Ai Prompt Engineering Masterclass for Teachers | Master Ai in Education

1. Overview & Introduction to Masterclass: Hi. Welcome to this amazing master class that is prompt Engineering for teachers at where you are mastering the art of using AI in your work daily life as a teacher. So before going in that, myself shakes CFOL, and I have one year of experience in the prompt engineering as a freelancer. So in this master class, I am going to teach you how to write the prompts and how to use different AI models like Cha GPT, Gemini, perplc.ai, DbCkGrog and to use the Google AI studio to generate the videos or to chat with the EI and to build some EI apps. Not only that, I also cover the Notebook M platform in which you can generate the mindmap for your students or to generate their FAQ quotients and other as well. Okay. So in this course, you are going to learn and we are going to cover different prom patterns in which you can understand the ALNs like hgptperplex.ai, Gemini, Desi and other AI models. So how to use them effectively in your work daily life as a teacher for generating the lesson plans or quizzes like that. Okay. So by the end of this course, you have the skill how to use the different AILLM models for your teaching teaching experiences like generating the question and answers, getting the feedback from EI for your personalized students. And we also generate the quizzes according to our topic, and we also cover how to generate the specific content for yourself, right? And you can get the help with these EI models to understand your student behavior and to teach them personally with your personalized teaching practices that is suggested from the EI. There is a lot more things we are going to learn in this course. So by the end of this course, you have the ability to use the AI models hagPD perfs and other EI more than six or seven EI modules, you have the ability to ability to use them effectively in your work daily life as a teacher, get the better results from the II modules for you. It can be a genetic thing, lesson plans, it can be a quizzes, and how to understand your student behavior and how to create personalized learning practices according to your student behavior, that is their habits, according to their marks, all those things. Okay, and get the feedback from EI to solve their problems and become more creative at your teaching. Okay. Not only that with this course, you will stand out from the competition as a teacher and get more value in your society or in the workplace. This skill is not required any coding language, and if you don't have a technical background, so this course master class is designed for the beginner and you do not need to learn the coding or you do not need to have the technical background. Right. So if you have the creative mindset, ability to learn the new things, and that is enough for the master class. So we will go from basic to advance, and by the end of this course, you have the ability to understand the different AI models, how to use them Fly, and how to get better and how to become more productive or teaching with the help of this AI in this AI world. Okay, let's dive into our first model that is introduction to AI in education. Let's dive into. 2. 1.1 What are Ai Large Language Models?: Let's start our model number one, that is introduction to AI in education. In this model number one, we are going to explore some basics and foundation of AI in education. We will see what our AI language models is. We will see examples and definition of the AI language models. We will see why it matters for teachers to learn this prompt engineering in this EI era as a teacher. We will explore some benefits of AI in teaching and we will see some challenges and ethical considerations we have to keep in our mind while using AI while using AI in our daily life. As a teacher. Let's start from the scratch. That is what are EI language models are. LL means a large language model. You can see here, large language model. If you think here, model model that means AI model, which is base model. Language means that human that speak or write large means large amount of dataset. That means the large language model is a type of AI designed to understand and generate text that sounds like it's written by human. That means a base model, a base model, I model, which is trained by large amount of data. It can be some particular specific data for the subject or even the large amount of datasets. For example, you can take hagibi. If you ask anything to Chachi Bt it will generate. That is actually called a large language model, okay? You can ask any question to this ajibit or any of language modules out in the market, it will give the answer because these language modules are trained by large amount of datasets. That's why these all language models are called large language models. Is a type of AI designed to understand what it happens here. When you ask any question to the AI tools, like for example, you can take ajibiT, it will generate some text. If you observe there, the answer is in the human text, not only in advanced English or not in this. The output is depend upon the input. That means you are prompt. If you write in the basic English, it will generate the answer in the basic English that humans we will chat like that. That's why the chargeb have the great best capability in writing the human like text when compared to this type AI tools. We will explore all those things in later lecturers, but just see this what are language models of. It is a simple a language model, which is tied by large amount of datasets. Generate a text that sounds like it is written by women, or it will understand our basic need and it will generate a text, which is written by women. That's it. That is a simple. I hope you understand these I language models. You can see the examples here Gemini, cloud.ai, ha GPT, perplexing dot A, Microsoft Copalt. This is not only these I tools, there are a lot even more AI tools out in the market. We'll just see this, for example, we have taken. You can explore more in the online search. So we all know about what is language model is. Why it matters for teachers to learn this prompt engineer to learn how to use these I tools in effective manner in our work daily life. As we know, the AI is evolving in all industries. Not only the education point of view, we can take anywhere. AI is improving our productivity and it will saving more time. Saving the times by automating repetitive tasks like that. As a teacher, you should know how to use AI. The power of AI in our work, daily life to improve our productivity or to save our time by automating repetitive task. Ensure that, we can focus on more creativity to explain the best subject to students, which help the students to memorize all those things. You can take EI as your assistance to make your subject even more creativity or you can take some storytelling from AI, which matches your topic, in which you can explain your students in the form of storytelling of particular topic that you are looking to explain, in which the student can easily memorize that story. That means it is equal to a subject that you have explained to the student. I hope you understand this point. This provides ideas and support for lesson planning for you and it can enhance the personalized learning for students as well. So not only, there are so many important points that this prompt engineering, the art of using AI in your work daily if can be very beneficial use if you learn how to use in effective manner. Let's see some our second thing that is what are the benefits of AI in teaching? 3. 1.2 Benefits of AI in Teaching: So here, what are the benefits of AI in teaching? You can see there are four major benefits of AI in teaching right now or in the future. That is the first one is personalizer learning. What is the personalis learning? If any student have a doubt, they can go to some particular teacher to solve the doubt. Right now some students have their you know that some students have the inner feeling that they cannot directly ask to teacher to solve their doubt. In that case, the student will go to the EIGPTTGTs. They can ask a simpler question to solve their doubt. That is called personalizer learning in which the HGPT or EI will help the students to solve doubt. That is great way, as a teacher, what you can take in this, the benefits of AI teaching is like that. You can take this personalized learning into the students and you can await the students to use AI tools as a personalized learning tool in which they can get more creative at their work. Even it can help you as well to improve your subject knowledge and to improve your subject matter to solve word dots as well to solve the student's doubts. I hope you understand this. Why? Let's say second one that is time saving automation. I means saving that time machine, automation. What you can do? We are not going to the technical bit of here. Promially means writing the prompt in effective M. That's it. What is the time saving automation? Even without coding or any tools, just by writing the prompts, you can write the automation prompt in which the AI will generate all of the information for you just by writing your needs. Yes, we will see all those things in later lectures, see some benefits of AI in this teaching field. I hope you understand. It will save a lot more time in your repetitive task. Instead of focusing on that, you can focus on more creativity. To enhance your teaching skill or to enhance your engagement of students in which you can impact in students life to change their lives. That's a simple. That is the benefits of A teaching to save the time automation. As I said, creative inspiration. Creative inspiration means by using AI. The AI is beneficial is for you when you have a specific subject in that. For example, if you have mathematics, you are expert at mathematics, if you use AI in the mathematics, so you can just solve the many problems instead of taking the whole stuff, you can go in specific knowledge in which you can solve the real world problems to save or you can literally save a lot more time in solving complex problems by using AI with your specific subject expertise. I hope you understand. In that you can think in creative inspiration because all the basic things and all the researching things, I will take now it is your time by using your creative mindset. And problem solving. Still, the EI and can combine to solve the particular problem in which it will take more time. AI can do the best in less time. It can solve the problem in less time. I hope you understand this creative inspiration. You can focus on more creativity rather than doing the repetitive task or stuff that it can be done by the EI. Then you can use this AI. For you. Why we need to avoid this great technology? That is a third point and what are the fourth point that is enhancer engagement. As I said, you can use this AI as a story telling, which is related to your topic. The students will engage more in story telling rather than in specific topic that we will explain. If you are explaining some particular topic, instead of directly explaining them the stuff, you can take that stuff, you can convert that stuff into the great storytelling. In that case, you can take help from EI. That AI will generate the story that is which matches the topic that you are looking to explain to our students. We can take the storytelling and just use inward teaching, the students easily engage in your topic discussion and they will memorize it because the story will memorize all your subject, all your teaching in their mindset, in which you can increase the student engagement in your teaching skill. I hope you understand this point very well. These are some benefits of AI in teaching right now and in the future. Or instead of that there are other benefits, you can know this. Okay. 4. 1.3 Challenges and Ai Ethical Considerations: Challenges and ethical considerations. EI is very beneficial at some work, but EI have some other side that is negative side in which AI is trained by large amount of datasets as we earlier discussed that large datasets will make AI to generate some. Even though some large datasets have some mistakes, no data is 100% accurate. For that, when the datasets are used to train a specific AI model, the large datasets have some Mistakes. The datasets have some wrong data also, which causes the module, the module is dependent on the datasets which are trained. If that data have some mistakes, then AI can generate the output which have mistakes like that. AI bias means the AI output, have some mistakes and inaccuracies in their output. As a teacher, we should know because sometimes we think we have, for example, if I have some specific subject knowledge about physics, I am going to explain that is about about light properties or all those things. If I ask agibtive to write some properties of AI, it will generate the answer. Even though if the Char gibt have some mistakes, what I am I am think that is I am wrong because I can generate the best output and there is accurate information, but not like that. AI can do mistakes. For that, as a subject matter expertise as a teacher, you should know the topic which are looking to take the help from EI because it is very important to review AI outputs critically because it helps you to up, take the right information AI, information from EI and explain to your students very clearly and accurately. So for that, you need to have some subject knowledge that you are looking to. Even anything that you are looking to get from EI, you should know about that topic. Then only the EI can be beneficial for you. Otherwise, it can do some mistakes. You can use that mistakes as well without reviewing that. I hope you understand. That is a AI bias. AI Bans means the outputs of AI can have mistakes and inaccuracies. In that case, what you can do here, always review AI outputs critically. Especially when using them in your lessons or teaching in your classroom. I hope you understand this point very clearly. That's our second point that is privacy and data security. So this is very most important thing. The AI is learning day by day with our answers and questions that we're asking to the AI tools. For example, if I ask HAGPT about something, question, the AI will generate the answer from based upon our prompt. If the answer is not correct, I am tell to EI, this is I am needed. I needed some adjustment in this. What happening there? It will generate the second response by adjustments according to my requirements. But what's happening there? The EI is learning from MSD, like that. The AI is learning from MSD to improve their responses. In that case, if you provide your personal data or anything data which are not are publicly shown, that I can take that data. When another person or even strange person, if you ask any prompt to the HGPT, unfortunately, it can be present that data that person. In that case, your privacy or your personal information can be disclosed to the public. For that, you need to avoid sharing your privacy or personal data to these AI models. In that case, what you have to do use AI tools with strong security protocols and avoid sharing sensitive student information or yourself or information. Further, you need to keep avoiding using personal information of students to avoid this leakage of data. In such a case of phone numbers, emails, credit card, debit card numbers, all those things. Just keep in mind that. Third word that is very important over reliance on AI. As I said, AI can do some mistakes without removing the outputs, without input and creativity of your mindset, if you just overly dependent on the EI, it is not a beneficial for you because EI is a generator output, but is not equal to the human creativity, but it can done in the fastest time of possible. If you take, for example, if you're looking to write a lesson, some particular topic explanation, the AI can write in seconds, but it can take for you as 1 hour. In that case, you can use AI to write some basic lesson explanation topic. It will take some seconds to generate the whole topic explanation. In that case, you can take that explanation and just adjust with your creativity and expertise in that particular subject or topic, then explain it. This is how you use AI in daily life as a teacher. So that's why don't overlay on the AI. First, in that case, while AI is helpful tool, it shouldn't replace human judgment. Creativity or empathy is to assist, not take over. EI is never replaced the teacher's job. But the one who know how to use AI in education as a teacher, they can replace you. For that, you need to learn this skill right now in future in which we can standard from the crowd or we can make the things happen with our subject knowledge expertise. That is the challenges ethical consideration of this. Prompt engineering in the AI teaching field. Let some key takeaways of this model. AI and LLMs are powerful tools that can transform your teaching by saving time, personalizing lessons and sparking creativity. As we earlier discussed all those things. Responsibility, use it critical. Be aware of AIS limitations and always critically evaluate outputs. It is very most important even every AI tool like hGPTGemni have their limitations, you should know about that as well as you need to evaluate each and every output. So just write the prompt and take the response from directly EI, but you need to critically evaluate and see review it, proof read it. Then only we can use this EI as a beneficial manner. If you miss this step, reviewing AI's output that you are missing more, you are taking the output have some inaccuracies, which can led you to the wrong direction of your subject or explanation. I hope you understand this point also. The last and I always say that is you are still the expert, AI enhances your teaching but doesn't replace your role as an educator. So that is very most most of the YouTube gurus or online influencers are telling that AI can take the teacher's job. Yes, it is through this the AI can take over some jobs of teaching. But as an educator, if you know how to use AI, then that AI cannot take your job because A the educator, who know AI? And who know how to use AI in the teaching field can take over your job. For that, you need to learn how to use AI in work day life as a teacher. For that, this is a course for you in next module, we will explore some basic prompt patterns, foundation and what is the actual prompt engineering is and how we can use this prompt engineering as a teacher to impact some students life as well as to improve our productivity in our daily life as a teacher. Let's dive in over model one in which we are going to see some foundation and basics of prompt engineering. Let's dive into that. 5. 2.1 What is Prompt Engineering?: Hi, let's start our model number two that is foundation of prompt engineering. In this model number two, we are going to see some basics and foundations of what is the prompt engineering is and we will explore some example and we will see how to craft some effective proms and we will see some examples of it and we will explore some basic and important prop techniques. As a teacher, we should know how to use AI tools with these techniques. We will explore each one of them with detailed examples and more deep in that. Okay. Let's start from the scratch. That is what is prompt engineering is. As we earlier discussed about what is the LLM, what is the AI as well? In this model, we are going to see what is the actual prompt engineering is. If you see here, that is a prompt engineering. You can see the definition of here. Prompt engineering is the art and science of writing clear and effective instructions to get the best possible output from an AI language model. I hope you understand this part. Prompt engineering means writing the best instruction or writing the prompt, which guide the AI to get the best output for our requirements. Right. So what are the AI elements? You can take any AI chat booard in market right in Internet, like Gemini, Cloud, Chat GPT, perplex dot EI, Microsoft CPL, and there are other AI tools as well. So prompt engineering means writing the best and effective instructions to guide AI to generate a best output per our requirements and needs. It can be anything, write the particular lesson for this subject or a particular explanation for this topic. Even you can write anything like that. But the output of AI module is the best when your input is better. Even with thought your input is good, the output cannot generate the best output because your input is not well. Like that. That's why this prompt engineering concept is being evolved by patily three years because it is very important. See, EI language bubbles have more and large datasets, so it can give the best output because it has everything the knowledge that we don't have. For that, we need to know how to use these AI tools at maximum potential to get the best output. For that we need to know how to write the best from best and perfect prompts that guide the AI to generate the best maximize output for our requirements. But thus this prompt engineering is very important to learn. That is all about prompt engineering. Let's see, what are the examples? For example, as I said earlier, you can see here the two difference between the prompts, P prompt, effective prompt. If you see, tell me about history. So the major definition of prompt engineering is writing the prompt for specific application, even which will help I model to get the best output for the specific team. This prompt engineering works very well for the specific application. If you write the prompt for specific application, the AI model think like that need to give the perfect output for the given query. The prompt engineering means writing the prompts for specific application to guide AI model to generate the specific output in which we can expect the accurate of output. Let's start the two difference between the prompt. That is poor prompt and effective prompt. Let's see the difference of this. If you see here the poor prompt, how looks like, tell me about history. Okay. It is also a query. It is also a question. But the AI think, I will tell about history, but in what aspect? In what specific application? The EI didn't know, but that it will just generate the output which is related to the history. If you see her effective prom, how it look like. It will go in specific application. If you see here, explain the causes and outcomes of World War two in simple terms suitable for eighth grade students. So if you think here, this prompt is very well written and effective for the AI model to guide to get the best output here. It has some great points. It will going in the specific manner. You can see here. Explain the causes in the specific one instead of writing all about summary, we just need a causes, which is a specific one point, the outcomes of this is a second specificity of in this prom of ol Dovar two. Okay. Here, instead of giving the World War two, if you ask AI to explain the causes and outcomes of world wars in simple terms. Sit will generate the World War I causes and outcomes at the same time, World War two causes and outmos, even some other World Wars. If you need the specific Answer from AI, then you need to go by writing the prompt in specific, which will help the AI model to generate the output which you need only instead of getting all the stuff by AI. But that writing the effective prompt is writing the clear and specific instructions to AI to get the best output which have the accuracy that you need. I hope you understand this fine. You can see here. Causes and outcomes is one specificity, World War two is another specificity and simple terms suitable for eighth grade student. It is one specificity. For example, if you think if you write this like this, explain the causes and outcomes of World War two in simple terms. TheI will generate the simple terms only. But if you write the specificity as a teacher, if you are explaining this topic, World War two outcomes and causes eighth grade students, you need to mention the students also because the AI will generate the output in that way, how the eighth grade student can grasp that, can take this knowledge, can easily understand, like that. For that, you need to go very specific and very clear. That's why you need to learn this prompting very well while before using the AI tools because you can lose the best and good creativity from the AI as well. You can see here. This is a difference between the poor prompt and effective prom. In that way, you can learn how to write the specific prompts for AI module to get the best output. I hope you understand these proms. Don't worry, we will jump into Char GPT and we'll see how these two proms will generate the output. I'm going to the Char GPT. So I already in the Cha gibt. If you don't know if you are a beginner to use hagibi just go online and search for hagibi and just log it sign up for free. After that, you will come here. Let's take the poor prompt that is tell me about just write it will generate the output about the history. What is the history, definition, and scope like that. You can see here. You get the information about history because our input is tell me about history. Simple. But if you need the specific information from the history, you need to write the specific instructions for AI. For example, you can take this prompt. I'll just copy it and I will page here. If you see, explain the causes and outcomes of World War two in simple terms suitable for eight degrees student. Let's see what AI will generate the answer. You can see the causes of World War two. It will directly generate only the causes and outcomes of World War two. Why it is going in the specific manner. Why we have instruct the AI to generate the output to generate the causes and outcomes of World War two for the Earth grade student. You see the explanation is very good. World War two happened because of several reasons that build up over time. Here are the main causes. You can read here, right? Outcomes and why it's important. This output is well suited for the eighth grade student that the eighth class student can easily understandable. Even can change here directly. Like we will see this in this format. Let's take fifth grade student. Let's see what the output is differ from the previous one. If you see here, the output is changed. Unfair treatment of January after World War I, Germany was blamed for the war and had to pay a lot of money. If you see the explanation is changed when compared to previous one. Why? Because the fifth grade student have some limited mindset understanding capability rather than student. But that the AI has know about data, how the fifth grade student will think what the understanding capability, but that the EI will change the output be half of our requirement. For that you need to go in arity. Even you can go with more specific outcomes like fifth grade students have low marks in history or. But what happens here, it will generate the output in the very simplified explanation. We can say explain simply for an learners, what causes Germany was anger like that. By knowing your requirements easily, for example, you are students by knowing your students pain, by understanding your students, you can take that requirements, just put to the AI and all the information from your side and it will generate the output based upon our requirements. But that the engineering is all about a providing your whole requirements to AI in which it can guide to AI at the maximum potential and will generate the customer output for you. You can take anything. Instead of trying by yourself, just you can use AI tool as your assistant and take the base level of information from AI and just add your creativity and go in forward. Like that. Even you can go more specificity by adding your requirements, your own requirements, it will generate the best output for your need. That is all about writing the good prompt and bad prompts. Okay, this is a put prompt and effective prompt. I hope you understand. For getting more experience on that, just go practice by yourself in the Char GPT itself, write the poor prompt and write the effective prompt with different use cases or by adding more specificity to check the output, which works for you very well. In that, you can take that prompt in everywhere as you can. I hope you understand this prompt engining well. Okay. So why is important for teachers. As I earlier discussed, customizable learning. Tyler AI responses to match for teaching style and student needs, I explained in the hajbet itself. Next one, that is save time. We are earlier discussed about learning this AI in prompt engineering is how it will save time generating lesson plans, quiz or creative content in minutes. For example, let's go to hall C. If you ask here, generate quiz. Let's take a topic. What it will happen it will generate some quizzes that you can directly ask to your students and just test your student that your lecture is good or not. You can see here. It will automatically generate the quizzes, lesson plans, documents, all explanation in your preferred requirements. Yes. Even you can generate this in your Spanish language, French language, that is all of this is AI can do anything. That's all your responsibility to write the best instructions and requirements that you need because AI can generate anything. Okay. I will generate some quizzes related to the above topic. You can ask your students or do that you need as you can. You can see some short questions, some fill in the blinds. Would you like to an answer key for this quiz as if I write a and it will generate the answers for our above quiz. That is all the how you use AI. That's why this prompt engineering is powerful till that you can use AI at the maximum potential. So this is all about our prompt engineering, basic. That is the save time in generating lesson plans, quizzes, creative content in minutes or seconds as well. Anthon enhanced engagement. As I said, if you make your teaching very fun, if you are looking to explain any subject or in case, for example, let's take in that case only. This is a good thing. If you're looking to explain this history to your students, just go and write here. Please make about topic of history into fun story. Fun story to explain my students. What happens here, I will generate the above story into some creative story that you can use this story in your teaching explanation that student have some fun and they can learn all this topic in fun way that they can memorize everywhere in anytime. You can see here. This is a great story if you think once upon a time, there was a big busy neighborhood called Arthuton. Okay. Check here the story about this. This is how you can use AI. Even you can use this EI in unlimited ways. It is all about, how you want to use AI, that is all up to you. This comes from the practicing by exploring more things from AI. I hope you understand these points very clearly. This is why it is important for teachers. Let's start the second lecture of this model that is crafting at prompts. Let's dive into them. 6. 