideration. Okay, we need to focus on main that is generating misinformation or inaccurate information to user, which they can which can harm them, right? That's fine. So whilGN is powerful, but it will do some mistakes. Okay? It must be used it with a responsibility and avoid generating misinformation or biases like that. So we need to avoid providing the personal information, and we need to use the correct information, and we need to check the output of GenAI applications, whether it is correct or not, by using Fact Check List Prompt Pattern, we have just e in the previous models. Okay, I've understand these points very well. Now, Let's see the future of GenAI. So more personalized EI. So this is very most important thing. Tell response and outputs based on user profiles, increase multimodal capability, seamlessly, combining text, image, audio and video generation. It is already done in the Google Gemini and other Air models as well right so democratization. You can see tools becoming more accessible to individuals and small businesses. We have different tools for the specific use cases that even the individuals or small businesses can also use in daily life to save a lot of time and to build something creative in the EIS market. Okay? Now, let's jump into our role of prompt engineering engine AI. 44. 6.6 Role of Prompt Engineer in GenAI: Session, we are going to see some roles and responsibilities of prompt engineering engineer. So as we earlier discussed, what is the role of prompt engineering, what the prompt engineering will do GenAI or in companies when we hired for the prompt engineering. We have already learned all those things, but in session we are some in depth information of role and responsibilities of prompt engineering in which you can prepare for advance while you applying for the prompt engineering roles like that. In this session, we are going to see what will you can see core responsibilities of a prompt engineering, applications of prompt engineering GenAI, skills needed for prompt engineering, challenges and ethical consideration and impact of prompt engineering and I success. Let's start with the first one. Now, core responsibilities as a prompt engineering. You need to have this five skills, five different types of skills as a prompt engineering you must have in order to become the prompt engineering, because you can see designing effective prompts, testing and refining prompts, model specific optimization, exploring prompting techniques, documentation and reporting. These are some five steps every prompt engineering should do in every organization to become a successful prompt engineering. So what are the responsibilities? Our responsibilities can be different according to the company's job description or the requirements, but I will just give the core responsibilities here that every prompt engineering should have the knowledge about to become a successful prompt engineering, designing effective prompts. So we have already learned how we can run the effective proms according to the AI models how we can test and refine the prompts, according to the output by changing by analyzing the output, according to our requirements and the analyzation of output, we can refine the prompt. Up to, we can get the perfect output that we are looking for it. Okay? First, we need to write the prom. After that, we need to test that prompt whether it works in every model or not or we need to choose which type of air model is best two done the task. When you write the effective proms, you need to test and you need to refine the prompts. You will write first prom. After that, you will test that output. You will check that output and you will analyze that output. If the output is code it if the output is not very well, you need to refine the proms again and again up to the output you are looking to get from AI. That is simple. This comes under the testing and defining proms and what are the model specific optimization. Here, as the audio know the ChatGPT Cloud, they have different AI models version like GPT-4, five, 4.1, 3.5 like that. So according to the models output, we need to refine the prompts as well, right? For advanced AI models, the output can be the effective for first to prompt. But in the very versions or less effective models like GB 3.54 like that, what happens, you cannot expect the perfect output or best output for the first two prom. In that case, you need to refine the proms again and again in that in which you can get the output. According to that, we need to optimize the prom model model specific, whether it is GB 3.5 GPT-4 or GBT five, according to analyzing of output, we need to optimize the specific prompt for our specific task in which it comes another model specific optimization and exploring prompting techniques, we have different prompting techniques you need to explore. You need to learn day by day and according to the update of AI models like that. The prompt engineering is nothing but writing the prompts for every use cases for specific application and by using the different AI models, not for the specific AI model. So you need to have the skill, you need to have the ability to write any prompt for any use cases in the specific manner for every AI model. Okay, I understand these points and documentation reporting. You need to show each and everything how you start with the initial prompt, how you analyze the output, and what are the drawbacks of output, how you refine the prompt again to get the best output, and how you optimize the output. After that, you need to show them what are the techniques you have used in your task, and you need to document and report it to your higher position. Okay, that is simple. This are the core responsibilities, and as I said, the responsibilities or the responsibilities are different according to the company requirement, all those things. These are some core responsibilities. Let's say the applications of prompt engineering. So as we already learned about the content creation, what are the applications of prompt engineering in GenAI? So the GenAI is ChatGPT or the Cloud. There are other AR models as well. We will use prompt engineering to get the specific content from AI. In that case, GenAI is a whole chat and what are the applications of prompt engineering here? By using the prompt engineering skill, we will get the specific knowledge from EI. It can the content creation, customer support, education, health care, creative fields or automation like that. Okay? I understand these points. Next three, what are the skills needed by prompt engineering? As we discussed as a core responsibilities, we need to have all those things. You can see understanding of GEI models. You can understand by AI models for writing the prompts, but testing the same prompt in all the different LLMs to understand how the AI model will work, how the GenAI models work like that. We have already learned this one that is understanding different element pros and cons capabilities like them. You can understand GenAI models by using that technique. What are the linguistic skills? Update you to write clear concise and unambiguous proms. This is very most important thing. So even though if you prompt have some grammatical