2.2 Crafting Effective Prompts: Okay. Let's start our second lecture that is grafting effective prompts. As we earlier discussed what is a prompt engineering, and we'll explore some pod and effective prompts in changeability, right? What are crafting effective prompts are? See clarity. There are three things we need to keep in mind while writing the effective props. That is first one, it is clarity. What is a clarity? Before you approaching to AI tools, you need to know. You need to have some clarity about what you are asking to EI and what is your actual requirement. For that only, it is beneficial. So if you don't have some clear topic, clear mindset that what you need from AI, the output from AI can be very clergy or inaccurate for your requirement, but that you need to have some clarity in your mind while approaching to AI. So that be specific and avoid vague instructions. Be specific as we earlier discussed with the art of writing specific prompts. We have explored two examples that is tell me about history. That is not a specific one, but when we go in deep with specific like we have understand about here in the ajibt itself, we just write the proms specifically for fifth grade students, eighth grade students like that. In this they have some specific ness in the problem in which we can get the output for better quality. Writing instead of getting the output that is the I can have some advanced English that students cannot understand. If you put the student age or student demography to AI, the AI can generate the output based upon that particular grade students that they have the learning capability. Right like that. By going the specific, by writing the proms for specific application specific need, you can get the best quality output from ER. That's why if the crafting effective prompts, first thing is that is clarity. Be specific, avoid vautructions. Vacuum is unclear. Unclear means that doesn't have the specific goal. That is vacuum instructions. Please avoid that and be specific in the writing the prompts. That is the first one. Second thing is this context. As we earlier discussed by providing more your requirements to AI can generate the best output. What means here, providing enough background information for the AI to understand your goal. The context best example is we are already discussed about that. That is World War two in simple terms suitable for eighth grade students. That means you are providing the background information. That is here. You can see, have low marks in the history. This is a background information, it is a simple prompt that you requirements, which describe your requirements. But you are also providing your background information of your students that they have some weakness in grasping or in weakness history in point of view to learn or to get the information about the history. For that, you are providing AI to students background information that this particular students have low marks in history. So please generate the output according to them. To explain in better way, like that. You are providing here background information. This is all about context. Context means providing more background information of yourself, of your requirement that AI can understand very well of your goal to generate the best outcome for you. P that you need to provide as please try to give as much as possible background information for your requirements to the AI that AI can understand very well and go to generate the best output for you. But that you need to please keep in mind that about context. Provide enough bankrod information for the AI to understand your goal. The third point is action oriented. You can see use verbs like explain, summarize, placed or create to define the task clearly. So always use these verbs terms like explain summarize. This is a simple thing we need to add in the task or in the prompt itself. If you see here, we are using all about tell me about tell me there is a verb. Write me about history this is also one verbum. That is simple thing we need to keep in mind set. Explain means there is a verb term that we are describing the action to EI. To do the next thing to our requirement like that. That is about action oriented. This is about crafting effective proms. This is all about crafting effective proms. What are the sum? What are the things we need to keep in mind while writing the effective prompts means clarity. Be specific and avoid unclear and vague instructions. The second in this context, provide enough background information for the EI to understand your goal. That means your requirements in each and every detail and your requirements, your needs, or weak strengths or anything about that. That I can understand very well and generate the best outcome which matches your requirements and possibilities. Okay? The third one is action oriented. Please always use terms like explain, summarize, develop, write next generate all these things, verbs like action oriented words that you write that you can use in the prom to take the AI action. That is all about the crafting effective proms. Let's see some examples, basic examples for teachers, how it can be done. For the first one that is lesson planning. So you can see some basic example how we can write the prompt to generate the lesson plus for our requirements. We can take here, design a 30 minute lesson plan on the water cycle for fifth grade students, including activities and quiz. This is a great prompt. What happening here, we are just not going to design a lesson plan on the water cycle for the fifth grade students, but you are going to specific ness, design a 30 minute lesson plan. You are not telling to I to generate the lesson plan, but you are telling to generate only the 30 minute lesson plan in which you can focus on specific setting and just what you want to explain in that 30 minutes. That's the thing we need to know how to use AI in the perfect and effective manner. You can include activity squize. This is simple, prompt. We just don't have to by hot it, we don't have to keep in our mindset, but just put all your requirements in the prompt. That is enough. That is the effective prompt only. There is no another techniques, more science behind that. Prompt means writing your own instructions, writing your requirements to the AI. It is only the effective prompt, but go in the specific manner instead of writing the whole bunch of instructions like design 30 minute lesson plan for the topic watercycle which students you are going to explain fifth grade and some extracur activities that is quizzes and assignments like that. That is a lesson planning. Students feedback. You can directly ask to the student or write a positive constructive feedback for a student's essay on environmental conversation. What is here? It is something different from this. Take the students, for example, if you explain something to your students. Ask the students, ask the students how much they have understand my explanation. Take that answer and pass to the AI, and it will generate some positive feedback or according to that, in which the student is actually understand your topic or not. That is a great thing that I can help you with that. So as a teacher, you already know about how much student can understand your explanation, but by using AI, it will save you what time, Taron is creative exercise. As I said, you can generate anything from AI. It is all about your requirements. It is all about how you use AI. To help in maximum potential. So for that, you can ask anything about period exercises, generate a list of creative writing proms suitable for high school students, low school students or fifth grade, third standard, even to wealthy grade students. It is all about your requirements. You can generate more and more. Rather than this. That is all about your need and requirements. Just put your needs and requirements to the prompt in specificity and avoid unclear instructions, then you good to go. That is all about using EI tools at the basic level or ineffective prompts. These are some examples of how we can use in our EI, daily life as a teacher. 7. 2.3.1 Basic Prompt Techniques - Zero Shot Prompting: Okay. Let's start our third that is basic prom techniques. There are four different type of prom techniques. These are some basic prompting techniques in which we have to know how the actually prompt patterns will take will process it. So if you take any LLM like HGPT gm.ai Cloud, if you take any chatbard so we have to know these basic prompting techniques in which we can use in our daily work life or professional life to get the best output from it. These are the simple basic prom templates. In later lecturers, we will see some advanced prompting techniques. So we will go from basic to advance, let's start from our basic prom techniques. These are the very poor prompting techniques and these are the major role in prompt engineering as well if you are using the IITs. These prompting techniques can help you to get effect output that actually matter from AI tools. Let's drive into that first one that is zero shot prompting. So what is the meaning of zero shot prompting here? It is nothing but writing a sum asking a question to AI. It is a simple any question. Zero shot prompting means is a simple technique in which we will use any quotients or query. We will ask to AI to get the best output from that. We can see the example here. Explain the Pythagoras theorem to a tenth grade math class. There is no nothing in this reasoning prompt or anything specifically. It is just a specific qui that means the zero shot prompting means it is a simple question or query that you are directly asking to the to get the output. It is simple. There is no reasoning in that. We are not providing any context here, any background information or any examples. It is a simple writing a question here. Even you can see, explain the Pythogor serum with tentigrade math class. It is a simple question. There is no in any prior contexts we have given background information, as we earlier discussed in previous lecture light, you can see here. Zero shot prompting is when you give the AI a task without any examples or prior contexts, that is simple. It is example. Explain the Pythagorum to a tenth grade math class. It is a simple question. This method works best for straight forward or well defined task. It is simple. Let's see we'll take this prompt and we'll try it in arbit. If it is simple question, I already use, let's it will explain the pyograsymt then the great math class. It is a simple question that we ask to AI. I am taking the hagibi here. You can take gm.ai, anything that if you think that is better. But all the prom techniques we will explore in the chargb because hagibi have great capabilities rather than other tools. We will explore that also in later classes. Let's see our second prompting that is few short prompting. Let's see the second one. It is a simple one, but let's see the second one. 8. 2.3.2 Few-Shot Prompting: So second one is few shot prompting. What is the meaning of few shot prompting here? It is a opposite to the zero shot prompting. It is simple. What is a few shot prompting? Few shot prompting provides the AI with a few examples to guide its response. What happening here? You are providing the example, in prompt itself, you are providing the examples in prompt itself to guide its response. That means you are your requirement is the output should be in this format only. In that case, what you will do, you will provide some example of that format in prompt itself. The I will learn from it and it will generate the output based on the example that you have given in the prompt. That is simple. We will explore that in the chargeability directly with the simple example. After that, we will take this as a teacher prompt. Let's go to the ha chit and we will start here. What I'm telling you, let's take the example student and teacher. What I'll write allow the below. Structure to complete task. What I am telling to here? Directly, I will provide examples, let's take two conversations like teacher and student. Teacher. Hello students. Let's see. Let's take student. Let's take good morning, ma'am. Let's take another example. That is what we are happening here. We will telling to AI follow this structure to complete the task. Let's see example teacher. Teacher will say again, good morning today. Again, the student will say, How are you ma'am? I'm guided the response. If I ask a question, simple, that is what a teacher should need to response for this, how are you ma'am question? If I just keep it here. I will learn about structured here already. You can see here. So let's see what is the output should be. Yes, you can see the output here. Teacher. Hello students, student, good morning, ma'am. Teacher. Good morning, students. Students, how are you, ma'am? The teacher is automatically generated the response. I'm doing well. Thank you. How about you, all of you? Are you ready to learn something new today? That means what happening here. Im guided the AI. Im guided the AI to generate the output regarding my I provided structure. That is simple. So what happening here? I am provided the example here. What is the example here? Teacher students structure. Conversation between teacher and student is example I provided in the prompt itself. In the prompt itself to guide the AI's response like that. The AIs learn from this example and it will generate them. Answer this. Why the AI learn how the teachers will respond and what the question that students are asked and what is this predicted one? The automatically AI will do. If I tell, let's student. If I directly ask PTR, I will blank it. Let's see what the answer should be. You can see here, students, ma'am, we are excited to learn. So the AIs learn with my examples here. Rather than just providing all those things, I just use it. I just guided the AIs response, how the response should be, how the response in the format. This is the best example. Why I am provided the examples, the I learned from these examples and generate the output based on these examples. It will continuous conversation. If I just again click if I write teacher. It is ultimately generate the teacher response here. Next response, that is, that's great to hear. Today we are going to learn pa Critiam Are you ready to explore this amazing concept? That is the teacher have telling. This is a conservation now AI is learned from my examples. I hope you understand. Even you can take like this, I'll just go here and I will so for example, if I write anything, you can add can provide any example that you want, how you response you need. For example, we'll take whenever, let's take another example. I ask question to you, please follow structure. What happens here? Whenever I ask a question to you, please follow below stucture. What is the structure? Instead of giving the answer to my directly for that quotien write the quien. Let but I am the link to this. Write quotien Let's say question. After that, your response. Let's see what the output should be. Okay. Got it. Follow the structure from now on. What happening here? When I tell example I provided. Instead of giving the example directly, but I provide the structure to follow this whenever I ask a question. Now if you see if I ask any to this charge GPT, let's say write about write anything that's angles. What happens here? Let's see. You can see here. The AI has learned my structure because I provided the example or structure that I want, the format I want. Instead of getting the response, I write the question, what are angles. After that, place your response should be here. Okay, that is the thing that you need to learn this few shot prompting because it is a very effective way to guide the AIS response that you need. How you need the AIS output in that format, you can use, you can use this prompting technique, which will literally can change your experience. So this is not limited to, you can use anything. You write any examples you can provide. You can easily ask any task okay. You have to provide the example or format that you need in which format you need, in which language you need the response. But that just provide the response and provide the examples in the prompt itself and get the response that you need. That is simple, and it is a very simple and effective prompting technique. I hope you understand. Let's see our main example, how as a teacher, we can use this. This is a prompt example. Here let's take this let's copy this prompt here. Let's see what the output should be. Now here, I'm providing this prong. If you see the prompt here. As a teacher, I am explaining the grammar to students. Here I am telling to AI. As a teacher, I am approaching to AI. Here, here, how I usually explain grammar. Define the concept, give examples, ask a practice qui. Now explain the difference between there and they are using this method. I'm expanding the student to grammar. My steps, as a teacher, I provided the steps to the AI from know what happens here whenever I ask a question, you can see here, difference between this. This question The AI will response, how the air response will look like. It will first conceptually define and it will give the examples and it will ask some practice question. Let's see the output here. Why the AI learned from my prompt itself. You can see here. It will follow in. It is followed my instructions clearly in this prompt, why I am provided the format or example here to guide AIs response. You can see here. The AI is generated the output that I need, that how I need. You can see here. It is followed my question. Okay. So it first define the concept, right? It will just give you the examples, right? And it will just ask some practice questions that is simple. That how you can instead of writing the jumping into writing the prompt and instructions, just put your all requirements. Don't let miss you or any requirement because EI help at potential level if you know how to use in effective manner. Okay, I hope you understand this. Even you can use more way, that is up to you because practice makes man perfect. As a teacher, you know, for that, you need to explore more examples to get this most of you. I hope you understand. This is about two short prompting. Let's see. When this technique is great. You can see this technique is great, but generating outputs that align closely with your teaching style. So this is the course is prompt engineering especially for teachers, right? So as a teacher, so you can use this technique that helps you to generate the outputs that align closely with your teaching style. Just as a teacher. What you know, how you teach, put all the teaching your style to HGPT or AII tools to learn from you to learn your teaching style and it will generate the output that how you like. But that you can use this few shot prompting that will help better you to focus on creativity and to make something impact the student's life. I hope you understand this few shot prompting. Let's jump into a T that is role playing prompting technique. 9. 2.3.3 Role-Playing Prompting: Okay. So in this role playing prompting technique, role playing means, it is we are using some most powerful and popular technique that is role play. Role playing means provide assigning a specific role to AI to generate the output in that specific manner instead of putting all together. Okay. So as I said, the LLMs are designed or developed by large amount of datasets in which they can generate some inaccuracies in the response. So to avoid that mistake, we are using role playing technique. So what it helps to AI, the I will think in the specific manner. So in that specific manner means the AI will generate the output in that specific manner in which we can export the output accurately because the I cannot focus on all those things. Okay? The II focus only as a specific application. Why in the prompt you are provided the role playing like you will say the example in a few seconds, but understand this most concept. Role playing helps the AI adopt a person of a specific task, it is a bestefor specific task. Even for every prompt, you need to use this role playing prompting. Why? To get the output from AI in accurate manner or in effective manner, you need to use this role playing technique. Why? I will think. Instead of putting all together, it will focus on only one specific task. As a specific role, we have ascend to AI in which the I will think in the deep of that particular task or specific assigning role in which we can get the more effective output rather than just getting the information all together and putting into one set. As you know, as a teacher, you know, because you have some specific knowledge subject rather than the person who know all the subjects. Okay. So any company or any school or anything you can take, the subject matters experts are only one subject or two subjects, right? So the most of the people, for example, will take, for example, if the person if the patient have some heart problem, right? If that heart problem, there are two doctors they can go, that particular patient can go. So the patient no, the problem is the hot Okay. So for that, they will go to heart specialist or other doctor, which we say all the body parts. I hope you understand this example. Because these patients will only go to the specific heart specialist. Why? Because the specialist have the greater and accurate solution for that heart patient problem. This person will trust the heart specialist only because they have the master in the heart surgery or that particular doctor is for that. If you use this role playing technique, it will act as a subject matter expert in that particular persona assigning role. So don't confuse it. It is simple role playing prompting technique. In which we have assigning role to AI to do some particular task. It is a most popular and very effective prompting technique. Let's jump into a GPT, and we will see how it will works. For that, you can see the example here. You are a science tutor. What I am doing here, you are a science tutor. We have assigned a role to AI. What is that science tutor? Specific one. Instead of you are a tutor for middle school students, explain the process of photosynthesis engaging and simple way. It is a good prompt, but there is no specificity in that prompt. Even the AI can generate the best output, but is not much accurate or effective when before using this assigning role. If you are a science tutor, if I place this science tutor instead of just putting tutor. If I place this particular assigning role to AI, so it will it will think like I am a science tutor. No, I am generating or I am explaining photosynthesis to middle school students. I will think as a science tutor. Now, it will just generate the output by how the science tutor think why it will just avoid all other information. Instead of science. No, it will focus only in the science era. Instead of going to the physics, instead of going to the mathematics. It will only focus on science in which we can expect the greater and effective output. I hope you understand this example. Let's dive into our hatch bit and see how this output and how this role playing will help to interact the AI in maximum potential. That's true. I will directly copy this prompt here. We will jump to here. Now, instead of if I write the question here, that is a great chagbt capabilities, it will learn from previous output or instructions. If I place a question, it will follow this format to break this format, just write from now onwards. Or you can use forgatab If you can use anything, forgot about or from now, the two places may same thing. Let's see. I will ask you a question. That is simple. From now, you are a science tutor per middle class middle school students, explain the process of photosynthesis in engaging and simple way. From no. Instead of that, I will just try forgot about. I'm breaking the chain here. Forgot about. What happening here? The I will forgot all these structures or any instructions that I given in previous way. If you use this forgot about from now onwards, now it will come this and this instructions. It will only follow these instructions. Let's see if it will works or not. But you can see here. I will again following the previous structure. For that, what you can do means just click here, forget abo and do not follow above. You can write like this only. Do not follow above structure or previous structure. Let's see what happens That is, you can see here, the output is. It is all about how you are using AI. This is a interaction between you and AI. This skill can be developed by yourself only. Just know the prompting techniques how to use, when to use and how to write it. After that, you can just write interact with I that is enough for you. As you can take the great output. Anyone can interact with AI. But if you know how to use in effective manner, you will get the output. Which more matters than using the people without knowing the prompting techniques. Let's see, you can see here. This is a great output. Why? I have assigned your role to AI. You are a sign tutor for middle school students, explain the process of photosynthesis. I explain the photosynthesis. If you can see here, the output is so much effective. Photosynthesis how blanc like, how does it work. The recipe for photosynthesis, why photosynthesis is important. You can see this all output here. Okay. That is you are science tutor. That is a specific assigning role. Even you can do anything, you can assign a specific role if you want to learn mathematics. Just go. You can use these two structure. You are a science teacher or you can take at assay. You can take this also. That is all up to you. At essay, experience, let's take experience experience. Let's take physics teacher. Physics teacher. Your task is to explain. Your task is to explain. Let's take anything that is letter to physics. You can take bow rainbow colors, web Jo let's take your task is to explain web job. Even the AI is no, don't we will see with this example here. Act as experience because what I am assigned a role to IIs experience physics teacher. Even you can take that act as experienced physics teacher in which you have ten years of experience in the web JR or rainbow color explanation. You can go specific in specific as much you can. Why? You can get the best output. That is simple. Let's see what the output should be. So you can see what is j? Bj is acronym that help us memory the seven colors of the rainbow in order. This is good, violet, indigo, blue, green, yellow, orange, red. How does web jar form? Why are colors in this order, application of web jar, fun fact. This is simple, right? That is output. You can take you can go in specific specific in particular topic in which you can get the great insights from AI in that particular specific topic. For example, you can take act as experienced physics teacher and your task is to explain web Jar. You can take like that. Act as the experienced physics teacher who have and you have ten years of experience in explaining web jar in detailed manner. Just in specific topic, you have ten years of experience, five years of experience. In the specific topic, even you can go another specific topic, you can just take that and it will think as say, I have got ten years of experience in that particular topic. I need to explain this and this in deep manner. The I think in deep manner, it will give the best output rather than just asking the question, explain the Vj. I will just throw the output, which is matters, it is related to the VDJR. If you go in specific manner by assigning a role by sharing some experience in that for the specific topic. You can get the best insight from EI. Why the I is no The I can generate the best output. Why it is primed by large amount of data sets. It have more information rather than as and rather than that specific test book have. But that you need to learn this act as a assigning role that is role playing technique. I hope you understand. There are more examples we'll explore with advanced prompt techniques in later classes. Just know these prompting techniques as a basic thing. I hope you understand these role playing techniques. So this is all about the role tying prompting. You can see when it is best for use. This approach makes the interaction more engaging for your students. Why engaging? You will go in deeper and deeper in that you can engage for your students. For example, I forgot to mention you. Sorry for that. You can go as a storytelling, instead of, for example, let's say, as essay. Act as a fun storyteller. You have ten years of experience in storytelling. In story telling. What happens here? I think I am a fun storyteller and I have experienced ten years of experience in storytelling. That's good. Not what is my task? No, you task to explain about topic that is withdraw. Instead of I'll just take about topic. AI have some patterns that it will generate the next output based on the previous output. That's a great capability that Cha chibD have. You can see here. Your task is explain about topic to explain, let's forget this to explain about topic in the form of M story to engage my students to engage my students. Let's see what the output will see. So you can see here. It will generate the output in the story format. You can use this story to include in teaching field. Make your student engage, not only they will learn the topic, the student will focus on your teaching. That is how you can use this EI at maximum potential. You can check this output here once upon a time inland not too far from here. There was a magical light named Sunbeam. This story have which, that means this all topic jaw is explained in the format of story even you can use in your teaching to understand or to make your students engage. That's why you can use this apple. This is simple question I have written. You can go in deeper and deeper with that. That is how much exciting to use this role playing technique. This is more exciting prompting technique here. You can use in your work life and see how output and how the things can go easily or more creatively. That is all about role playing prompting here. Let's say our fourth one that is fourth basic prom technique that is last one. That is system instruction prompting. 10. 2.3.4. System Instruction Prompting: What is the system instruction prompting here. As the name suggests here. System means you can imagine a computer system. Computer system is designed or developed for various tasks. As I said, prompt engineering means, what is the prompt engineering? Thating the prompts for specific application is called prompt engineering. For what we can do, we are making the systems for specific application here, right? So for example, imagine some computer system. The computer is designed for various tasks, right? So in that, the computer system basic level is prorating language, program language or anything like that. Okay. The AI now come back to here system. The computer system is a system prompt here. Just imagine that. Now, when you go to any computer system, if you type any in keyboard, that you are giving the input there computer system. Here, that is a task prompt. I hope you understand. What is the system instruction that is a computer system? What is the task prompt that you open the computer system and you are typing something or you are giving the input something to computer system to do the particular task to complete the particular task. The input that you are giving is the task prompt here. The system instruction means that is already a developed or computer system that is developed for a specific application here. That is simple. This is understand by using some examples. Let's dive into that. You can see here. What is the system instruction mean? System instruction prompting involves the setting the personality or role of the AI at the start of your interaction. So what is the difference between role playing and system instruction prompt here? If you see the personality or role of the at the start of interaction, what is the difference between the role playing and the system instruction? As I said, role playing plays some good thing. But the system instruction is like that. For example, some computer system. Now your prompt is, for example, let's example, let's let's jump with the char GPT. As previous one, we will use this act as a role playing prompt technique here. If you are directly using here. In this CarGPT, you can use as many as assigning role prompting techniques. You can take another act as a let's take any storyteller and you have five years of experience in that particular topic and you need to generate the story for that. You can use lighter. You can take number of them. But what happens in the system instruction is what happens here? You develop or you make AI as a system to maintain this type of specific teacher only. I hope you understand. That means that is one particular computer system here. Let's see the example. I will copy this. And I will say no AI, from now onwards now onwards. To break the abo I will just page here. What I'm telling you to AI is from now onwards, you are an educational assistant for KL teachers. You can take students, all these things. Let's take students. Let's see. Focusing on creating engaging content and providing student feedback. Use a professional at approachable to. Let's take teacher as it will be better for us. What happening there? From now onwards, I will think, I am educational assistant for K two teachers. No, I have to focus on creating engaging content and providing the student feedback. What I have to tell, I have to professionally approachable tone. Thinking here now.I will generate some output. Now, think of it. Let's see. The output should be I will tell, stood. Understood. I will ensure that my responses are professional approachable, focusing and creating engaging content and providing helpful student feedback. Let me know how can assist with your teaching needs. If you see here, what is the difference between role playing and system interaction form? It is from now onwards, the I will focus from these instructions instead of previous one. I hope you understand. You can see here. From no onwards, you are an educational assistant. Now it is an system instruction. It will only focus on this task. Okay. It will only create the engaging content and provide student feedback only. It will never generate the story. Why I am not mentioned here in the system instruction prompt? Why if I ask any question, it will only go these instructions and it will generate the response according to this main instruction. Why it is designed only to do this particular task only, not the another one. Let's see how it will works if I tell you AI No, generating. Even if you to generate a story, it will generate the story. But why we have not told to EI not to do this for that, for example, you can see, let's say, we will see the example first now. Understood I will ensure my responses are professional teratable. Look, I will tell to EI, create content. Let's take some particular topic that is photosynthesis. Let's say photosynthesis. It will generate the output here. Why system instruction prompt is this only. I hope you understand. This is it will works like a computer system, and this is a prompt that you are giving that means you are giving some input to the computer system to do particular task and the task is done. Why the output is here. That is good. What is the difference between that role playing and the system instruction prompt? These are two, but why? Now, come here and we will see how it works. If I tell AI, generate a story. It will generate the story because it is also one type of content, but we have to take, which is not related to this. What is that? You are an educational assistant. Now if I tell AI, that is, instead of writing the content, I will tell to AI, um, there's much of educational tone. Instead of that, we will directly add some instruction here. You task only this. You don't you don't need to avoid task to follow the instructions only and avoid generating stor it. This is some particular. Let's see I'm mistaken in something. Let's clear that after the mid. See, I have just told to AI, this is system instruction prompt. First in the first method, I have just to tell to AI. In the first method, I've just told to AI to generate a content because you are a education assistant for Katar teachers, focusing on creating engaging content and providing feedback. Now second thing, what I have told to AI, I tell to AI, do that particular this task only and avoid generating the story. Now the AI learn my instructions. This is the system instruction prompt here. It is now working as a computer system. Then I will give the input here. I will give the prompt. That is one sub prompt. What is that? Just write a content for photosynthesis, let's say, same question. Now I have given the input. Now the task is done. This answer response is jumped here and it will see, Okay, I have to create the engaging content. Now it done. If I tell AI now here, generate a story. You can see the output here. See, that is the major benefit of this instruction prompt. You can see here, I can only create stories. I can only create stories if it aligns with the goal of engaging students in a creative educational context. Since your focus is on educational content, how would I like to proceed? Would you like me to explore a specific concept or topic today story? So it is refused to generate a story. Why? Because I have told AI to avoid generating story. That's why it is not generate the story after my question also. That's why the system instruction prompt is very benefit in this type format. Even you can go specifically, let's take another example. So let's go instead of going educational content, let's take science. So now you will understand clearly. I will take out this. Let's say the AI generate the AI thinking instead of educational assistant, I just assign a rule that is science assistant only. It will never go that is mathematics English teacher. It will only the science assistant for teachers. Now the AI is thinking is science assistant. So if you see, I will ask a question to particular about the signs. Write content for Let's take any signs of particular thing. Let's say example, uh, let's take directly write content for science history. Okay. What if we generate responses. That is history of science, Asian science, science revelation, all those stuff. Now if you ask a question to AI, write content for mathematics. Let's see the example what it will be write content for mathematics history. Let's see the output will be the expected one or anything like that. The output is there. Okay. So what it will should happen here, the I is now times assistant, but the mathematics is also coming. Why? In the system prompt instruction, I never tell I to do only this thing and avoid generating content for other subjects. That's why this following the structure. If I give negative prom here or negative instruction to particular task, do this and never do these things, the I will know our requirements even more effective, for example, let's say. Now avoid generating content. Generating content for other subjects. That is simple. Let's see how it will be the output. Instead of writing the content for science, I will just tell AI to generate a content for mathematics again. Write content. Mathematics. Let's see the output. What is output. You can see the output here. I can help with mathematics related content, but since you asked for a focus on science, not a mathematics, could you let me know if you need support integrating with a science topic, otherwise, I can help with engaging size content instead. Let me know how would I like to proceed. That is the benefit of using this role playing technique. So this prompt can make you to build some specific application for the specific one, instead of getting the code, writing the code. This simple prompt can help you to make some amazing apps but to do some particular specific task. It will working like a computer system. No, it will refusing to generate content for mathematics. Why? I am told to AI, avoid generating content for other subjects. So instead of that, you have to focus on the science subject. That is the instruction prompt system instruction. It is a fixed one here. It is a system. Now if you give the prompt here, it is a sub prompt question related to for that it will generate the response. That is input. It is a response or task done like computer systems have. It is a computer system, fixed computer system, like that. This is how you can use this system instruction prompt to build a specific application and I will only do that particular instructions instead of going together together. It will only focus on this and it will work only in that specific manner. I will never go outside of that loop. Okay. And remember, adding this avoid these extra instructions, the AI will get more requirements of our needs, right? So even it will learn our instructions and it will only a follow our instructions only for that you need to give as much as possible. You need to give more detailed instructions without missing anything. Then only they can follow instructions and it will generate the best output for this. That is a port system instruction prompting technique. I hope you understand there are more things to explain you, but this skill can be developed by yourself when you need to use this prompting technique and you have to explore more examples, then only you can get the knowledge of how to use this AI as effectively. You can see here. This is example, design a 20 minute lesson plan about the solar system for fifth graders including one hands activity. This is a task prompt. This is a system instruction prompt. It is a fixed one like computer systems have. It is a task prom that you give to input to computer that is opening a file, right, any opening closing, opening a chrome like that. That is task prom and once you open a chrome, the task is done, the chrome will open. That is a generating response from the AI like that. That is a simple example, you can imagine that for better understanding. System instruction and task prompt. This task prompt is only followed by this system instruction. If your task prompt is not matches the main system instruction prom, then D will simply refuse to answer that. As example we have earlier seen. When we have to use this, when the best time to use this prompting technique is multi step task, long term projects where consistent behavior is key. Okay. Multi step task are long term projects where consistent Ba is key. That is most important. As we said, if you have some more task, instead of writing act as the role playing technique for every time you need. That means what here. Instead of writing the act as a one story teller for each prompt. If you are looking that, just try and AI. Just write the prompt for one time using system instruction prompt technique and after that, just write the task and it will done in the loop of this system instruction prompt only instead of writing the act as a role playing plating for every time that you write into the prompt here. Instead of writing every time, just write in one time, that is a system instruction prompt and just write the task and is done. That is all about these basic prompt techniques. As we're seen, this is all about system instruction prompting techniques, basic things, and there are other advance that will literally make you amaze to get the output from the AI. We will see in letter class. Let's see our last topic about common mistakes to avoid. 11. 2.4 Common Mistakes to Avoid: Mistakes to avoid common mistakes, what is being too vacuum. Instead of writing some broad or broad area, focusing in broad area, you can go specific. You can see here, help me with math. I will help me with math. What math? In which specific area I need? If you go the specific manner, the AI will think in specific manner in which you can expect the bust output as a Liar learned, right? A wide use this, create a phi algebra word problem suitable for eight grades with solutions. You are given you are you are used in the specific manner Pi algebra, specific one, eighth grader. Instead of just world problems with solutions. You have again, given specific one. That is eighth graders with solutions, how much you can give the specific requirements of you, it will generate the best output. That is a Bing to g. That is Haw help me with broad narrow or broad concept. Just go in the specific manner. Just write the prompt, what you need from AI in specific manner. That is effective prompting is. That is a being to argue and second one is overloading the prompt. Now, being to argu is right. But what is overloading the prompt? I have told EI I already told you, giving the background information or putting all requirements will help you to get the most output. But what is overloading the prom? You can see the points. Avoid long complex proms with too many tasks, break into smaller manageable parts. That's what I have told AI to use a prompting techniques here. Go step by step, instead of writing the whole prompt at one time. I'm not telling about the system instruction prompt, but I'm telling the specific individual prompts. Fool loading prompt means just I'm telling you about with specific one task is okay. You can write so much long prompt for specific task. But if you are looking to write more prompt for one time for so many tasks, different tasks, the output can be the very okay. Instead of generating the good output for specific one, it will generate the output for each every task, which is too low or which can mistake it or which can miss the main part. I hope you understand point, right? For that, to avoid that, break into smaller and manageable parts. Break into each task in each one prompt, then you can expect more output, more output. Instead of that, if you put all the task in one prompt and generate for response for that, it will generate the response for all tasks, but something small, small. In which you can miss the main information or more information. If you use the specific one, it will generate the best and long output for that specific task in which you can explore more things in that particular task. How you use these AI models at the most of the things by writing the prompts. A wide long complex proms with too many tasks, break it into smaller or each write the prom for each task only instead of writing too many different tasks in one prompt, you will miss the more information in the response. That is overloading the prompt here and ignoring the context. That is more important. You never ignore the context. Context means providing the background information, the how output you need. I have already told you in the letter classes, that is crafting AA two proms. In this just lick include details like grade level, subject, and tone that it will even more guide the AI to generate based around grade level based on the subject and in the form of tone. Providing tone, it will help you to get the best output. Don't ine the context here. Context is very important in writing the proms here. That's why we have creating that. What is the context means Phi algebra, the specific one suitable for eighth graders, is it context? It is a context here. Subject is algebra problems and tone. Even you can add here in professional tone or in engaging tone or in funny tone that you can use as you need. That is common mistakes to avoid while writing the prompts as a basic. If you understand up to this clearly, then you are welcome to go more advanced prompting techniques in which we will see explore some more advance in which it can change your way of teaching or change your way of using AI tools at maximum potential. So let's see some wrap up play. Up to this, you know, uh, you have some solid understanding of what the prompt engineering is and what is important for teachers, right? Even if you learn how to cross the prompts using basic techniques that zero shot few short prompting, role playing technique and we have explored the system instruction prom technique as, right? So that is all wrap up and in the summary, the clearer and more specific your proms is, the better the rare responses. You need to practice these techniques, with more examples by yourself. By approaching to AI tos, you can learn more things. Okay. So up to no, if you know this, if you have this particular knowledge, how this basic prompting works, how to craft the proms and how to avoid the mistakes. If you master these three type of these topics in this model, the upcoming techniques or advanced techniques are very easy. And it will take your teaching level with A to the next level. I'm very excited to share that particular important and very effective prompting techniques in which you can build some more apps by yourself without knowing the knowledge, without knowing the coding language jars. The AI is very excited technology I use in the right way. Let's explore our third module in which we will see the advanced prompting techniques. Let's dive into that. 12. 3.1 Advanced Prompt Patterns : Ask for Input Pattern: Start our model number three that is advanced prompt engineering. In this model number three, we are going to see some advanced prom patterns which are very powerful and very effective in using as a teacher. If you use AI model. These are the fundamental prom patterns. There is no restrictions for this use in any model. This the prom patterns will work in any AI module you want to use. You can take any TAGPTCloud gem.ai, perplexity, Microsoft Coe Pilot, or newly AI tool that is deep seek by China. You can take any LLM. These fundamental prom patterns are very important. They're all same because these are the LLM trend prom patterns. This work same in all the LLMs, okay? So there are two parts of this model that is Part one and Part two. We are divided the five prom patterns for part one and most powerful prom patterns for another part two. In this part one, we are going to see some of the most important and very helpful prom patterns that is ask for input prom pattern, persona, potion refinement, cognitive verifier, out an expansion prom pattern. Let's dive into each one in details. Let's start first one that is ask for input prom pattern. So what is the actual meaning of these prom pattern is? It is a simple one. To use this pattern, you prop should make the following fundamental contextual statements. So you can see the statement here. Ask me for input X. X means it can be a task, any requirement that you are asking to AI. That is simple. You can see here. You will need to replace X with an input such as goal task. That is up to you. What is the main purpose of this prombtans asking AI to provide input to you. That means you are telling to AI, to ask input. To you. When you give the input, then it will start proceeding the task. That is simple. How you will give some input to any other machines like computer it can take example. When you give the task, for example, you can take the best example is when you are going to login in any website, it will ask your email address and your password. It will ask you input to give it. Then you can enter email address and password. Then it will proceed the task and it will successfully log up log in. For that, the EI by using this prom pattern, you are telling to A to ask input to you in which after when you give the input further, it will start proceeding that task. For more deeper explanation, let's dive into Cha GPT, and we will see how it will works. Let's go with that. For most, let's let's see the example here directly here. After that, we will jump into hagiPT. What happening here? You can see the best prompt example here. From now on, I will provide a scientific topic. And the grade level of my students. You will suggest an age appropriate experiment, including materials, procedure, and safety tips and with questions to discuss with students. Now you can see here. Ask me for the scientific topic and grade level. So at the last of this prompt, you will telling to AI, you are telling to AI, ask me for the scientific topic and grade level. These are the input. When you give the input, it will start proceeding this task. Okay? That is simple of this prompt. You need to just specify the task here. After that, tell to AI at the last, ask me for the input. Okay, that is simple. It can be anything. Not only this topic grade level, it can be anything that is up to you, that is all your requirements. Let's stay into the chargebyH prompt will work. As earlier, let's start. I already copied that prompt, I will directly placed here. You can see here from now on. By using this from now on, as we earlier discussed in the last lecture, so it will break the above chain sometimes, sometimes make mistakes, you have to give them more instructions here. Let's focus on this from now. From now on, I will provide a scientific topic C. You are telling to AI, I will provide a scientific topic here. You need to provide the clear instructions. If you are not give this at the start point or in the middle, it will take the own topic. It will take AI's own topic and it will generate the content. To avoid this, what you have to tell, you need to mention I will provide a scientific topic, and the grade level off my students, you are telling to AI, I will provide. You will suggest an age appropriate experiment including materials, procedures, and safety tips, and with the questions to discuss with students. You are telling to AI last, ask me for the scientific topic and grade level. What happens here? You will suggest and you will try and AI to I will provide a topic and grade level. No AI will think, a user will provide a scientific topic and grade level to me. Now, I have to suggest an age or profit experiment including materials, procedure and sapetives, all those things. Now, I need to ask a specific topic and grade level to the user. That AI is thinking. Now what happens here? Let's go. Let's see what Charge BT will do. We can see here. Got it. Please provide the scientific topic and the grade level of your students and I will suggest an engaging experiment with all the necessary details. You can see here. It will asking the input to you. After when you provide the answer here that is input to the AI, it will automatically suggest an engaging experiment and generate a output according to your needs. Let's say let's take any scientific topic here, photosynthesis. Grade level should be. Let's take eight standard or eight grade. Let's see what happens here. So I provided the input. Then you can see all those things here. It will generate the best output for us. Experiment gravel eight object to demonstrate that plants produce oxygen during photosynthesis. You can see the materials process observation safety tips, discussion quoi and it will just end up with some small suggestion. That is how you can use this prompt pattern. Even more, you can try it by different examples. There is no limitation of writing the prompt. You can write the prompt in number of waves, number of wives. This can be done by practicing yourself. Just put your requirements all together and tell to AI, I will provide the input or the last. You can use any type of in any way of this prompt. That is up to you and how you use this prom pattern is main important. Just know, you have to declare two things here in the prompt. You have to tell to AI, I will provide some topic and grade level. I'm just taking the example here, specific application. You can take anything here. You need to mention clearly to AI, I will provide. After that, at the last prompt, you need to tell to AI now ask me what I have mentioned to you, that means I'm telling to AI. Ask me for the scentfiTpic and grade level. Then it will ask the topic that is input. When I give the input to the AI, it will start proceeding generating the content for me. This is not a limitation too. You can use in any way in any application. You can just try out with different type of use cases in your teaching field, so you can get the most of the experiment or experience in using this prom pattern in effective manner. I hope you understand this prom pattern very well. You can take any examples, you can take all those things. Even C, even you can go to gem dot a cloud.ai, any other AI models, deep Sk, we can start using this prom pattern because these are fundamental prom patterns. These are the LLM prom patterns. Chargibty is not only the LLM we have, we have different LLMs like Cloud Gem AI perplexity, Microsoft Copalt Lama, and new AI tool that is deep seek by China. These are all the LLMs. We are discussing the LLM prompt engineering here. The prom patterns which are fundamentally designed for LLM, not only for the ha GBT. I hope you understand. I'm using ha GPT f because it has some capabilities. We will cover all those things in later classes. That is one of the best capabilities, memory update. It has great capability. That's why I love using ha GP because it has some great capabilities. I will explain all those things in later classes. For that, what I'm telling you is use this prom pattern. It will help you for better formation, better formatting your interaction with AI. Okay. Those things. That is all about. My preference all other things I'm telling to you is experiment with other examples as well with other EI modules by using the same prom pattern, you will get the idea which one is better. This is all about this prom pattern. Even you can explore more examples on it, just tell to EI and do the things as it as possible. Okay. Let's dive into our second prom pattern. That is persona prom patterns. Let's dive into that. 13. 