mistakes, it will just clear all those things. But if you have the mistakes in the core writing sentence, the AI can just give the misinformation like that, in which you can use some extension to enhance your prompting skills. As you can see, we have seen the chrome extension that is AI promptkEI prompt optimize to optimize your prompt to enhance your prompt teting skill in which you can get the best output. So with that AI extension, you can write a clear concession and ambiguous prompts as well. Now, let's solve let's see the problem solving. So you need to have as a prompt engineering, you need to have the problem solving skill because you have the power of AI. You have the power of writing the prompt for AI in which you can solve the complex problems by using the AI. In that case, if your mind have problem solving skill, which contains analytical thinking to debug and optimize proms, you need to have the problem solving skill because you have the AI. When you have the problem in your mind, you can solve by using AI by utilizing the prompt engineering skill. Okay, I hope understand these points very well. The fourth one is adaptability. You need to stay update with the evolving AI tools and techniques. It is very much important because AI is evolving day by day. Something new tools, apps are coming in the market, which blows your mind. In that case, you need to stay update. It is very most important because you can follow some EI communities or even the YouTube channels, Instagram channels like that, you can also join the forums to stay updated with these evolving AI tools and techniques of prompt engineering. Now, what are the domain expertise? It is the most important point that is tailoring proms for specific industry or use cases. So as I said, the prompt engineering is writing the proms for specific application. So to write prompts for specific application, you have the knowledge of that particular specific application. Okay, I can't give me anything marketing, it can be education, all those things. Okay? When you have that specific domain expertise, you can leverage the power of AI. You can integrate the AI in your expertise to save a lot of time or to create something new, which change the word. That is, I hope I understand these points very well, the skills needed by prompt engineering, I required thing. Okay? No, let's see what the impact of prompt engineering on GenAI success. As I said, if you write the prompts for specific model in the great manner, the GenAI also generate the response in the effective manner. Okay, that is simple. Can see your skilled prompt engineering can enhance the productivity and accuracy of GenAI applications, and it will save the time and resources by reducing trial and error kills and enable businesses and individuals to unlock the fill potential of GenAI I tools. So GenAI success is always dependent on the prompt engineering only because you are going to try the AI model to do the task to done that task. It can be the generation text or image or video or any app. Are the major role in the GenAI because you are writing the prompt. You are guiding the AI, you are training the AI. Okay, if the prompt engineering failed to train AI, then the Output of GenAI also be wrong. That is simple, right? So that's why the skilled prompt engineering can enhance your productivity and accuracy of GenAI by training AI in effective manner. The prompt engineering also save time and resources by reducing trial and error cycles. Sometimes when we try AI model, we have some trial and error cycles as well. But you have the expertise in the prompt engineering for the specific application. You have so much experience in that. You can save you a lot of time and resources, but trial and error cycles. That's why here, a skilled prompt engineering can save a lot of time and resources, not only the prompt engineering. It is a very much port that is skilled or prompt engineering, in which you need to have so many years of experience like that. Next, you can see a skill the prompt engineering can enable businesses and individuals to allow the full potential of GenAI tools. When you try A model for the specific use cases, then the GenAI application will offer the individuals and businesses to unlock the full potential of GenAI tools, like the Chachi PT Cloud and other Als are doing right now in this AI world. Okay, I hope you understand these points very 45. Final Thoughts: Congratulations. Up to now, we have learned so many prompt patterns, techniques, and different ways of using GenAI applications like chatty, all those things in effective manner. So you can use these GenAI applications in daily life to improve your content skills or any skill, or even you can save a lot of time in producing some value to our customers or the organization or even for yourself as well. Okay. Now, to get the full potential in your work as a prompt engineering for better understanding, I'm provided the whole document of this master prompt engineering course, basic to advance by explaining each and everything that I've explained in the videos like, number one, what is the prompt engineering. I have explained each and everything in this document, you can find it in the document section. Even you can find all those things in the document itself, even you can get help with this document as well, you can buy just reading it. I also shared the chat link in which you can easily just click here and even you can go directly in the chat and you can directly just write and you can directly write the prompt that I have explained in the video particular. I hope I understand these points. Not only that, I have just provided each every prompt method and explanation in detail, you can find all those things in the master prompt engineering document that you can find in the course document section. You can find it by easil. Okay. Not only that, I also just given the assignments for each module and PPTs of that particular model, I have shared all those things in that course. Please download yourself and you can and you have the final project which is based upon all the course. When you complete this course with the instructions, all those things, now you have the ability to understand the writing proms for specific use cases and understanding the different LLMs and pros and cons and how to optimize proms, how to refine, just to take this final project, complete it, and according to the requirements and the steps follow the steps. Now by completing the all assignments projects, now you have the ability to understand the GenAIs capabilities and limitations will help you harness this potential in your work as a engineer. Just to complete assignment project and Jewell done updated. And even you can find the advanced course, or if you are looking to build applications, AI agents, you can follow my course profile. I have already created the pe coding course and master agents and other courses, you can follow Okay. And if you find this course is valuable for you, please don't forget to share your feedback and ratings in which it helps me to improve my content for you. I understand these points very well. If you will meet in the next course till then, good luck. Thank you for joining this course.