3.2 Persona Prompt Pattern: Okay. Let's start our second prom pattern that is persona prom pattern. So as we earlier discussed some basic prom patterns like role playing system instructions, right? So this personal prom pattern works like role playing. But this prom pattern have some specific applications, right? So this prom pattern is similar to the basic prom pattern that is system instruction or role playing that we are earlier discussed in the previous classes, right? So in this class, we are going to see the depth of personal prom pattern. What is here? As I said, persona means assigning specific role to AI, specific role. This is also called some popular prom pattern. Most of the officials or any companies will use this to get you can see the exam or you can see the template here, act as a persona or role. What you are going to do here, you are assigning specific role to AI to complete the specific task in which the AI will think in that pattern only. You can expect the great output rather than just putting to do some task without assigning role. It will just thrown some output, which is not a fact when compared to using this personal prom pattern. We will compare all those things in charge B, focus on here. Act as a person or role, you need to assign a specific role here. After that, your task is to do specific task. We are telling to AI, your task is to complete the specific task here. It can be anything a gold task or any that you are looking to do with the AI, and use specific tone or style. Even, you can tell to AI, kind words, use kind words or use professional words and focus on a beautiful tone or effective style like that. That is all about your requirements, you are designing EI for a specific application here by writing the prompt. You are assigning the role. After that, you are telling the specific task, after that, how you need to generate the output in tone, in which style you are telling to AI here, in your responses and here's the input. You are giving input in prompt itself here. So you can see it is covered two prom patterns here. Two prom patterns, that is advanced prom patterns. What is one? First one is persona prom pattern and second one is ask me for input prom pattern. Previously, we discussed write, that is. So here is the input, user input or scenario. You can see the example here, which matches the above template. You can see act as a high school math teacher. I am assigning the role to AI. You are a high school math teacher. I I'm telling to AI, your task is to do that is explain Pythogra theorem to a 15-year-old student. What's happening here? We already discussed this type of example in previous classes already, but focus on the fundamental theme. Fundamental thing here. I am going specifics in specific, in which I can expect the most effective and accurate response related to my question or query. Or requirement. First time assign the role to AI, you are a high school math teacher and you need to explain pyedagory soem to a 15-year-old student. The AI will think, I'm a school teacher. No, it will forgot all other subjects or other knowledge. I will focus on only high school math teacher knowledge, how the high school math teacher will think, at the same time that I will think like that only, but not like Hume, but that based upon the training data. The actual model is trend by data. It will focus on only the math teacher, only the mathematics. Why? I am assigned the role to specifically high school math teacher. After that, it will explain the Pythagoras theorem. Instead of, if you see it can explain the Pythagor serum. If you didn't use this 15-year-old student, it will explain the pythagore serum, but in the no the format of this 15-year-old student or any other student, it will just explain the pythagore serum in which you can understand, but you are explaining the student. There is a much difference between the teacher and student understanding capability. But that as a teacher, you need to mention your student age and student capabilities and what is the student capacity to withhold or to understand this type of mathematics? This is how you can use this even you can go more in depth using this personal prom pattern. Let's jump into Char GPT and we will see how we can use this. I already copied that example. Let's see here. I have just ascended the role to act as a high school math teacher and explain the Pythagoras soem to 15-year-old student. There is no reasoning but it will automatically generate. Yeah, sorry. I will let you. We will just confirm it. So what happening here if you followed my previous lecturer that is basic prompt patterns you can understand this response here. For more, please check out again the previous lecturer that is basic prom patterns clear. Then you can understand this response. Why it comes, I focus and science content. That is very great prom pattern there. Don't miss out that. If you know about it, that is good to go. I'm breakdoing the chain here from now, forgot about From now, forgot about, bring the chain here. For that, you need to see from now, forgot about, follow these instructions. So what happening here? Now, it is It is prom pattern here. I will follow. The great thing about ha GPT is it have some memory update capability in which it will generate the output and it will follow the instructions based on previous one, but that we need to break this chain here. To break this, you need to mention in the prompt, forgot above or follow these instructions from now onwards. That is simple things you can use when the Char GPT just produce output based on the previous one, take this. Let's see here. No forgot about, follow the instructions. Now what happens here, you can see here memory updated. Now the AI will follow this prompt instructions only instead of going to back or instead of going to previous one. I hope you understand this point clearly. Now the AI is just explain the pythagoh serum in simple explanation for a 15-year-old student, you can see the content here. That is simple. When compared to this, this is some simple question we are asked to EI using the act as a high school math teacher. So what happens if you have some more requirements? Okay? So it is better. Even if you think, you can use ask me for input prom pattern, and this prom pattern combined. That makes a great prompt here. Let's see that. So for that just keep it like this one. We will start from here. So for that what a tell to AI. Now, I am using the two prom patterns here. Previous one asked me for input prom pattern and the personal prom pattern. We'll comb we will combine these two prom patterns to write some specific prom pattern here. For specific application, let's see how amazing it is. For that, I will use that is assigning role, Act essay. Let's take high school science teacher. Let's another that is act essay, high school social teacher. I am a send a roll to EI. No AI will think I am a high school social teacher. That is fine. Now I am telling to AI. Let's focus on here. I will provide. I will tell you simple, I will tell you I will tell you which topic you need to explain. I will tell you which topic. We will miss out. I will tell you which topic you need to. We have some grammatical mistakes. That is no problem. Cha GPT will automatically rectify it. I will tell you which topic you need to explain with grade level students. Now I am assigning a task to AI. What I'm telling to A, what task is to explain in simple words and fun story. Now, I will tell to AI. At the same. If you focus, if you regain that, ask me for input prompt pattern, we have to use some fundamental contextual statement at the end of this prompt. What is that now, ask me for you need to tell to AI, ask before what you have to guide the AI. That is which topic and grade level. That is simple. Now ask before topic and grade level. That is simple. If you focus on this prom pattern, so I have just assigned a role to AI that is at as a high school math teacher. It is a personal prom pattern. That is first one and I have used I will tell you which topic you need to explain with grade level students now ask me for topic and grade level. These two will combine as, ask me for input prom pattern and this one is a personal prom pattern. So there are no limitations these prom patterns have. You can use any prom pattern as you like. There is no limitation. There is no particular formula to use these prom patterns. It is all about your requirements. Just know your requirements and use AI with these prom patterns techniques because it will save your time. It will save time and it will reduce some inaccuracies in the output. If you use, not only these two, we have more prom patterns in later class we discuss. I later classes, we will see that all prom patterns you can combine, you can use for the specific application, then you can see the automation we will doing that. Okay, we are talking about automation. We are not talking about putting to create some automation by using different tools. No, no, no, not like that. We are automating the task in the har gPT or any AI module itself by writing the prompt by adding the instructions. How we need to know how to use these prompt patterns effectively in which time we need to use. Okay? In how we can use this, in prompt itself. So you need to know this. After that, there is no limitation, okay? That is all about how you put your creativity writing skill in this prom pattern. That is simple. It is all about how you, you will use. Further, you need to practice by yourself with your use cases with your examples with your requirements, then only you can improve your prompt writing skill. As a teacher, it will save you a lot of time in creating content, in creating story okay, or anything. Let's see what is the output of this here. Now, it will generate a Okay. But you can see here. Again, you have previously asked between focus on science and math. Would you like me to update your preferences to include social studies as well? This is a great thing that ha GPT have. Unless if you use other AI models, they don't remember the previous roles we have assigned to that AI. We'll talk about all those things in letter classes. Let's focus on this. Let's take. If you use the Hagibt in correct manner, so you will love this chatting with this AI. So what I'm telling you here, it will ask you previously asked you to focus on science and math. Would you like me to update your preferences, include social studies as well? Yes, include it. That is simple. Got it. Please provide the topic. No, it will asking me to provide a topic and grade level of your students and it will explain in simple and engaging story. It is how you use this. For example, I will take any social studies topic that is World War one, let's see what happens here. The story, I just forgot to mention student grade level. Let's see how the output should be here. It will generating some output, the story of World War I, the domino effect. Once upon a time Europe was a group of friends, that is the content is in the form of story in which you can tell your students in engaging way that can remember this story as the knowledge, equal to the knowledge. So you can see the example here, once upon a time. Okay. Let's see if you mention the grade level of students, the output will change according to the preferences, your preferences. Let's say take a seventh grade students. The output will be the different. Let's see that. The boulevard the domino effect of a single park, the output has changed. When compared to this, the I didn't know for whom I want to generate the story and content. It will just generate a story by some without specific ness. I will understand. But if you use your each and every preferences like sinth grade student, even more capacity, the students don't have learning capabilities in the social studies, they have boring, all those preferences, each and everything, the AI will generate the best output for your requirements. I hope you understand this how important or how it is possible or important to provide each and every requirements to AI to produce best output for your requirements. That is simple. That is how you can use AI modules at maximum potential. You can see that there is a large difference between this output and putting more requirements by yourself and this output. You can see them. Once upon once upon a time, there is a huge neighborhood called Europe. You can see this here. Okay. That is how you can use this persona prom pattern and ask me for input prom pattern combinedly. I hope you understand these two prom patterns very clearly. Okay? So please practice by yourself. These two prom patterns can help you to save a lot of time generating the content which is not have some inaccuracies, but it will help you to generate the best output. When compared to just writing the quotient, just explain the World War I. If you use these prom patterns that is act as a ask me for input prom pattern or even we have other prom patterns which are literally are more advanced and more effective or more we can generate the 50 outputs by using the upcoming prom patterns. Rather than these two prom patterns, we have other different prom patterns, which literally that will amaze your output, amaze your creativity. Let's dive into our next prom pattern that is quotienRfuirement, prom pattern. Let's dive into that. 14. 3.3 Question Refinement Pattern: Okay, let's start our most effective prom pattern that is refinement prom pattern. What this prom pattern is? In earlier prom patterns, what we have done there? We have just asked to AI directly. Instead of asking quis with AI. I hope let's dive into deeper that. What happening here? Question refinement, what we are telling to AI, refine the particular quotient. Refine the particular that I asked to you. For example, you can I am providing some content regarding this topic. Please rectify it and remove the grammatical mistakes or suggest me the better content rather than that I provided. I hope you understand this requirement. What happening here, the AI is refining the quotient or refining anything that you are looking for. So you can see here. The template of this prom pattern is whenever I ask a question. So this is not only the quotient. You can take anything. Whenever I ask a question, query, task or paragraph, anything that you requirement, whenever I ask a question, suggest a better quotien and ask me if I would like to use it instead. What's happening here? We are trying AI, we are telling to AI, so suggest me the better version of this particular paragraph. Okay? Suggest me the better answer for this question. Okay. You can tell that to AI. It will suggest a better version of your input or your paragraph or your answer that you provided to AI. Let's jump into Char GBT and we'll see how it works. We are in the chart itself, ChargB. I have already directly copied that. Now you can see here. What I'm telling to AI whenever I ask a question, so instead of question, you can take task, answer, paragraph, anything that you like. I will just go with the question here. Whenever I ask a question, suggest a better question, and ask if I would like to use instead. So now I will think, Okay, this is my task. So it will tell to you, I say, Okay, understood. I will suggest a better version of your question and ask if you would like to use it before proceeding. Let me know your next question. So it will ask to provide me question here. So I will just provide the question here. Explain. Rule of EI in education. Let's see what happens here. No. This is my question. I will generate the output. I will generate the output which is it is better than that. I will suggest a better version of this question. Let's see the output here. You can focus here. That is, you can see here. HEI is transforming education and water its benefits and challenges. We can see here. Now, AI is suggesting me a better question that you can ask to any AI model or anything like that. Would like to use this version. I tell to AIS, it will start generating the content for this question here. I'm not going forward like that. Just focus on this how a prom pattern is work. If you see explain role of an AI in education, this is my basic question. Now EI is suggested me a better version of this question, how you can use this question. For example, if you tell to AI, if you tell to AI here, whenever I ask a question, suggest a four better or three better versions of quotiens. Ask me if I would like to use it instead. What I'm telling to AI? If I ask one question, suggest me three different better quotients. Ask me if I would like to use it instead. Let's say no I will understand our task. Now, got it. I will suggest three better questions when you ask one and check if you would like to use them. Let me know if you'd like to try them out. I'm telling if I write the same question previously I ask here. I'll just provide question here. Let's see what happens here. You can see it will generating the three different better questions to me. You can see your response, A plays a transporter role in education. It is directly generating the output. We can see suggested better questions. How can I personalize education to help students learn better at their own pace? In what way does AI help teachers save time, focus more on teaching? What are the potential drawbacks of relying too much on A education, how we can understand. So if you see here, I have just told to AI explain the role of an AI in education. No, it will generate the three different suggestive questions to me. I hope you understand these points. So how effective is this? Instead of telling instead of telling to question, let's take any paragraph or any answer or that you don't have sentences that have some grammatical mistakes. So what happens here? Let's directly say using prompt here. We are learning the prompt engineering here. So here is generating all content, but I can also generate a prompt for our requirements. That's how we use this prompt here. Whenever I ask a question or whenever I ask a prompt. Let's focus here. Suggest MA. Suggest MA. Better version of, let's take better version of prompt. Okay, I hope you understand this point. Whenever I ask a prompt, suggest to me better version of my prompt and ask me if I would like to use it instead. And remember, let's see this, what happens here. Now, we are using AI to write the better prompts for I models. Let's see what happens here. Got it. I will make sure to suggest a better versions of your prompt and ask if you would like to use them instead. Let me know you are next to prompt here. Prompt means it is an equation or instruction that you need. If I provide a prompt here, let's take the previous prompt only. It is simple prompt here. The quotient is also a prompt, right? This is my simple basic prompt here. I have just provide to AI. Now you can see here. Now it will suggest to me the better version. Can you explain the role and impact of AI in modern education, focusing on how to enhance learning and teaching experience that is very effective when compared to this simple prompt here. The prompt engineering is not only that is creative that you have, you need to learn how to use AI in each and everything. Then you can maximize the potential of yourself and as well AI. It is very amazing. I have just provided some basic prompt here. Now, it will generate the best, better version of this prompt. You can see which one is effective This one. Why we lack in the information we have. We are omens. We don't have more data regarding any topic, but AI is trained by large amounts of data which have everything about that topic. The AI know how to ask a question regarding the specific topic. That's why it have taken role impact, modern education, enhanced learning and teaching experiences. When I for example, if I don't know about topic, but I want the content for that, I will just tell to AI how much I know about that topic. I will just write it simple or simple prompt. It will generate the output that is according to my prompt. But if you use this prom pattern, that is quotient refinement prom pattern or prompt refinement prom pattern, it will suggest a better version of this prompt in which the I know better the I know better at any topic. It will give the best prompt by including more requirements, which support the main topic or which supports the main requirement of yourself. I hope you understand this how much very important this prom pattern is. I'm just tell to AI. Just whenever I ask a prompt, suggest with a better version of my prompt. Now it has suggests some great prompt here. Even you can tell to AI, suggest me three better version, four better version, even you can 1010, 20, it will generate it. There is no limitation to AI because AI have some large amount of data. But as a human, we lack in the data or knowledge. That's why we will just write what we know, but AI can write which supports the main topic that is role and impact. Because I is no strand a large amount of data, they have more information about this particular topic in which they can write the detailed prompt here. That's why before going to interacting with any model, just use this prom pattern to enhance your prompt writing capability. Just come here, use this prom pattern, write your i. It will suggest a better version. Just take this and interact the AI model, then you can get the best output. When you don't have the knowledge about the topic. The question is out of your subject, out of your uh knowledge. Then you can use this prom pattern. Even you can use your in subject for better specificity in better in depth knowledge. That is all about your requirements. I hope you understand. Now, let's previously we are combined the two prom patterns, that is ask me for input prom pattern, persona. We will combine these three prom patterns. That is ask me for input prom pattern, persona, and this portion refinement prom pattern in which we will see how it will works. Let's dive into that. For that, I will just tell to AI now act as act as a expert prompt writer. I'm just telling to I you are an expert prompt writer. Even you can go for specific application, act as expert prompt writer for education. Even you can act as expert writer for particular subject, that is mathematics, you can go in depth as you can. I will just tell to AI you act as expent prompt writer. I will tell Okay. Now, I will provide let's take. I will provide prompt. I will provide basic prompt that is simple. You or you can directly take this. Let's. I will provide basic prompt. Now, just use our question definment prompt pattern here. Whenever I provide let's take whenever I provide basic prompt suggest me three better version of my prompt. That is simple. Now, you need to see the Is know what is a prompt where we are using. So for better output, you can give some extra information that is note or remember like that. Why? Let's see. Remember, this is a optional one so you can use for better effective output. Remember, these prompts are or used in AI models to get output. I'll just provide the extra information to A for better output. That is optional one. Now, see, we are used three prom patterns, but we are missing one point. That is asking me for input prom pattern, fundamental contextual statement. That is now ask me for prompt. That is simple. We have used three different prom patterns here. I have sent a specific rule to AI. I will provide a basic prompt, and this one is ask me for input prompt pattern. This one is that is quotient refinement prompt pattern. That's the output will output here. No, it will ask me for the provide to prompt. It will ask me to provide prompt. Got it, please provide your basic prompt. I will suggest three better versions for you. Let's Explain World. Let's take another thing. Explain Cold. What? You can see here. You prompt is explained cold war. Here the suggested better versions, three different. Can you provide an overview of the cold war, explain its causes major events and consequences? For the starting time, I don't know about major events cold war. I don't know. I need to include whether these points are not in the prompt. I don't know why? Because I lack the knowledge. I don't have a knowledge about this cold war, but is no. That's why it will writing in prompt, explaining its causes major events and consequences. That is and after that, you can see another better version of this prompt, Cold War shape, how did the cold war shape global politics and what are the key conflicts during this period. You can see there is three different better versions of this simple basic prompt here. This is simple one. If you write a big prompt here, even it can generate a better version of the proms even more okay. So this use cage is very useful or very important while you are looking to interact with AI. As a teacher prompt engineer, you need to know how to use AI modules at maximum potential. So for that, you need to know a particular subject or if you don't have some knowledge about the topic that you are looking to get from EI, you can use this prom pattern as well. Even if you have the knowledge, you can use this because it will improve your a prompt writing capability or prompt writing skill in which you can get the best output from AI models. That is simple, how you can use this prom pattern here. We have used the three different prompt patterns here, even you can go in depth Act as expert prompt writer which you have ten years of experience in particular AI model that is chargebty or Cloud. Even you can go in specific in specific, you can get the best prompt here. That is all about how you provide your requirements, okay? How you provide how you try AI in terms of prompting. That is all about this cocine refinement prom pattern. I have just told you some specific one use case. There are more ways to explain you, but that is now how you can learn this skill. This skill can be improved by yourself by practicing yourself by trying different use cases, then only you can get the better at prompting. I hope you understand this. Now you have some idea about how these prom patterns works. Let's jump into our fourth prom pattern, which is more advanced rather than these three prom patterns here. Let's start our fourth prom pattern. That is caved to verify prom pattern. Let's dive into that. 15. 3.4 Cognitive Verifier Pattern: Okay, let's start our fourth prom pattern that is cognitive verifier pattern. As we earlier discussed some prom patterns like persona prom pattern, ask me for input prom pattern and quotient refinement pattern. In the three prop patterns, we have seen some examples like? Now in this prom pattern, we will see what is actual this prom pattern works, how it works, and what is the main important thing we need to learn from this prom pattern is very amazing, right? So this prompt pattern is very important when compared to other prompt patterns, which we are going to learn in previous or upcoming lecturers. Why this cognitive verifier pattern is very important? Why? Because if you take any LLM like haGPTCloud deep sk, any LLM, so they will generate the output based upon their trend data, okay? They will only generate the response based on their trained data. With that, we can't expect the output which is real equals to our requirements or effective or accurate with our data. For that, this cognitive verify pattern will help us to cater our output into the effective and accurate according to our instructions. How it works, actually, how it works, we can see here. Let's see the example, how did World War two impact global politics? Ask me subdivided questions here. What happens here? I will ask to me subdivided questions related to this particular topic about World War two impact global politics. Ask me subdivided questions related to this main topic, which helps you to generate best overall output after I provide answers to your subdivided questions. No ask me subdivided quien. Don't confuse it. It is quite easy. A. What happening here? So when, for example, let's jump into ha GIP and we will directly learn in this. I'll just copy and I will come here and I will directly praise Char Jibe. Stee, if you see here, how did world want to impact global politics? This is our main topic. I want content about this particular World War two impact global politics topic. What happening here? This is our main prom pattern here. So how it works. Let's see. Let's see what happens here. So this is our main topic and we are looking to the content about this topic. What I am tell TEI, ask me subdivided quiensRlated to this particular topic. When I provide when I provide answers for that quotiens that you are generated, then proceed to generate the answer for this topic. This is all about our prong here. What happens here? For example, if you think instead of giving directly quotient, for example, if you are looking to get let's directly see here, we'll just copy and I will cancel it. If I just directly ask a question to directly to the AI about quien. This is again repeating our previous prompt method. We will now break this prompt chaining here. Simple prompt now, act normal. This is very important while interacting with EI because the ha have great chat functionality in which it will generate only based upon the previous prompting we have given. For that, we need to break this change for normal chatting. Let's say from let's say, what happens here, I will break the chain and it will just generate the answer for our question here. Now you can see here, memory is updated. It will directly answer for this topic. How did World War two impact global politics? No, you can see this is the answer for that. What happens here? You can see you can get this most of the best output when compared to all those things. But what is the use of this prom pattern here? If I put here, for example, by directly put here, Counter place and we will take this one. And I'll paste here. Sorry. Let's take here. Now what happens here? When I tell TI ask me subdivided questions to generate a best output, it will start asking questions to me. You can see here, superpowers and Cold War. How did World War two contribute to the rise of United States and Soviet Unions as a superpower? No it asking to me to provide the answer for this. So what is the difference between that? Okay? If LLM is trained by large amount of data, right, so they can make the mistakes. Okay, so the output is based upon their training data, but we have our own data. When we have our own data, we need to give to AI for better output. I hope you understand this point. Okay? For that, this prom pattern can be helpful. Okay? So in this not case applications, but we have different as a teacher, you know, you have to provide some guidelines while correcting the answer. Okay? For that, you need to use this prom pattern. Tell to AI, ask me subdivided questions relative to these guidelines. When I provide the guidlines to you, follow the procedure. That is simple. Now, AI will follow guidelines only instead of just taking their own guidelines. Preference, it will give preferred to guidance. After that, it will act according to your requirements. I hope you understand. When I give answer to this particular question, Okay. So it will start generating the output, right? So this will create some accurate response when compared to this. Even it is very accurate response, but in some cases, we need to give our own data because any LLM doesn't have some access to data, right. So just a chat model. They don't have access to our privacy data or own access to database like that. So in that case, why we have to give to AI. In in that application, you need to use this prom pattern because you are providing your own data to analyze it or to generate a output based upon this input data in which you can get the accuracy of the output. You can see here. When I provide this, for example, you can take. I will start providing I will provide answer. First one is how did World War two contribute to rise of Interstates and Soviet Union as the superpowers. So if I just tell to AI, for example, this is not equal this is not equal match of this prompt pattern. For example, we can take another one, for more in depth understanding. I have tell to AI, just take simple prompt for better understanding. So what to AI? How did World War two impact global politics? So for that, we are going to see, let's take something about calculate. Calculate calories calories of my body. Let's see what the output will be. What happens here? Allocate your body's calorie needs accurately. I need some details. Please answer the following question. If you focus here. What's happening here? It will asking questions to you. After you providing the answer for these particular questions, then it will calculate the calories and it will show you. That is simple. You can see here. What Let's start giving the answer for you. What is your age? Let's take my age. Let's start answer to these questions. I will take first, what is your age? Let's take 25. Number two, answer is what is your gender, male, Number third potion, let's take what is your height in 5.9? Let's take 5.9 feet. The fourth portion is, what is your weight? Let's take 159 number potion, what is your daily activity level? Let's take exercise, exercise, small exercise or little bit of exercise. Can take? What happening here? After I given answers for these Qs. It will generate them. I will calculate the calories of my body. Why here, the AI doesn't know who I am. Who I am chatting with? Who is that user chatting with whether it is a male or female? The AI doesn't know. But that, what is my requirement here? Just callate my calories for that it doesn't have some data about me, but that it will asking question. About myself, my information like age, gender, height, weight, which are required to calculate the calories. In that case, you need to use these particular prom patterns here. This is the Pi to verifier in which you are using this prom pattern to give input from your side to generate the output, which is accurately towards you. I hope you understand this point. Let's see what is the output will be. No, thanks for providing your details before. I carry over daily calories needs. I need a bit more calification on your activity level. Let's say it will again asking the question, let's give you the answer for direct layer. Lightly activity, let's take lightly activity. If you see here, no, you can see generating it will start generating the calories. How you metabolic rate, what is a metabolic rate will callating that? You can see adjust for activity level, total calories, calorie adjustment for goals, all those things. You can see the output here. This output is very effective according to my data according to my inputs. Why? This output is very effective and accurately matches my requirements. But that you need to use this prom pattern when you have data and you need some assistant by AI, then you need to use this prom pattern to try AI. Just ask me some questions related to this topic. When you give you own data, then it will start generating the content or anything that you are looking to get from AI. I hope you understand these points very clearly. This is all about this prom pattern here. I have taken simple one query. You can take more complex queries or even you can take analysis purpose or formula purpose or generating lesson plans, outline, story, you can take anything, but you need to use this prom pattern. Ask me subdivided questions related to the main topic. What is the main topic here? Calet colors of my body. Not a topic, but it is some task, you can take the task. Instead of main task, main topic, task, passage, anything you can take, which helps you to generate best overall output. After I provide answers to your subdivided questions, ask me subdivided questions. You can see here now ask ask me for input prom pattern as we earlier discussed previous lecturer. Now what is our next prompts. Before going that, let's combine all four prom patterns, ask me for input prom pattern, persona, and we will take quotient refinement and this prom pattern to try whether these four prom patterns can generate something amazing. Even more effective. Let's take second prom pattern that is persona prom pattern. Let's take Act Al g. You can take anything about that. Let's take act lz NeuronistNu. Nutrition, special Nutonist with ten years of experience. Let's take. Okay. Now, you can see here. It is our personal prom pattern here. I have assigning this specific role to AI. That is special neuroni with the ten years of permans. Now, we jump to. I will provide answers. Let's take directly here. You tasks we will ask me input prom pattern as well as we will take powder to verify. Ask me subdivided quotient. Ask me subdivided questions related to the top. For better understanding, we will take directly here, ask me for input prom pattern. What I have to tell here? I will provide I will provide task or aliste. I will provide a task. I will provide a task. Your goal is to Okay. Your goal is to assist in that. Here, the task specificity is completed. You can see. Now, we will just tell I ask me. Ask me subdivided portions. Related to main task. What we have taken here, just to copy that. Quickly, I will take this which helps you. Helps you to generate best overall output after I provide answers to your subdivided questions. This is ins. Now ask me task. If you see here, we have used three prom patterns here. First one, persona prom pattern. After that, we we have taken. Ask me for input prom pattern here, and third one is cognitive verified pattern. That is, ask me subdivided questions up to here. What happens here, let's? Will tell me, no, I will provide a task here. You can see here. Got it, please provide your task and I will ask detailed is to ensure the best possible guidance. After I provide the task, it will start generating the quotients to me. What happening here? Please be attention here. I will provide a task. Suggest suggest best puts to lose fit. This is our task here. What happens? Let's see. Okay. The AI is started asking the questions to me. You can see here to suggest a best puts for weight loss, I need to understand your dietary preferences and lifestyle. Please answer the following question. After I provide the answer for these questions, then I will start generating the personalized and accurate diet plan to me because the AIs know my preferences. How by asking the questions line does. After I past answers, it will start generating to me the diet plan. You can see what is the power of this prom pattern here. This is simple examples I have explained to you, but you can take any application, any preferences that you like. This is the power of things here. When I provide this, let's take when I provide the answers like vegetarian, like gluten allergy, small fricken meals, FS restorant food. After providing the answers for these questions, the AI will start generating the output to me. What is the benefit of using this prom patterns? It will start generating the accurate and highly personalized output to me. Instead of just if you ask without using this prom pattern, if you ask the AI, suggest best foods to lose weight. It will simply suggest the foods. But there is a whole content, but you need your own preferences output, but that you can use this cognitive if prom pattern for better personalized diet plan or for accurate response according to your requirements. Okay, I hope you understand this. We have used the three prom patterns here, but we have missed the third prom pattern. That is potent refinement prom pattern. For this, what we can do. We can just copy this prompt. Until play C. Now, you can directly ask here. Suggest me. Suggest me. Better version of by prompt. You need to mention here to break the previous chain. We need to call from now, at normal. Let's see. This is our main prompt here. This is the potion refinement prom pattern. Suggest me a better version of my prompt. That is just use above prompt here. What happens here? Interactive while you are interacting with AI, you can get the most effective output. Let's see what is output with B. It thinking. So memo updated. This is the best prompt here. There is no accurate, but they have some great professional prompt here. A rather than we have written earlier. It is a better version of this prompt here. You can see how effectively the IIs written the best prompt here. For above task, act as experienced nutritionist with ten years of expertise. I will provide a specific nutrition related task and your goal is to assist me by asking detailed follow up questions. This quotien should help refine the task to generate the most accurate and personalized recommendations. While I answer, proceed with the best possible output. Now ask me for my task. You can compare this prompt and this prompt. So how professional the AI is generated this prompt here. Okay? That is, you can use these prompt patterns. We have discussed the four different prom patterns till now. Okay? These are the most important as well. There are upcoming also. But you can maximize the output with these prom patterns. These are the most fundamental LLM prom patterns, so we need to know. You can take the prom patterns, you can use in any LLM. But in ChargePt they have different capabilities rather than other AI models. We will cover all those things in later classes. Pops on the prompting here. Okay. This is all about this cognitive verify prom pattern in which you need to provide your answers to the tiens that will ask by AI, then it will generate the personalized answer according to provided answers for the quotients that asked by AI. That is simple. This is best when you have your own data. When you have your own data. I will generate the personalized one, that is simple. We have seen four different prom patterns. So we have another prom pattern to complete a part one. That is outline expansion prom pattern, which helps you to generate the content in depth, in depth and very effectively. Let's start that. Let's see our next prom pattern that is outline expansion pattern. Let's dive into that. 16. 3.5 Outline Expansion Pattern: Okay. Let's start our fifth prom pattern that is Outline expansion pattern. As a name suggests, you can see her. What is the outline expansion pattern? Outline means it can be anything. That is, if you're looking to create a document for your class or any experiment document or anything about writing the content or writing the lesson plans lessons. You can see the outline. Outline means specific the bullet points or any lesson names, structure, all those things comes under the outline. Hotline means the whole structure or whole structure of particular lesson or book or anything experimentation or anything like that, that document have you can take the contents. Lesson contents. All of it comes under the outline, right? I hope you understand this point. What are the outline expansion prom patterns here? So it will expand the outline in depth. It goes into in depth. Instead of just taking just knowing the subject, it will generate it will expand the outlines or bullet points regarding the contents in which we have. Or if you can take any subject, for example, you can take any subject, test book or workbook like that. You can see some contents are there. For particular lessons, there are different bullet points, different topics for different lessons. That is all comes under the outline. In what happens there, it will directly they have some content. But here what happens for every bullet point, I will in depth of that particular bullet point to generate the more sub bullet points. Again, it will go to the sub bullet points to generate the sub bullet points of the previous one. So don't confuse it, we will understand or we will discuss in more depth. Okay. You can see this is a structure of this pattern. You can see the intial prom setup. So there are some several steps we need to follow to use this prom pattern. So for that we need to focus on number one that is intial prompt setup. So what is the initial prompt setup? It is a type of potion or it is a type of war requirement that you are looking to generate the content from AI. That is simple. Next, it will generate a bullet point outline. So in the prom itself, you will describe what outline you need. That means I which topic you are looking to get the outline or structure of that content. It will generate some bullet point. In this step. After that, you will start telling to AI interactive expansion. You will start telling to AI, now expand the particular outline in above structure. It will take that outline specifically and it will generate some sub bullet points regarding to the bullet point that you pick up right now. That steps will take here after that iterative exploration. So what happens here? So you can take this step as you match as you can. So there is no limitation for this because AI is trained by large amount of data, so it can go in depth in particular subject or anything like that. Iterator means you can follow this step many more times. Many more times, so it will generate the best output in depth of that particular subject. That is simple. And final output that is final answer. Final output that you are looking to get from AI. This is all simple five several steps. We have to keep in our mind while using this outline expansion prom pattern. Let's see some example to better for understanding. You can see I have retained some simple requirement of this prompt, act as an outline expander. This is most important because we are assigning the specific role to AI to act as a outline expander. After that, generate a bullet point outline based on the input that I give you and then ask me for which bullet point you should expand on. Please focus on this. What's happening here? I am directly, I am directly using the personal prom pattern here. Okay. After that, general I have described the task. Okay. Based on the input that I give. What happening here? We have used the ask me for input prom pattern here. This one and this one will combine to give, ask me for input prom pattern. Okay, you can see here, input that I give you and then ask me for which bullet point you should expand on. Each bullet point can have atmost three to five sub bullets. The bullet should be numbered using the pattern or you can take any pattern here, one, two, three or small AB or letters. Now, create a new outline for the bullet point that I select, at the end, ask me for what bullet point you expand next. Ask me for what to outline. So this is a bit confusion. This is some prompt some bit confusion, but you can understand directly but chargB by Ava doing the output. Let's jump into Chagbt and we will see how this prompt pattern will works. I have already copied this prompt pattern directly here. I will just go to hagibt. Let's see here. To break previous chain, just write from now, act as normal. So I will directly pass this prom pattern here. It's taking simple. Let's copy this. Let's come here, page here. This is our prompt here. Act as the outline expanded, generate a bullet point based on the input that I give you, then ask me for which bullet point you should expand on. Each bullet can have utmost five bullets. This is the task. This is the main task. Next, the bullet point should be numbered using. Up to this, up to this it is all about task, main task. This one is ask me this one persona prom pattern. This one is ask me for input prom pattern, and it also combined ask me for input prom pattern. Let's see what is the output here. No, I will tell it will ask some topic to me, okay? Got it. What would you like me to outline? I ask me a question. When I tell to AI, for example, when I ask any topic to AI, it will start generating the outline for me. Let's see. A topic. Let's take any topic, let's experimentation about or stay impacts of World War one and two. Let's see. This is our topic. Now, the AI will generate the outline regarding this input here. Let's see. You can see here, it will generate some outline regarding our main topic. That is impacts of World War and World War two. You can see here. I will start from the political impacts. This is some sub bullet points as sub bullet points regarding this main topic here. That is simple. You can see here, it is some outline. When I tell to AI if I tell to just take fourth outline, it will start generating the sub bullet points of regarding this topic here in more depth, let's see. It will expand the fourth thing that is technology and military advancement. You can see here. It will expand the outline regarding the fourth one topic. That is technology and military advancements. Even have generated the most in depth of outline in depth outline of that particular sub particular bullet point in the previous one generated. N you can select from this here to generate even more in depth outline here. Let's take third one directly here. Okay. Now if you see here, then it will again expand the outline. For the specific third one here, that is evolution of military tactics and strategies. It will take in some other thing. Let's take which one it is taken. I will taken from the previous one. This is the point here. We have to tell you here, we have to direct take this one, evolution of the strategies. I know what happens here? It will start generating the even more in depth outline, outline of this particular third one. Uh, you can use this method for many more times to get the depth content of that particular topic. There is no limitation for this. You can use this prom pattern as much you can for generating the best output. This is best when you are looking to generate some content for your e book or lesson plan or lesson content for your students, or if you are looking to prepare document so you can use this, right? You can use this outline expansion prom pattern. If you don't have any idea about what to explain to students or what to include in the document or e book, you can use this outline expansion prompt pattern to generate all the contents or to generate all the outline in which you can get the idea. And you can start generating output for this particular each and every outline. This is all about this outline expansion prom pattern. So please focus on this prom pattern here because it is very most important. It is all about how you instruct AI, how you will guide AI. In which step you will hotel to AI. This is most important. What happens here now, this is the break chaining. Focus from here. Act lase outline expander. We have assigned the specific role to I that is outline expander. What is a task generate a bullet point outline based on the input. It will generate the bullet pain. This is the bullet point outline, Okay. That based on the input that I give you. We have given the input here. That is simple. Ask me for which bullet point you should expand on, you can see here. Which bullet point would you like to expand on first? This is how you will try and AI to polyer steps or work like you want. That is simple. After that, each blend bullet point can have at most three to five bullet points. You can see five and you can see here five. It will following your instructions very clearly, at most five and the bullet point should be numbered using the pattern. We doesn't use pattern, so it can take one or two. I will take it will follow instructions. That is no problem here. After that, create a new outline for the bullet point that I select. After I select specific bullet point from the previous one, it will start generating the in depth outline for the particular outline that I selected. You can see the output here after that. Next thing, okay. At the end, the end, ask me for what bullet point to expand next. Again, it will tell you, I will tell to me, what would you like to expand on next? When I give any topic or anything regarding this outline will start generating the again in depth outline of the above previous one. That is how it works the outline expansion prom pattern. I hope you understand this prom pattern. Before completing this part one advanced prompt engineering, I'm giving you assignment. Please use all the four prom patterns, including these five prom patterns and combine all the five prom patterns towards spam uh one prompt for the specific application and see how it will works. Just try to build up that momentum or build up all the instruction formula. You will get the best output or you will get how to include prom patterns, different prom patterns to dive some particular topic or particular output from the AI LLM models. Please practice with your more examples. Combine all the 45 different prom patterns we have discussed until up to five, combine all the five prom patterns and write one prom pattern, one prompt, F specific application for any generating the content for your e book or generating content document for your lesson. You can see, you will learn how to interact with AI in effective manner. That is all about this out and expansion prom pattern, up to now we are completed part one. That is Advance prompt engineering part one. Let's start our Advanced prompt engineering part two in which we have see some different and most amazing interesting prom patterns that is tail generation. Next topic is tail generation. Let's dive into that. 17. 3.6 Tail Generation Pattern: Let's start our part two of this model number three that is advanced prompt patterns. As we earlier discussed some different prompt patterns that is about ask me for input, prom pattern, question refinement, persona, et cetera. In that prompt patterns, we have seen some beautiful and effective outputs. By writing them in effective manner by asking questions to AI directly or like that, we have seen some of the most important and creative prompt writing techniques in earlier apartment. Now in this part two, we are going to see even more advanced prom patterns, which helps us to understand the lens very clearly and very effectively. For this, these are the five other prom patterns which are very works for any application line, not only teaching point of view, you can use any of requirements, Let's see what are these different five prong patterns. Let's start from the number one that is tail generation, semantic filter, menu actions, packed checklist chain of thought. These are the five prong patterns we discussed in this part two. Let's dive into first one. So tail generation. What is actually ten generation is. As the name, you can see here, tail, tail means it is a bottom section of particular object. So for that, when it comes into the prompt pattern, tail generation, what happens here? At the end of prompt, you will guide AI to do something after the output. Okay? I hope you get the point here, what I am guiding in this I prompting. Like I am telling to AI, first generate the output for my question and at the last, tell me to do the next task. It will similarly works like ask me for input prom pattern that were earlier discussed in the advanced prompt engineering part one. Like that only, but these have some different use cases. Let's see in this. To use this pattern, your prom should make the following fundamental contextual statements. At the end, repeat Y and or ask me for X. So you can see here. You can replace YX with your task or any goal or any R requirement. At the end, this is the most important thing. At the end miss, it is a tail generation. Tail generation means at the prom tend. At the prom tend, then repeat Y. We need to tell AI for every output. You will generate repeat this particular task, or ask me for to do next task. Line, I hope you understand this point. So you can see this is a similar use cases when we are looking before lecturer ask me for input prom pattern we have seen Outline expansion prom pattern. That works like a similar only, but it has some other use cases. You can see it is a template of the prom. You will need to replace Y with what the model should repeat, such as repeat my list of options or X with what it should ask for the next action. This is how it will work here. Let's dive into some example how it works. Let's jump into Cha GPT. I'm already in the Char JBT, so I have already copied right in the prompt here, so I'll just paste it. You can see here. Describe the water cycle for middle school students. This is my task here. This is my task. What happens here at the end. At the end means we need to use some fundamental contextual statement here to use particular tail generation prom pattern. In this case, we are using at the end. At the end, ask them to explain the process in their own words. What I am telling to AI? Describe the water cycle for middle class students. Now, at the end, ask them to explain the process in their own words. After generating output for this particular task, the AI will ask a question. Two students to explain the process in their own words. That is simple. This is a simple use cases. We need to use as a tail generation prom pattern in v applications. This is a simple prompt here. Let's see what happens here. As earlier said, we need to break this above chain in above chain to make to do some normal chat with EI. As we earlier discussed, if you know earlier about it, you can understand this. Let's see what happens here. Let's break the chain first. From now. At no one. This is our main prompt pattern here. This one is for to break the change or not to follow previous task, okay? Let's see this. What happens here? It will start generating our answer for that describe the water cycle for middle class students. Now, you can see here it will start explaining about the water cycle. You can see the water cycle is a process by which water moves through different parts of the Earth, evaporation consideration, all the stuff. You can see here. Now, in your own words, can you explain how the water cycle works? You can see here. At the end, the EI is asking to me, can you explain how the water cycle works in your own words? How it works in this simple. If you see here, it will telling us at the end, repeat Y. What happening here? It is repeating. Now, if you check here, you can ask either repeat Y or ask me for the X. In this use case, we have use ask Mf to do something X. In that case to explain how the water cycle works. Instead of here, we can tell to AI repeat at the end repeat repeat. This output this output can have mistakes Okay. Please check it out before you use it. What happens here? For every output from now onwards, for every output that generates AI, we repeat this statement. What is the statement here? Outputs can have mistakes. Output can have mistakes, please check it out before you use it. Let's see what happens here. Now, you can see here. It will explain the water cycle, evaporation, condensation, all those things precipitation. You can see here. Output can have mistakes. Please check it out before you use it. At the end, it will just showing this message. It is repeating, how it repeat if I ask a question another question. Explain Explain about evaporation. So what happens, let's see. It will start explaining about the evaporation. Can see what does our operation happen? Why is evaporation important? It will repeating the same message again. Output can have mistakes. Please check it out before you use it. So how this prom pattern it works. What is all about. Why it is recognized that the prompt pattern here. So from now at normal, it is previously break in the change. Now it will follow this only. You can see here this is the task that we have earlier described that at the end. At the end means we are using the tail generation in which we have to describe AI. At the end, please ask me this question or repeat this sentence or you can take any task, anything about that. You need to use this at the end, right? You can take not only that then, you can take this at the top, like that. But it is a tail generation we need to use at the last. That is all about how you can use it in your way. We have seen these two simple use cases. This is a simple examples I have taken. Now, you can go in deeper with your use cases and applications by yourself. So what we can tell you AI like that. Okay. So whether you can repeat some particular task, you can define AI in AI in the prompt itself or you can use, ask me another task to do from myself. No, you know much like this previously we explained, right? So now you assignment is use all the prom patterns that we are earlier discussed up to now. That is six prom patterns we have earlier discussed, right? From part one to this part two number one, that is tail generation, use six prom patterns to write a single prompt, F specific application that and literally improve your prom writing skill. Please explore this each try it by yourself to write combining the all prom patterns for a specific application and try to evaluate the output that can literally can improve your prompt writing skill for your teaching experiences. I hope you understand this prom pattern very well. Let's dive into another prom pattern which is very important, as to filter out our outputs. That is semantic filter prom pattern. Let's dive into that. 18. 3.7 Semantic Filter Prompt Pattern : Okay, let's start our second prom pattern that is semantic filter. As a name, you can see filter. Filter means we need to filter out, we need to remove any unwanted or misleading information from the AIs output. So as we earlier discussed that AI can do mistakes, due to it is developed or trained by large amounts of data, which causes some mistakes in that. So we cannot over or we cannot 100% tell that Is output is 100% correct, but we need to evaluate it. We need to correct it. We need to check it, we need to proof read it, whether it is giving the correct information or wrong. Further, we can use this semantic filter option in which we can remove some unwanted or misleading information from the output itself. How we can use it. Let's see. To use this pattern, your prom should make the following fundamental contextual statements. Filter this information to remove X. This is a simple o. This is a simple fundamental contextual statement we need to use in the semantic filter prom pattern. You can see you will need to replace X with an appropriate definition of what you want to remove, such as names, dates or extra information in lesson tent. It is a simple prom pattern that we can use, but it will help to evaluate the output very effective matter. Let's jump into hagibt and we will see how it will work with effective manner. Let's do that. Here in the chargeby. Let's see how it works. As I earlier said, we need to break the chain to break the chain, we need to write from now at normal. Now I have already copied this example for you for better understanding. You can see here. So what I have to filter this historical passage to remove any violent details, making it suitable for elementary students. In this simple passage, I have written some violent details. You can read here. The French revelation was a time of great upheaval. Many people were execated using this kilot and battles led to thousands of deaths across France. But this passage of some violent details information. So what I have del to AI, now remove any violent details from this passage, which will suitable for the elementary students. Now, AI will remove any violent details as my requirement. Let's see what the output will be. You can see, it will just remove some violent details. It is the output is great when compared to this one. The French Revolution was a time of great change. People wanted more fairness and new rules for the country. Many important events happened and leaders made big decisions that shaped the future of France. This is very cool for the elementary students. Like that we can use, for example, you can take a filter out Okay. You can di, filter out the words that you remove in above output. So you can see, I have Telo A, filter out what are the words you have removed in a put you can see here, executed Gilda tone patels are the removed words and it will make this simple cooler the elementary students. Now this is how you can use this prom pattern. This is all put one d a use case, you can see here here, I have given the passage from myself to AI. What if the AI is generated some content. We need to cover all those things. For example, if I write any question to AI, let's dive into here only. Let's take any question here. Explain. Now, explain about precipitation. Let's take precipitation in re forwards. I have just taken some small question to AI. That will explain about the precipitation This simple quiet and simple answer from AI for the question precipitation. What happens here? No, I am looking. This is the output from AI. As I said, sometimes the output is have some words that I cannot understand very well for the sake of students. In that case, find what you need to tell to AI. Filter. Filter. The words have more effect, which have more effect to understand to understand. Let's take one other thing, which have more effect to explain second class student second grade students. What happens here? I will just filter out the words which are not very well to explain the students for the second grade students, because the second grade students doesn't have some words meaning in their mindset because they are the second grade students. Let's see what happens here as a you can see here, filtered words here, sleet hail, droplts, temperature, precipitation. What happens here? The AI is no, for the second grade students, what they can have knowledge about it or the terms of subject or in the terms of knowledge about they have. For the second grade students for that I have just AI, filter the words Okay, which are effective to understand or which the second grade student cannot understand about them. For that it will just remove some filter words here. That's fleet hail drop plates, temperature, precipitation. The thinking these are the words that second grade student cannot understand. But that it will just retain in the great manner that even second degree student can also understand. Understandable. You can see here, precipitation is when water falls from the sky as rain or snow, that is good when compared to this one. So this is a simple use case for you, for example, it is best use cases when you are looking to explain more complex topic to the students. In that case, you can use this prom pattern, use your particular lesson part of which have some complexity to explain. Just come to here, use this prom pattern that is filter this filter this particular lesson in adjustable, or filter the words which have more complexity, complexity to explain the students for a particular student grade students. Particular students. What happens here, the AI will think, these are the sum words which have in this particular lesson that particular students cannot understand. It will just remove that words and it will suggest a better lesson in which you can use in your teaching that students can understand very well. I hope you understand this prom pattern very well. As I say, use up to how many prom patterns we have used, use all the prom patterns and write a single prompt, combining all the prom patterns, write single prompt for the specific application that you can get some idea how the LLMs works very effectively, that you can use in your Di work life or personal life to maximize your potential. I hope you understand this point and prom patterns very clearly. Let's dive into another prom pattern that is MDO actions, which is very interesting and it can help you to do some automation process. Let's dive into that. 19. 3.8 Menu Actions Prompt Pattern: Okay, let's start our third prompt pattern that is menu actions. So as you can see the name here, menu actions. Menu actions means the prompt pattern which have some actions in the form of menu. If you go to any restaurant and you can see some menu options there. So you can see that is example, but this is quite different from that. Menu actions means it is a set of actions. Okay? It is a set of instructions which perform the AI step by step from the task. Okay. Uh, we will see how it will works. So to use this pattern, your prom should be make the following fundamental contextual statements. You can see the basic template here. Whenever I type X, you will do Y. You can assume X a particular task or Y something instructions. That is up to you. That means you can replace with any task, goal or anything. You can see optional provide additional menu items. Whenever I type I type Z. You will do Q. If you focus here, whenever I type, you will do Q. At the end, you will ask me for the next action. This is very most important. You can see here at the end, we are using here tail generation prom pattern. It is some set of instructions which perform step by step at the starting position. At the end, it will again ask to proceed next action. I hope you understand this prom template. Let's dive into our example. Okay. So I will directly go to habit and we start here. Now you can see here. This is the hagibT is following our previous prom pattern here. Again, to break this chain, we need to write from now. Let's write from now at normal. Which makes AI to break a previous ten. For that, we need to write. If you focus here, this is a basic example problem. I have written, you can see here. Whenever I type, generate lesson, generate lesson, plain topic. We can take any topic here. You can take any topic name or title. You will create a structured lesson plan for that topic. So we need to clear here. So plan topic. We need to give a plan topic. After that, it will create a structured lesson plan for that topic. So what happens here? I will first complete this task. It will go step by step from the scratch, from the starting task. Okay? It will just complete this task after that whenever I summarize topic, you will generate a concise summary of the given topic for the classroom discussion. You can see if when I tell AI to summarize topic, you will generate it will do this task only. If you focus here, this is the second task. What is the third task? Whenever I type, suggest activities stopping. You will provide engaging classroom activities related to the topic. This is a task after giving input this sentence, suggest activity stopping. At the end of every interaction, you will ask me for the next action. This is a tail generation as we earlier discussed. You can see here. So now what's happening here? This is simple like if you are using any app. So when you click on the button for doing some particular action, it will generate it or it will go to another page like that if you are in the website or like that. We are telling to AI, we are instructing AI. Whenever I give you this input, you will do this one. Here, whenever I type generate lesson plan topic, you need to create a structured plan for me for that topic, like that. Whenever I type summarize topic, you will need to summarize the given topic like that. By this, we can go automatically. Once we need to set up this prompt here after that, for every interaction, we can save our time. We can save our time and we will just write this words or sentence to do some particular task. So what are the benefits of this using this prom pattern here? It can save you a lot of time instead of writing, summarize this plan topic for every prompt. You can just write this one time and it will goes in automatically. After using this prom pattern in the like we have seen already. The Jagt have great capability that is memory update. From now, it will only follow this prom pattern in which we can automate the upcoming subpms of this bomb pattern. So let's start what happens here. So now the Cha JPT is thinking. No, you can see here memories update. Okay. Got it. Let me know the topic and action you would like to take. I will just give the topic, randomly, World War. Two, impacts on global economics. I have given this topic now. Got it. Let me know the topic. We have given the topic action you would like to take. Though what I'm telling generate. Let's take SRS topic. Now, you can see it will summarizing the topic here. That is, World War two impacts and global economics. You can see the answer. So at the end, it will asking, what would you like to do next? This is a tail generation prom pattern already we have earlier this gazette? So this is one. What would you like to do next? For every output that is generated by AI, you can see this sentence. Why we have tail to AI at the end of every interaction, you will ask me for the next action like that. Now if you see, I have just tell to a summarized topic. Now, if I give input like this one, generate Listen plan topics. Let's see what happens. Now, it will asking the topic. Now, if you see here, if you think, we need to clearly define here. Instead of plan topic, we need to give name of that is World War two impacts on global economics. Then it will generate all those things. If you don't give it will provide the topic for the lesson like that. To do the things, we need to edit some prompt here directly. How we can write this? Just tell me here. Here we can use something that is ask me for input prom pattern in which AI can ask me for the topic. But to do that, we need to write here at the starting. I will provide. I will provide a topic. I will provide a topic. This is our simple ask before input prom pattern. This is a task. Whenever a type and generate lesson plan topic, this all comes under the prom. Now we need to write now ask me one topic. After that, it will take another task here. Let's see what happens here. No will ask me for the topic. You can see here. Got it. What topic would you like to start with? Let's take the same one This is our topic. Et's check what happens. Now, would you like me to generate a lesson plan or summarize or suggest activities, please choose an action. If you see here, the AI is working like tool, if you're open any app, so it can suggest what I need to do next. Like that, the AI is working like a tool. After when given the topic, it will telling me what I would like to do for you. Please choose an action. If I tell to EI, generate a lesson plan. I will just write one here. This is how the EI works. Instead of writing generate lesson plan, I can write one here. I will automatically it will automatic think it the user is looking for the lesson plan. You can see here lesson plan, World War two impacts and global economics. This is how the lesson outline is generating. You can see here. This is the output of that particular task that I have designed to A. You can see here. Would you like for the next topic? Would I like to summarize our suggest activities for this topic next? Why? At the end we have declared to AI, you need to ask for the next action. Here, this asking to summarize our suggest activities. If I write summarize, now it will summarize all things information about the World War two impacts on global economics. You can see, this is a summarization. Again it will ask a good like me to suggest activities for the next topic so you can write, and base. Let's see what happens. No, it will suggest activities for us. If you see here at the end, what it will ask. Let's see. You can see here. Would you like to explore another topic? Let me know in the next action. It will ever break the chain because we have train AI to do this particular task only. For every interaction, it will generate or it will give the sentence that is you will ask me for the next action like that. This is how this menu actions prom pattern works. This is not a limit. This is a limit example, but you can use for your tile work life or professional life to build something unique. This works like a tool app like that. You can take any example, you can explore any other different examples and scenarios in which you can get the best output in which you can get very interesting in interacting with AI. So for example, you can take this example of. Whenever I type one number, you need to generate two number like that. This is how tools and you can take any example of the machine, computer, or phone. You can play with these prom patterns. Interact effect with AI to do some particular task. This is how you can use this menu actions prom patent. There is no limitation, just follow this prom pattern and this is easy, just explore more with your use cases, examples by yourself practice and you will get the foundation or you will master these prompt patterns. I am giving assignment to you. Please combine all the prom patterns up to we have learned. Combine all those prom patterns including this one and write single prompt for the single application or use cases and try to implement what will happening. And you will learn how to write effectively by using different prom patterns for the specific problem to solve it. Then you can leverage this AIMs full potential to maximize your output or to maximize your potential also. So I hope you understand this prom pattern very well. So practice by yourself for more experience with that. Let's dive into our fourth prom pattern that is packed checklist prom pattern. It is a very most important prompt pattern. We need to know as if you are looking to interact with AI to get the output for your work daily life or personal life. Let's dive into that. 20. 3.9 Fact Checklist Prompt Pattern: Okay, let's start our fourth prom pattern that is fact checklist prom pattern. So why this prom pattern is very powerful is. So as I said, the lens or trained by large amount of data, this LLMs can make mistakes. The output of the AI LLMs or have some ionization words or inappropriate or fake words like misleading information, which is not correct about particular topic or particular thing. To will filter that to filter that or to know whether the output is correct or not, we need to use this fact checklist prom pattern. Even though if you great at taking the output from AI, but if you don't know how to evaluate that output, whether how to find any mistakes in that output. If you don't know, but you have the great skill at leveraging the output, but you don't know how to evaluate it at effective level or in good level, there is a waste of time. For that, you need to learn how to evaluate output, how to decrease any misleading information in particular output, how to find packed that are correct and that are not correct, incorrect, like that. So to do that, we need to use this fact checklist prom pattern. What will happen here to use this pattern, your prom should make the following fundamental contextual state pens so you can see here. Whenever you output text, generate a set of facts. We are telling to AI. We need to use this prompt here. So you can see here. Whenever you output text means whenever the AI generate output, generate a set of facts. Facts means that is factual claims which are correct information about particular thing, that are contained in the output. The set of facts should be inserted at the end of the output. That is your requirement, you can tell at the end or at the top of the output, that is up to you. The set of facts should be the fundamental facts that could undermine the veracity of the output if any of them are incorrect. That is most important point here. The set of facts should be fundamental facts. What are the fundamental facts? The AI, which is generated the output, that particular output is based on some fundamentals of that particular thing. So we need to separate that fundamental facts to check out whether the generated output is correct or not based on the fundamental facts. If they are correct, so we can expect that output is generated accurately. If they are not, correct. We can say the output is have some misleading or misinformation. To evaluate this, we need to separate the facts from the output. To do that, we need to use this fact checklist prom pattern by using this pattern. Okay? I hope you understand this prom pattern. So this is you can use, okay? So for practical application, we will just jump into chat GPT and we will see how it works. So as we have said, so to break our chain, we have to write from now at normal it will simple automatically a break or change. What will happen here? I will try here. I will take this prop pattern past So Milton. What happens here? The AI is just our prom, you can see here. Got it. From now on, I will include a set of fundamental facts at the end of my responses to ensure clarity and accuracy. Now what topic would you like me to cover next? What we need to give, I have to give topic name. Let's take our previous one, we can see which we can evaluate the output. Let's take this is, let's take number one, World War I. This is our topic name that I have given to AI. So you can see, it will start generating the output. So if you see here at the end of this output, the AI has generated the fundamental facts. These are the fundamental facts in which we need to verify in the online, Google, or any other stores. Okay. Then only we can say this output is have some accurate information, which helps. By doing this, we can expect the generated output have some great accuracy rate. So took to evaluate the fundamental facts, you need to go so says like website, any official website or Google or YouTube videos like that. Then you can evaluate this prom. Sorry, you can evaluate these facts. Then if these facts are correct, then only you can use this output in your work. Why, as I said, the AI is trained by large amount of data, they can make mistakes. Okay, they have some misinformation, they can lead some misleading information like that. To avoid this, we need to use this fact checklist prom pattern. This is very most important prom pattern. We need to know how to use in the effective manner. For every output you generate from EI, you need to use this prom pattern at the end of every prom pattern. In case when you generating some document for your next lesson or lesson planning or content for your particular chapter. Let's say all about things. You need to use these fundamental facts from pattern. How you can use just write. Even, for example, you need to, you need to for example, you are looking to generate some lesson planning per particular topic or particular subject, what you can do, just go write, write a content for the specific subject and topic. You can use any of the prom pattern that we are earlier discuss for your requirement. At the end, what you can do, just write this prom pattern. The set of facts should be inserted at the end of the output. You can use this prom pattern. If you don't look like that, you can directly use this at the starting point of your interaction. I hope you understand this one. Use this prom pattern because for every output you generated from AI, you can see the fundamental facts in that particular output. Then only it can easily for you, which create evaluation output, which can make your output very verified and you can use in your work Daily life or personal life like that. So by using this prom pattern at the starting before you interacting with AI, before you are jumping into write your requirements or topic or that is your task. Just for the first prom, you need to use this prom pattern. Okay? For every output generate the AI will include this fundamental output in which you can directly verify each output, each output that you can get the accurate information from AI. Okay? I hope you understand these points regarding this prom pattern. So even you can go anyway. This is simple prom pattern that I have explained to you. But if you that's exercise with different prom patterns or a different use cases, you can get the most experience with this. Please once again, I am telling I am telling, which I will suggesting, go use all the prom patterns again and try this one. And write one prom for single use cases and you will learn the art of writing the prom patterns for the specific one. After that, your mind will be blow. You have build some automation or build some tool with the help of words, not with the help of programming language. This is how the PIS power is. You have great opportunities if you learn how to write a powerful proms per specific application. This is amazing skill. I hope you understand this prom pattern very well, practice by yourself to get more experience. In this prom pattern. Let's dive into our second, that is last prom pattern, which is even more important for the LLMs. That is chain of thought prom pattern. Let's dive into that. 21. 3.10 Chain of Thought Prompt Pattern: Okay, let's start our fifth prom pattern that is chain of thought. This is the most important prom pattern. So we have earlier discussed the different prom patterns which are very interesting and creative. But chain of thought is even more, okay? Important for Y. So I have some major benefits from other prom patterns. So as we al discussed some prom patterns which are dependent on the AI. Which will directly generate the responses even without thinking, right? That is depend on the different models, right? Even some there are so many models that deep sea carbon, now it is newly chatbard developed by China or we have open EI 01, 03 meaning. These are the reasoning models. They will think before generating the output. So we have using the HAGE PD that doesn't think about that it will just generate the output based on the input that we given without thinking. But chain of thought, what is the chain of thought? So you can see his chain of thought as name suggests. The thought like human being we have, we have some combinations of different thoughts to make step by step decisions for us. We will take some decisions by thinking step by step. By thinking the future or past, dependent upon the future or past, I present also. That is how the thought process will go in our mind at the same time. We will make decisions before, right. So after thinking only we will make decisions. Like that the AI also, the AI also generate output after thinking whether I need to give this output or not. I will think in different ways or step by step to complete a particular problem or task in the step by step process. That is simple chain of thought. You can see what it is, a prompt designed to guide the AI through a step by step reasoning process before arriving at the final answer. It will go step by step to complete or to complete the final answer to done the task, by step by step. Why use it Idle for complex problems, requiring logical thinking, multiple or multi step solutions. By prompting the AI to think out loud, you can often get more accurate and insightful responses. So as we earlier discussed, instead of writing the prompt for one time, for one time, so you can break into similar or simple steps. Like if you are looking to write the prompts up to five lines or four lines, but in that four filins you have defined AI ten task at the time, at the one time. But what happens here what happens here? The output is generated, but that is not a specific in depth for all the ten, all the ten task. But if you go step by step, if you write the prom for single task, it will generate the output in depth. After second one, if you write the second prompt for the second task, so it will generate the in depth of content for the second task like that. Because the output have some limit, some character limit that AI or ChagBT have or other AI models also. So to maintain that the I will do the I will generate some information about all the ten tasks instead of giving more information, some particular task. For that, we need to break down into simple and different task to complete one large task. I hope you understand this point. Okay. So what we happening here, we are going to step by step reasoning process. Chain of thought means combining all the tasks together to complete one particular task in the step by step process in the step by step reasoning process. I hope you understand this point. So for better understanding, let's jump into charge bity we will see here. Let's break the chain. Realistic what happens. Now I will just directly copy some prompt here so I can see you are teaching a fifth grade class about the water cycle. Explain the process step by step before summarizing it in simple terms. We can see what I'm telling I have told to AI. Explain the process in step by step. What happens it will go step by step reasoning. It will take reasoning instead of directly summarizing it instead of just explaining it. Let's see what happens here. But All right. Let's breakdown the water step by step. Evaporation, condensation, precipitation, collection. You can see it will just explain this all this one step by step and it will generate the summary. I think this is simple question or answer. Let's see, we will use another prompt here. This is one simple, but the step by step reasoning is usually mainly in logic reasoning questions or mathematics. Let's take explain Pythogoras. Explain Pythagorus serum step by step. Let's say what the output will be. You can see. Let's break down Pythagor serum step by step. It will goes step by step process instead of just throwing out the explanation about a Pythagorum. You can see it will start generating what is the Python theorem Pythagore serum, the formula, understanding the relationship, example, this is the example that has generated, this is a summary. Okay. This is quite very effective output, right? So this is a simple reasoning, right? So what is the actual chain of thought means? For example, this model, this is the Hagibt 3.5 or four is not a reasoning model, right? So you can use this reason button. Think before the HGP have newly has added this button reason. Think before responding. This is the chain of thought work like a chain of thought. For now, we will write some we'll write some topic or let's take any question and we'll use this reasoning model, how it will think, we will get some understanding about this. Explain let's take another that is properties of proper. Let's take properties of circle. Now, this is simple question. I have to do. We will use this reasoning model. Just click here. Think before responding. No, it will think. It will start thinking before responding here. You can see here, reasoning. Unraveling. It will thinking, what I have to tell unraveling circle properties, clarifying instructions, clarifying instructions. It is thinking. It is a thinking, okay. Chat here. Reason for 14 seconds, let's say, it is generated some output here. Circuit is a shape where every point and its edge is the same distance from the central point. You can see the output here. This is a reasoning model. It is summarized in simple terms and fundamental facts. So before. What happens here? It is not generated the fundamental facts here in previous output. But when we use this reason model. When we use this before responding model, no it is generated our fundamental facts in this output. Why? Why means the AI is thinking. I have tried to generate a fundamental facts in previous one. It will just thinking before responding, but that it have some great output here. Okay? This is how you can use this reason model, but what is the connection between reasoning and chain of thought? As I said, it will break down the step by step. To break down it step by step, we need to think it. Like that, the AI also think step by step to produce a best output. If you doesn't use this reason, the AI will just generate the output without thinking in which we can get some inaccuracies in that. If you use this reason model, then this is a thinking capability in which we can get the best output for that. Okay. That simple thing. Major benefit of this checkup thought is, what the actual prop pattern is explaining us. For example, let's take explain. Explain. Five different five different shapes. Five different shapes in geometry. Let's say I will just explain us about five different shapes in geometry. Circle, triangle, square rectangle, pentagon. It has some genderism, summary of shapes, all those things. It has a g fundamental facts in this output. We have just 12 to AI to include the fundamental facts in this output. Now what happens here? Let's see. I have tel two I explain five different shapes in geometry. This is whole one output. But if you see it is generated there's some information about each shape. But if I tell AI to explain only one tape, only any shape in geometry, let's see what happens here. One. One shape in geometry or you can specify particular shape name. Let's take this one. I will generate the particular shape that it has taken triangle. Now if you focus on it is generated the even more detail about particular shape in major points in more content. That same if you put all the different shapes to explain it, it will just tell us some basic things about each shape. So in that, you can lose some important points about the shapes. If you go specific one, for example, you can see the about triangle. It had just explain us what is a right angle scalentangle, all that simples, simple sorry, simple information here. But if I told AI in the specific manner, explain one shape in geometry, you can see how effective it output is definition, types of triangles, sides, angles, parameter of triangle, area of triangle, summary and fundamental facts about this particular shape. That is how we need to use this chain of thought. This is we need to write the prompt for each task. Instead of putting all the task in one particular prom to get the output, we cannot get the best output for each task if you write in only one prompt. We need to breakdown it for each task. We need to write different prompt here. In that, we can get even more depth information about that particular task. This is how the chain of thought also work. It will just breakdown the solutions. It will break down the goal hour. It will break down the task into simple things in simple steps, and it will verify each and every step to go to our final answer. That is how you can use the chain of thought. As we seen, but the chain of thought, you need to use the reasoning model. It will think before responsing you can use this for any type of or complex, uh problems to solve complex problems especially in physics, mathematics or any about that. You can use this chin up thought prompt pattern, which is very important to solve complex problems in mathematics, physics, or anywhere. You can use this. As a teacher, you need to know how to use this in effective manner. Up to this, we have understand what are the different prom patterns in which we can leverage the AI modules in effective manner to get the best output according to our needs. We have seen all those ten prom patterns and four basic prom patterns earlier. Up to this, our effective prom patterns lecturers will be completed. But remember, this skill is improved by practicing only practice by yourself with different use cases, with different examples, with different requirements and see how the IIs working, how is AI is generating the output. Then only you can use this in effective manner at maximum potential. Okay. I hope you understand these prompt patterns, all the prompt patterns very well. And as I earlier said, now I will also saying that I'll suggest you to put all the ten different prompt patterns together and write single prompt. At the starting level, I am telling the starting level, you get to automate something by using the words, not using the automation tools, not using the programming language, but using the words in AI. Okay. Use all the prompt patterns, write the prompt to automate some particular task for you by using the prompt writing skill in the charge itself. Now you can see how amazing this still is. After that, you will unlock more potential in this AI. I literally will change your mindset as you move forward. Up to this, we completed our advanced prompt patterns. Let's move on to our fourth model in which we will see some more advanced not a prom techniques, but some methods to leverage more output from AI and we will also see some different capabilities of JGBTjtEI. Up to now, we learned how to use JGBT. We will learn how to use Gemini Cloud Deep Seek and other AI models right now we have in this AI era. But don't worry. We are not just learning from the step by from scratching using the all prompt patterns in all those. But remember, this is the LLM prompt engineering, not a specific prompt engineering. Okay, this course is only designed for all the combining all LLMs. But why I am using hagibt for better understanding to expand in better way. These are the prom patterns which we have learned up to now can be used in all the other I models are the same. You can get the same output, but there is a difference in output, but the LLMs are same. The prompt engineering all the same. But we need to know how to use that models because every model have their own capabilities, strength and weakness. We need to categorize it. We will show, we will learn all those checking whether this I model is perfect our task or not. We'll see in upcoming lecturers. Move to our fourth model, that is, we will see how leverage I models in effect may by other methods. Let's dive into that. 22. 4.1.1 AI-Assisted Lesson Planning - Part 1: Okay. Let's come to our model number four, that is practical applications of AI in teaching. In this model number four, we're going to discuss some practical applications, how to write the proms for our requirements as a teacher, along with that, we will see how to use different LLMs like harGPT gem.ai Cloud. Perplexity.ai Deep Ck and Microsoft copilot. Let's check out all this to understand the different LLM capabilities. Even with this, we can choose the better AI tool for our requirement. We will explore all those things in model number four. Let's dive into our first lesson that is AI assisted lesson planning. As we discussed some different prom patterns even more advanced different prom patterns. You can use that all prom patterns according to your need to generate some lesson planning or other thing that you would like to explore. In this model, we will just go through some basic proms for our requirements to get the output from AR. Okay? You can see here. Why we use AI for lesson planning? EI is not 100% accurate, but we can save the time by generating some foundational outline, foundational lesson planning. What we can do, you can generate some structured outlines quickly by using AI. I will suggest relevant activities and discussion questions instead of thinking what I have to make the activities for students for a particular lesson plan. So you can give to AI to suggest it. I even more quickly, it can generate some activities and discussion questions easily, the third one is adapting content for different grade levels or learning styles. It helps you to generate the content for different grade levels. Instead of putting yourself, putting yourself into the work, taking more time. It will take more time by analyzing each and every student. You can't write and you can't differentiate content for different grade level students. It will take time. Instead of that, you can to give the context or background information of your students to AI, the AI can generate the content categorizing, according to your student behavior, and grade levels. These are some benefits of using AI assisted lesson planning. You can use any tool that you would like to get from AI. But remember, before writing the prompt before choosing the LLM, we need to know how LLMs work. We earlier discussed some different advanced prom patterns, how the LLMs work, how we need to give information from ourself, how to get the best output from the AI itself. We have learned all these. So before moving forward here, so you please try practice all the ten prom patterns with other AI tools. Okay. Let's take we have multiple AI tools like jm.ai Cloud, and perplexity.ai, DeepCk and Microsoft copilot and ChagBt. We have different LLM tools. Okay. So please make sure before moving forward from here, Please go and check all the prom patterns we have discussed till now, use that all the prom patterns practice in every LLM. Practice in different LLMs. Use all the same prom patterns and test it out, test it out gem AI Cloud at the same time. You can get the idea about how these lems are working or generating output. From that, from that, you can choose the best LLM for your requirement. That is how you can choose the best LM. Let's see some example, how to use prom for lesson planning. This is simple example we can take. What is the difference between weak prompt and better prompt here? So make a lesson plan on photosynthesis. It is a quite straightforward question, but it will generate a better lesson, but we don't given some background information. Even we cannot go in specific manner. How we can create specific manner. So, I will generate a lesson plan on photosynthesis for how many minutes, for which students. What I need to include? What is the actual content you want in this photosynthesis? Okay, This works better instead of using this direct straight forward. Prompt. I hope you understand. All this we are earlier discussed. Now we will see what is the better prompt here. This looks like better. Create a 45 minute science lesson plan on photosynthesis for seventh grade students. So it has specifically, we have trained EI to seventh grade students, include an engaging introduction, key concepts, a hands on activity, and a wrap discussion. So we have just told EI, what are our actual requirements we need to get a lesson plan on photosynthesis. Even you can go more specific like using personal prom pattern, act as a photosynthesis expert, like that. You can go even more specific act as a prom pattern and other advanced prom patterns as well. You can do all those things. But in this model, we will just see how we can use these prompts for our application. So you can see, by making your prompt more specific, the AI will generate a more useful response. That is good. Me specific your prompt is, then you can get the in depth answer for that particular task that you are given to AI. That is simple. Let's copy this. I already copy. Let's go to our first, that is chargb. We will check this particular prompt in all the different elements like har gibt gem.ai, right. Cloud, perplexity.ai, Deep Sk, and Microsoft Copilt. We will check this prompt in every LLM to understand which LLM is generating the best answer for me for our requirement for this particular prompt. Let's check it out for our first that is Ta GB. 23. 4.1.2 AI-Assisted Lesson Planning - Part 2: So as earlier discussed, to break the above chain, okay? To break the above chain, we need to write sin Abel. This helps to break out our previous chain of prompt. Because the Chagpt have great capability, great capability that is memory update in which it will just focus on the previous one. If I doesn't use this or forgot about, now only follow these instructions, this warning prompt, what it will do, it will generate the output based upon the previous. Previous one it will follow the previous format. It follows the previous pattern. So to break that we need to write. Let's quickly write our main prompt here. I just copied and pasted here. You can see this is our main prompt. Create a 45 minutesige lesson plan on photosynthesis. This is our prompt here. Let's see how it will create the lesson plan for this particular task. You can see. It is generated some lesson plan that is 45 minutes seventh grade student that is object to introduction, we need to cover these topics in 10 minutes, 15 minutes key concepts, we see hands and activity and 15 minutes wrap up discussion. It is quite good. You can see this is it has showing some let's take another model. We'll see what will happen. Then again, we'll generate so you can see lesson plan photosynthesis seven the grade science student, 45 minutes, lesson objective, lesson structure, 10 minutes quite very effective output here. 15 minutes key concepts. You can see it is generated, that's good. That is wrap up discussion about 5 minutes. This is our generated output here for the task. If you see here, it will also generating some assessment and homework, fundamental facts in this output. It is very important as we earlier discussed, we need to implement these fundamental facts in this output for every output unity check. Whether that is correct or not. I hope you understand all in earlier classes. Let's try out this prom in other LLMs as well to check whether which LLL works well for this requirement. I will just copy and paste in chem.ai. Let's see what happens. Let's see what the output will be. Or I can see here is 45 minutes science lesson plan and photosynthesis, title unlocking the magic of photosynthesis, good seventh grade. Showing some learning objectives, materials, lesson planning, introduction about 5 minutes, key concepts, roll off chlorophyll hands on activity 15 minutes, wrap up discussion 10 minutes. It will showing some class discussion summary, differentiation, assessment, not good. You can see the output here. L if you see here, this is some showing even more creative what you need to know. That is Whiteboard projector makers pens. When compared to hagiPDe it is just giving the main topic instead of putting all those other things like materials and other lesson procedure to follow. This is not required. We just required some lesson plan. So in which we can see the BD has generated a great output when compared to gem AI. It is also good, but it has some great points when compared to m dot a. It is given more in depth, it has generated some that equation also. W is in this equation. And you can see here. It is good when compared to gem.ai. Here, the emit A is miser the equation. It is simple, generated some 45 minutes lesson plan, but it is not quite effective when compared to hagBT. Let's jump into our cloud doti that is another amazing tool. Let's paste this prom and we'll see how the output will be. Now you can see it will generate some output. Wow, it's great. But if you focus here, it is also telling us not even good, but you can see here. You can directly open the document from here or you can see here. The Cloud have some great user experience when compared to CharGPT or other AI tools, it has some good dislike. Document you can directly see the document here or you can publish or download copy paste from this here. That is all about yourself. Can see some photosynthesis lesson plan, grade level, duration, learning objectives, materials needed, it is also generated some materials needed. It looks like we are doing experimentation for particular thing. But what we need, we need some lesson plan. We are not doing any practical experimentation. Now you can see it will it is generated like we are doing experiments on the photosynthesis, but it is not quite bad, but it is also good when compared to chargebt, but the cloud and Jemmy dot A, these two outputs are quite similar. What we lack in this emit and Cloud is they are missed the intent, what we are looking to get need. We just need photosynthesis lesson plan. Lesson plan means it is explanation about the even there are different lesson plans about that, but what we need actually is we need the content about the photosynthesis in which we need to clear end of 45 minutes. So you can see here. So it is the agebra has generated the great output, which is clearly matches our intent and it has generated the great output right instead of putting all those things. I also generated some materials needed, all those things, so it has quite matches our intent and output that we're looking to get. I hope you understand these points clearly. Cloud when compared to Cloud and get AI, Cloud has given the great output when compared to gemini.ai. You can see the output here directly. Okay, let's dump it over another AI tool that is perplexity.ai. Okay. This AI tool have the great features that you can research anything very fastly. Why? It will use the resources link in which from where it has taken the content for you. That is a great thing about this AI tool. So as a teacher, you need to know. You can see here the output how amazing this output. Here is a 45 minutes science lesson plan and photosynthesis tailored to seventh grade students. It has generated some very good things. Learning objectives, materials needed even more things. That is good. Lesson processes are engaging introduction 5 minutes, key concepts 15 minutes, you can see how well it is written and it is expand each and every part of the equation. Okay, that is good, emphasize this oxygen, role of flopll hands on activity, very good attractive output when compared to all these AI tools like cloud gem.ai, chargebT. It has great future. You can see how directly it can use you to discover all those spaces, library like that. Okay. So you can search images compared to this lesson plan, that is you can directly see lesson plan here. You can see all the things from the perplexit. It is a great tool for research purpose. You can research anything papers as a science teacher or anything. You can research directly instead of going to Google Search or any search engine. You will waste that time. Instead of that, you can directly come to perplex.ai and just write your quotien and it will summarize the particular paper and it will show that particular link that particular paper research link to you in which it has taken the content. We can expect that the output is very accurate because it is taken the output from the valuable resources that you can see there. That is all over this purpose. I can see. This is the best AI tool for getting the lesson plan because it has even more in depth, it is explained to me. Materials need a learning procedu in the part, you can see the output, how well it is written. Okay. Let's jump into our deep sik. It is a newly developed DI tool, AI chart board, China. Let's see what happens. You can directly search here or deep think here. Deep research. What is a major benefit of this is the deep sik is equal to the open EI 0101, which is a very costly model. This deep sik is providing free to us because it has a great capability that is thinking capability. Let's see here directly. I will paste a prompt here. Just go so now, it will showing I am thinking. That is AI thinking. You can see here how the AI is thinking about this question. This thing will apart from all the AI modules out there. This is the only AI chatbot, which is now available for free, and it is showing how AI is thinking to put the output for you, which is great part from other IITols. It is very most important thing. By this, we can expect the best output. Why? It is thinking how I need to give the output, how I have to gather information, all those things. It is providing the respond after thinking. It is responding after thinking in which we can expect the best output. As we earlier discussed it, that is chain of thought in chargeb itself, how do CT that is chain of thought work and which we can Okay. So now you can see how much it is thinking about that. So to produce a output, it is taking time. So when compared to all these other tools, they have just without thinking, they have given our lesson plan. But when compared to deep C, deep C is thinking. This is thinking to provide this 45 minutes science plan, which is great. Capability when compared to other all AI tools like chargebty Jem Ni Cloud, Microsoft copilot, if you take any AI tool, so deep seek is the best reasoning model for reasoning purpose. It will respond after thinking. You can see here, it is generated objective introduction 10 minutes, key concepts and hands on activity, wrap up assessment. Okay, it is not good, but it is even compared to this GEM cloud. These three things, DeepCC also works like a ChaGPT GMD Cloud. These REMs are work like, but they have trained by other companies and other ways. You can see some great output. That is a good one, but it is not effective like perplex.ai have. Why this happens? So let's see another copilot, that is Microsoft copilot. I will directly based here. This is also some AI tool which is developed by Bin. That is Microsoft Bing you can see here. It directly generated our 45 minutes seventh lesson plan. You can see it is a simple has generated that is key concepts, three hands and activity, wrap up discussion. If you see, these are the simple lesson plan that are generated by these I models like hA GVT. If you come to h JBT here, you can see here. Now you can see it has generated some even great, like this. So best output. When compared to Gemini Cloud, Deepsik and Microsoft copilot. But when compared to Gemini, what is apart this? Gemini, Microsoft copilot are the two AI chat boards which already developed by the world's great search engine platforms. Microsoft copilot is developed by the Bing chat right and Gemini is developed by the Google itself. These are the two search engines in which we can expect this type of output only. Why? They have some knowledge. It will take the knowledge from the resources like websites already they have in their search engine. These two outputs are quite similar. You can see here, gem.ai. These are two are same, but Cloud AI and ha gibt or two works similar. The output or the similar because they have even they are two chatbds are developed by different companies. But if you see here, they have some equal some type of equal output when compared to these two hagibt and Cloud. But deep Seek also one of the chatbot, we can see. These are the generators some great lesson plan for us. Now you can see, it also comes under the JGBT Cloud, Deep Seek, one side, Microsoft copilot Germany one side. But what is about perplexit EI? If you see here as a research purpose, the photosynthesis means it is experimental topic. Already so many scientists or so many people are experimented, all those things right in the papers research about photosynthesis. This perplexitEI is specifically designed for research purpose, ins will research online. For various resources and it will gather the information from other real resources like research papers, online, YouTube videos, websites, all those things and it will show you the best output. You can see here. This is very good. Very good output when compared to all the other models. Why? It will check all the other extional sources like research papers, all those things. We can export the best output here. As a teacher, you need to know how to use this perplexed.ai. Okay, but that you can see here. You can directly come here, you can just click here. I will generate another output. Here you can see. These are the resources links that perplexit AI has given to you from where it has taken the points to show you. You can directly click here. It will just go to website Link or you can go from directly here. I hope you understand these things. Okay. Is very most effective AI tool in the purpose of research or some particular topics which need real information, which you need accurate information, you can use this per placid.ai. Instead of that, for the creative for the storytelling, you can use GPT, Cloud, Gemini, Deep Seek, Microsoft C palette as well. But for the deep research, you need to know how to use perlasit AI. I hope you understand. What is the conclusion of it? For this a lesson plan prompt application. I am after analyzing all the six different EI tools, I decided to take the perplexity AIs output for my task. I choose perplexity.ai for to generate this lesson plan for me. That is simple. How we can choose the AI module, perfect AI model for our task to complete it. As a teacher, you need to know instead of writing the prompts, also and also we need to know how to choose the perfect EI model to do our task. That is simple. I hope you understand this lesson very well. This is simple first application, we have learned. Let's jump into another 4.2, that is EI for student assessment and feedback in which we will see some examples by using all these five or six I models. Let's dive into that. 24. 4.2.1 AI for Student Assessment & Feedback - Part 1: Let's start our second topic that is AI for student assessment and feedback. So as you can see, we are using these different applications, right, how to use different LLMs for different applications. We earlier discussed some one of the one application that is generating lesson plan. We have tested out this previous prompt in all the different LLMs like CharbCloud, all the other sixes, different ELMs. And we have seen the output of all the elements and we have chosen the better one. Okay? So now, let's do same for this application also EI for student assessment and feedback. What's happening here. How EI helps with assessments? If you put exams or assessment to your students, they will give the answer, but it will take time because in our particular classroom there are a lot of students which causes more time to rectify it to correct to correct the answers. Instead of that, you can use this EI to help to maximize your potential and to save your time by answering all those. Of course, many of the for example, today in this world assessment and all this takes place in the computer itself. There is no problem with that. But if you are looking to go in without computer using computer for assessments, so it can help you the better. Let's see, this is a creative example, we can try with the AI. Simple this is the simple use cases that I'm explaining to you, but you have to use different applications, different use cases with different prom patterns we earlier discussed in the previous model that is number three. You can use all nearly 14 prom attals we have discussed. You can use any of them or combined use them and to make something great impact and take the best output from the AI. But in this case, we are just looking. The main purpose of this model is to test it out different LLMs capabilities and we will choose better LLM, which will be the best sootable tool for our requirement. Let's see this also. What happens how AI helps with assessment, grading and giving feedback can be exhausting. A can help buy. So that is true. I have earlier explained to you. So what it can help. Generating quizzes and tests, that is easy. Instead of putting yourself to a search in each and every test book or to search in each lesson plan, you need to pick out some points and you have to make some quizzes and test. Instead of that, you can tell to AI like any tools, Gibet or other thing, other AI tools. I will in seconds, it will generate better quizzes and test according to your lesson plan or anything that you will provide to AI. That is simple. It can help you very time. It will save more time for you. That which helps in generating quizes and test. Next, providing instant feedback on the student response. So for example, it multiple choice is say, but when you compare it to short answers and essays, you can directly pick the answers from the students and pasted in AI chatbards. You have to design prompt like that. You need to design prompt pattern like that in which you need to just tell to AI. Please give me the multiple choice or short quotients, based upon the feedback or based upon the passage that I given. Or lesson plan. It will start generating the is that the student need to give the short answers. After when the students given the short answers, just pick the short answers and tell to EI. This is the short answers from the students. So please correct it. It will share your feedback. It will provide the instant feedback based on the student responses in which we can rectify it or you can correct the short answers of the student or anything essays that multiple choice also. That is how you can use AI to save time. Suggesting areas for improvement in student work. That is very creative after AI. I can help you to some tips, it can give some tricks to help your students to improve their self. How after giving feedback based on the student response, you can tell to AI, no, suggest areas for improvement in student work, automatically suggest some areas of improvement in your student. Okay, I will suggest based upon the student response of this particular quizzes and test questions after students giving the answers. I hope you understand this easy application. Let's dive deeper in depth. Using AI to create quizzes, how we can use. You can try this prom to generate quiz, create a ten multiple choice quiz and the causes of World War I for high school students include an answer key. So we have just given the simple prompt here, protein. What is a protein? Ask AI to format the quiz for Google forms or other assessment tools. This is a big west a tip we need to use. Okay. So for example, you can use these quizzts in the form of Google forms by Google. You can keep this already as a teacher, you know, there is no need to tell about this. Let's see this prom pattern, how it helps. I'll just copy this and we'll jump into first chargebtyH it can happen. As I said, we need to break the chain. When compared to other things, we need to focus on a jit because the jib have greater capability that is memory of t. I would have other platforms. Now I will just place this simple prong here. Let's see what happens. Let's see what happens here. It will start generating the quests for me. What was the primary cause of World War I? It is given some multi pul choice options. It has generated some ten. Let's see, let's see. It has generative some answer key. There is no limit in that. You can use prompt in any way in so many ways that's up to you because you need to know how to use the prom patterns very well. This is a simple application I'm explaining to you. The main purpose of this model is to understanding the capability of different elems like Char gibt Gemini Cloud, with some prompts. That's not good, but it has some great answers and all these things. Let's ask Char gibt to provide or anything, ask questions or provide. Provide short questions. That's what happens here. It will start generating the quotiens about what event directly triggered World War I. Let's see. When the student give the answer for this number first question, you can turn to AI, this is the answer. The answer of one student. Please rectify it. Questions suggestion areas we improve it. Okay. Okay. Now, you can see. Here, you can use one of the prom pattern here. I'm just giving you. Whenever you can write like this, I will provide the student response. This is called the ask me for input prom pattern as we are layer discuss it. Let's take I will provide student response above above question one. You can take one, two, three, four, or even ten questions. That is up to you, but I am telling as assuming for explaining you. Please rectify and give the suggested areas to improve it. This is the task, and I'll just write the now ask student response. So what happens here as we earlier discussed about ask me for improved prom pattern. Sure, please provide the student response to the first question and I will rectify it with giving suggestions for improvement. After providing the first answer, what was the event directly trigger the state of W Devar one? The will generate it will rectify it and it will suggest in areas to improve it. Instead of telling instead of writing the answer, I will write just I don't know about this question. What I tell you, this is a student response. Instead of providing answer to it, let's try what happens. It looks like student is unsure about the answer. Here is a way to guide them toward the correct response. The correct answer. The event that directly triggered the state of World War I was assassination of Archduke France, Ferdinand of Austria Hungarian, June tie, 1914 in Sarajevo. So suggest it has generated some suggestions for improvement. Encourage research, provide context, use memory aids, like that. You can see how the AI is helping to save our lot of time or to go more creative, right? So how you can use this. Let's jump into another LLM, how we can use this or prompt in other LLMs or how the output will be done. I'm here in the m dot a by Google itself. Let's paste this prom and let's see what the answer will be. So it is generated some quotients. That is great. There's nothing in there. You can see it is also generated some answer keys also. Let's see the great depression. What are the primary cause of PolvarO If you see the quiens here. Which of the following is not a major contributing factor outbreak, cultivar one? It has given some complex quotients when compared to this one, gPT. These have some catchy answers are related to our main topic in grade. It is according to our students. But in this case, we don't know about what imperialism, colonialism. It has some complex in them. As you can see, we cannot use the quiz for this particular Gemini. Why it is a search in the, they have a lot more data already in that. It is a fresh data. When ChargPT is mainly designed for personalization means it will the output is very quite easy to understand and have some good words or good explanation about one particular topic. That's why it has some popular. I gained the popular ha gibt also. Let's compare to Gemini and ha gibt. The Gem is more complex in generating quizzes. Let's take another prompt here at the same time. What happens? Let's check it out. What short questions? It is as a question it is generated, we will take another prompt here. So as I said earlier, before coming to this model, you need to practice all the different prom patterns that were earlier discussed. You need to practice with all other LLMs as well. Then only you can get which LLM is performing well, let's see this correctly. I am ready please provide the student response to the first question who was assimated in the triggering. That's great. Let's take our answer directly here. What happens in the journey? You can see here. The student response indicates a lack of knowledge about this topic. Here is how to rectify it and areas for improvement. Rectification, acknowledge the student response. I understand you're unsure about this question. Provide a brief and clear explanation. That's good. Areas improvement, encourage further research, provide study resources, break down the question, relate to the current events. By providing contractive feedback that is good. Okay. If you see here, the Gemini has given some more areas in improvement which are very factual and very important, providing steady resources. But AJAGBD is not generated in that format, you can see the suggestions for improvement. But the questions about this one, short questions or quizes, it's better than Gemini. But when compared to areas of improvement, the Gemini is given the more explanation about that. I hope you understand these points very clearly. Let's try out these proms in the other platforms as well. Let's take this one. I'll go to Cloud, take this. Okay, remember one thing. So the output is depend upon the model that you are selecting, right? The chargebras have 3.5, four GB four, four, Okay. So if you select the higher models, current versions of models, the answer will the grade, right? So it will works in all the lems also. Gemini have a different flash too. We have different models that 1.52 like that. If you use this point of view, so you can even can get more detail or very effective output as a model goes up. At the same time, cloud. You can see the cloud has given some what are the events at the start of World War? Russian evolution. If you see that formatting issue, have some formatting issue when compared to other LLMs, you can see A, BT has directly generated. But if you see the hajbt has created the best when compared to other. That's good. This is a multiple choice question. Germany also given in this format, but the cloud but you can see a document. You can directly go the document. I will save more time. Okay, I hope you understand. If you click on here, it will directly take you to the document. You can see here. Let's take the quiz covers key causes and events leading to World War I. Would you like to any adjustment and difficulty level or content? It's not good, but there is a formatting issue in the Bself. Sorry, Cloud. The answer given is given in the good format. But let's take another continuation prompt, provide short questions. It has generated some questions. But if you focus here, instead of providing to me short questions, it will shorter the quiz. If you see this one, so it is a big question. But when tell Cloud provide me the short quis, it has a thinker and it has generated it has shortened the quizzy instead of giving the short quotients about a topic. That's why I you see this, it has some major disadvantage for this application because it is thinking I need to short the quizzy question instead of providing the shorts about this. But this cloud also, it is no providing short questions, but they lack in the prompt shining. They lack in the prom chining that Chagp have the Gimi also have, but the cloud is lack in prompt hining means it is not generated based upon the previous one. Okay. It is not linked to the previous one. It is linked to previous one. That's why it is short in the quiz. But, it is not connected to the point of user intent when compared to Gemini and hagPty. I hope you understand these points as well. If you see here, if I use this second problem, that is, I will provide a student response. That's for the cloud. What happens here? It sure I will evaluate the student response to quarter. It has taken the World War quiz. Instead of providing the short questions, it is just a shorten the quiz question. You can think here. This AI is something is very less creative when compared to the gemian ChangePT. Let's take our first two question this I don't know. Let's take that only. So it will generate some improvements, we will check all the things. You answer suggest you need to review the immediate trigger for the correct answer is B, Q points to study. The assassination happened but it is not generated the best suggestions to improve it rectify it. So this is the lag in the cloud itself, right? But this part of you quizzes so I can't use this cloud because it is not matching my intent. Okay? So further, we need to give more context to this cloud than only it can generate some best output. But when compared to Gemini and Cachibti it has creativeness or from shining in which we can get the best output. Can see there is a result in front of you so we can say, but this World War quiz, I can't use Cloud. Because I have checked the proms, so the emigBT is ranking rather than cloud. That's. Let's jump into another I tool, which is very reasonable. This is equal to the open EI OI model that is currently have the paid version of $200. It is free one because it has deep R it will gender the response after thinking. That is great thing. Let's jump into this one and we'll see whether this win blows our mind or not. Okay. Let's take our first prompt and we'll check in the Deep sik. Let's see what happens. Let will start thinking, you can see that the best part of this EI tool, which is free one, that is a great thing. I just thinking. Okay. It will start generating after thinking it will start generating the quiz for us. Let's wait for a couple of seconds and we'll Okay. It has generated some multiple questions. Not bad. It has generated some great questions. The blank check refers to okay. That's great. There is no formatting issue when compared to Cloud one. It has great comeback. We can see how this deeps is thinking for us. You can see how the AI is thinking. Deep S is sing before generating the answer in which we can expect the best output. You can see here. There is a ten questions and it has generated some answer key for us. Let's try another prompt. Provide shorts short questions. What happens? Let's see here. I think it is starting. All the user previously as ten multiple questions on the poise and I provided there asking to provide short questions. Let me parse what they need here. It is thinking what the user is thinking, what I need to give. It is thinking in different ways to come up with a beautiful or effective output. That's a great thing about this tool. Let's say it has generative Water th. Yes it is working in the right way, you can see it has generated some short answer questions causes of World War I. We can see here, what the name. It has generated some answer keys. That's great. It is also generated some answer keys. It is followed our previous prom. That's good. You can see the ten questions here. Let's jump into our last prompt in which we need to give this. There are two prompts. Let's put here and it's thinking right now. I will provide student response for above question one. All right, let me break this down. The user is high school teacher who previously asked for requests and the causes of World War one. Now the pre sentences to question and get feedback. That's great. Thinking. Got it, please share the students response to question number one. I will provide corrections, feedback and suggestions for improvement. That's great. Let's take our response. I don't know about this question. Let's try whether it will show some areas of improvement or not. Let me see the student response to question number is, I don't know about this question. The general question was asking for the event directly triggered the State of World War I. How the is thinking, you can learn from here. That is great thing about this tool in which you can use the better prompts for this. So now you can see, thank you for your honesty as how to improve it. You can see the correct answer is the event directly tgd why this matters that's good and suggestions for improvement. Focus on key events, contextual causes, study tools. That's great thing. That is practice if unsure thing, what happened in blah, blah, narrowing the time frame, so you got this reviewing this Juli crises and primal. This is something generated according to our topic or question. So if you see here, the answer is also great. It is also explained that it has given some study tools which are very important. Suggestion improvement, not bad. But it is great thing. It is thinking before responding to our prompt. You can see the answer here, correct answer why this matters. Not bad when compared to cloud, Gemini and Changpi. It has great tool, I'm very excited about this tool after developed by China companies. That is great too. You can use for any purpose. I recommend use if you need more reasoning answers from AI, so you can use Deep Sik. You can use AI 01 model, but you need to pay for them. That is a very costly. But after Deep sik come into the market, they also made this reason button, think before responding. Okay? You can directly use this here and you can directly ask a question. It will also think before or think after. Sorry, I will think before responding. After deeps come into the market, so the open AI has started this button, added this reasoning model that you can use from here. That is all about the deep C. We have another two tools. We need to test them this application. Let's do that also. 25. 4.2.2 AI for Student Assessment & Feedback - Part 2: Okay. Let's take this. Let's mbed our fifth AI tool that is perplexity.ai, which is very powerful for the researching and gathering the real information. It has generated some ten questions which are very good, which are the following is considered as the main cause of World War I. You can see the questions. You can see the great thing about this tool is it will generate the response with the back links or with the links. If you click here, it will start generating about this particular topic in which we can get the best output. You can directly check these resources directly from here. I just break the chain here. No problem. And you can see here. Or it has generated some quotients. If you focus on here, Gemini also generated the same output that is similar to perplexity means? Why means? This is a simple thing here. The perplexity AI is dependent upon the online resources online sources. Exactly, the Gemini also dependent by the Google. What is the Google? Google also have some resources have their websites, resources, research papers, all those things. The purples AI also take the knowledge from the online sources and the Jim may also take from that only. In that case, what happens, they will gather the information from different sources. If you think this quotients also gendered in website already. It will take from that it will show to you. At the same time, Jim may also do that because it is a search engine. Gemini is not a search engine, but Google search engine, they developed this Gemini. We can say the Gemini purples a like doing the search engine method in which you can get the real time data. Instead of if you go to hat GPT deeps Cloud, they are the chat boods not the search engines. Even though the GPT launched their own search capabilities. You can use them according to your needs. This is that's a good thing. Let's directly, we will ask some follow up questions. That is follow up prompt. Next, we need to tell to AI that provide short questions. Whether this tool is capable to provide short questions or not. Let's check it out. Now, you can see, provide short questions. That's also thinking like it is shorten the quizy question instead of asking the questions here. Military is a means. If you see here, this one and cloud one have working similar in particular this thing. But these are the questions are very good because they good at researching capable days instead of going creative or instead of going into the thinking. But the Haji Dno, it has generated some word short questions with generated well. DeepC also generated directly. Gemini also generated the short questions directly here. But the cloud perplexity AI have failed in generating provide a short quis. If you give them more context about questions, you will generate it. But we have use the CA methods in all different LLMs, we are testing out which LLM is working for best according to our needs. For that, you can see here. Instead of asking some provide short questions, it has shortened the previous or it has shortened the quiz questions. That is simple meta it means love of peace like that. You can see the output is in front of you. Let's take our third follow up question, follow off prom. What happens let's let's see here. I will provide a student response and above suggest areas to improve. Okay, I am ready, please provide the student response for questions number one. Tell me if you answer is correct or incorrect. I incorrect, explain why suggest that is okay. It has generated so many things about it. Let's take our answer. That is, I don't know. Okay, let's copy our student response that I don't know about this question. Copy it and we'll move to perplexity.ai and we'll just pase here our answer that is I don't know. Let's see how this perplex dot I will give the answer for us. We can see. Thank you for sharing the student's response, here, how we can address it. So you can see it has given some beautiful answer for this question. But if you think I have given the input here, I don't know about this portion, you can see here. When I previously told to AI, I will provide a student response for above potion, one please rectify it, give sedation areas to improve it. No, ask me for student response. I will ask me to provide my response for this. I just tell to A student response that I don't know about this question. It is directly explained to me about what the militarism is and where to need to learn. That you can see, suggestions for improvement. In this way, I have to improve myself to write the answer for this particular question. This is how the AI is helping us to make to learn something okay. Indeed, we can, get improve by ourselves using this particular technology. You can see here. It is just giving to study the definition of particular militarism what is about militarism is and you can see the context learning, discussion, visual aids, practice questions. This is all some skills or suggestions that is given to me to make or to improve your this skill that is about militarism by learning in online or these various steps, study the definition, context learning. This is how AI is suggesting improvement. Now you can see the evaluation of the response here. Correctness the student indicated that they don't know the answer to this question about militarism. This means they did not provide a specific answer, we cannot evaluate correctness based on options. Option. Sorry, it is just. This means they didn't provide a specific answer so we cannot evaluate correctness based on options provided in the quiz. This is something good output when compared to other II Mrs. It has given some explanation about the particular word militarism and it has some suggestions some improvement steps like definition, study the definition, contextual learning, discussion visualis. Think that, but this is how this perplexed.ai is giving the answers for us. Again, I'm repeating the answers from this perplex.ai is depend on the models that you are using in that particular that. If you go and change our models, if you upgrade to newly version or updated version, the answer can be different looking this. Okay? So we have another AI model that we can test. We can see whether that is Microsoft copilot. That is also AI model. Let's see in which this AI model can help us to make that decision which is right to create this particular quizzes. Let's jump into another AI model that is Microsoft copilot, so you can directly come into the online Google Google and search for just Microsoft copilot and you can sign up and you can move ahead of. Let's start from the scratch. Okay? So I'm already in the chat. Let's message directly here. Let's click here new upload. Let's this is the homepage off? Let's start from the scratch. I just go to the deep Sk. Let's strike the same method in the Microsoft Co palette also. This is our main prompt. I will copy from Deep seek here, and I will come to Microsoft coopet and I will just copy and paste. Let's see what the output will be. Now, you can see, sure thing, here's a ten question multiple choice quiz and the causes of World War I, along with the answer key. That is good. You can see the quiz here. If you see there is a formatting issue, right? So it is not looking like the multiple choice is, but it is looking some sentence or paragraph, you can see here. It is just instead of putting the options one by one, it is just a combining all the options, which is not the format of multiple choice quotients that you can see directly from here, right? So not bad, but let's see what happens here. So it is directly generated the answer key. Okay, good. It is not bad. Okay. It has some formatting issues, right when compared to other AI models, right? So let's provide our questions. I hope this quiz helps you understand better understanding the causes of World War I. Let's try our words second. That is question or answer here. You can see provide short questions. Let's see what AI this Microsoft AI chatbot will think about it. Provide short questions. Settle here some short questions related to causes of World War one, that's very very exciting. You can see it is just generated some short questions. You can see here, what event triggered the state of World War one? Which countries were the part of the triple Entent you can see the sm provide short quotients. I have already seen about some I model that is cloud is not generated. This particular provides short quotients directly when we use this simple three word sentence. We need to give extra more contextual information in the cloud, Cloud AI. You have already seen that. But here, Microsoft copilot, it is automatically thinking that I need to provide a short quotien based on the previous information that I have given in this. That's not good. It is continuously working like prompt churning. That is all over this EI. Let's give some another question, prompt we have already seen in other AI models. Let's take this. I will provide a student response for above question one. Let's check it out this prompt in the Microsoft. More polite. Let's so absolutely I'm ready to help, please provide the student response to the question. What event triggered the start of World War I and I will give you feedback on it. It is directly tell me that is what event triggered to start. This is the first question it is asked the AI to me. Not bad. It is good, a way of interaction with the humane. Let's give the answer. What is about answer question is, that is, I don't know about this question. How it will improve. I will give the suggestions to us. Let's check it out. So now the ICI is thinking. All right. Let's turn around and make sure that student has a strong response. Feedback and response. Current response, I don't know about this question. Suggest areas for improvement, knowledge gap, the student needs to review the basic events leading up to the World War I. Conference in answering encourage the student to attempt an answer even if they are unsure improved response. What event triggered the state of World War I suggested answer. This is the answer. It has suggested some points areas to improve it. It has given the feedback also. That is good. So if you encourage the student to study the key events and alliances that contribute to the outbreak of World War one, they can also benefit from learning the sets of sinensi. If you think here, so it is like search engine, Microsoft Copt means the Microsoft Copalt have their own search engine that is Bing chat, Bing. It is built on the top of that. You can see the Gemini Gemini also developed by the Google itself. Now you can see here. If you see here, the Gemini and Cloud have some similar answers, if you think here, but not the similar whole, but they are differ in that. Okay. Now you can see here. It is asking some beautiful questions answers has followed all the commands very effectively. So if you think here it has given some suggested areas improvement, but it is not given in the more deeply, like purples.ai have given. So now you can see here in the purples.ai, study the definition, context learning, discussion, visual aids. This is quite deep in that in which the student can get guided to learn some particular topic, right? So when compared to the other AI model. Okay. So this is the AHR board that is Microsoft Co Pilot have given the best answer, but not effective as per placid.ai. That you can easily see here, right? So in the deep seek, you can see it has given some answer, but not on the shore, okay? Why this is one. These Air chat boards like ha GBT itself also. You can see some suggestions, correct answer. You can see some suggestion has given. Again, I'm telling it is all depend upon the models that you are using right now. If you go and ha GBT, there are different models that you can use it. If you take the Cha GBT plus, smartest small and more that you can take and the answer will be different from other. What is the main purpose of this particular prom pattern or prompt that I'm trying to explain to you is use this prompt and just look how the AI chat boards are thinking and try out yourself, take your one particular task and try that same task on different AI models that we have seen here. Just take one particular task. That is anything about it. Anything about explanation about particular lesson or topic, take that particular task, write the prompt. Just take the same prompt and paste in every AI model. That is in the hagibT cloud Gemini, deep sig perplac.ai and Microsoft copilot and just see the output from different AI modulus and check which output matching your requirements. Which output is matching your requirements and better have the better explanation about that. Then you can get that output and you can use in Wteilfe or anywhere. Okay, that is the most important using AI models. As here is here, but I will conclude it. My main focus and my main requirement is, I need to create the quiz about the ten question multiple choice quiz and the causes of World War one for high school students. This is my particular task. I have just written the prom here, I have tested in all the six different AI models. Here. So when you seeing this video, then it may be another AI models out there in the market. You can use all that models also in which you can get better insights from the AI models. It is not about you loving the particular AI model. It is all about understanding the AI capabilities and limitations that can help you to get the best output from it. You need to stay updated in this AI technology taking or by testing out each and EI module with your similar task. In which you can choose one particular task for different task. Sometimes your task is need some other II model rather than Charge PD, rather than cloud. Whether we don't know which particular task is solved by EI model. They have their own capabilities. For that, you need to test it out the same particular task in all different EI modules to choose to choose the perfect AI model for your task based upon the output that matches your requirement, that is simple. That is how you can use AI models in different ways for your different task. Okay. So in this particular task, I have decided I have decided which module is best for me that is per plesit AI. Why? I have already explained to you because it has given some valuable points, right. It has given some explanation about it and the correctness, that it has given more good points, good points which I am thinking, which is right for me. Okay? That's why I have chosen the purples.ai output, which I have just impressed with this output. Okay? So there is a more good have the purpose dot have capabilities. Why? Because I have already told you you can as such research this research online, anything. You can research it anything online with this particular AI module that is perplexi.ai. You just tell to AI perps doti just research about some particular topic and give the summary. It will go and it will take all the references, websites, all those things, and it will give the best output for you, which is a lot more time. You can directly search images for this particular topic. You can search videos this particular topic. I can generate image of. All those things it comes under. The main part which the main thing which this purplesy dot is standout from other Rs is, this is the follow up questions. You can directly click here. It will take to the follow up questions which help you to expand this particular answer or question. It is good. You can use this purplesy dot A to generate any particular quiz because it will research the papers, all the particular websites, all those things for you instead of you and checking you can instead of you searching in the Google or anything. It will just tell you and it will summarize you and it will give the best output for your requirements in this particular quiz. Okay, I somewhere, if you're looking to generate a story, all those things, so you can test it out particular story prompt in all AI MLs and check it out which story which matches your requirements and very creative. Okay, you can check all those. Okay. So I just with my experience, I will just summarize you which I model is best for it. So when compared to HGPT it can generate a best storytelling or creative story, poem and content also, that you can use. When compared to Cloud, the cloud is different from the HGPT and other I models. It will help you to in the technical part like any coding basis or analyzing basis, right, that need to be analytical mindset, in which you can use this cloud. This is good best choice. Gemini, I can saying the Gemini also the best I model because it is developed with the own Google itself. They have the lodge dataset in which you can get the best output, right? So that is not bad. Gemini have their different models that have launched 2.0 flash. So it is all about if you use the latest version of AI models in every AI models, you can get the accurate and good output from the AI. I hope you understand these points. Jimi also a good search engine or you will get the best. You can directly summarize a particular video articles or you can just use this IM as you will use hagibi Deep si again, I'm saying that deep seek is a game changer HR word in the 2025 that we have right now, that is a good. Why? Because this particular hagbtive functionality is same as a hagibt open 01, which cost you 200, dollars per month. This the same particular functionality the Chinese company that is Deep Seek is providing for free to you. You can use this for the analyzing purpose, you can analyze it thinking, reasoning. If your task have some reasoning part, you can use this Deep seek for yourself. That is very best HR, but I have never seen. For the free, not I'm talking about the paid for the free. Okay. So you can directly just go and search it here, just click here and you can search it in online. It will give the best answer, but researching and thinking. The hag also have their own reasoning and research capability. They have added the button after the Deep Sik has launched. That is how the AI war is going in this world. Okay. That is all about eats. You can use this for analyzing purpose or for reasoning purpose very well. Okay. And as I said, the perplsidt is great for researching online or to get the best insights instead of going to search by yourself the different websites. You can come here perplest AI. You can ask anything and it will go and it will refer you the references in which it has taken content to show you. We can directly ask follow up questions. You can go check it out there, what is references from here where it has taken the quiz, you can directly click here and you can go to the same particular papers or websites that it has taken the content to show you in which you can trust this particular output. That is purpose doti. And our last AIM model that is Microsoft co pilot, it has great, but there is a formatting issue. You can change this formatting issue by providing the example how the multiple choice quotients look like to be. It will just learn it from it and it will generate the best output for you. This not up at all. You can see the best output from the AM rules also. You can use this AI model for better. But in this particular task, creating for the ten quiz quien multiple choice quotients. I have selected the purples.ai because it has good research capabilities in which I can think it has some 100%. Not 100%, but I can trust it because it is taken from the different sources which has already published in the online. That is best part of this particular EI model. I hope you understand this particular application that is creating quizzes. Now let's start our second point. 26. 4.3 Interactive Learning with AI: Okay, let's start our 4.3 that is interactive learning with EI. Not only we can use to generate a content quizzes. So we can use this AI in different manner like interacting your students with EI. It is a great choice to improve your students' critical thinking capability. So how we can see that let's jump into here. So how AI makes learning let's say students allow interactive experiences, even your student can love you interacting with the teacher to improve the engagement in that like asking a doubt in the particular topic or solving some particular question from your side like that. In that way, the EI can help you to create engaging classroom activities. As a teacher, you need to create some engaging activities that the student can relate to learn something from it. For that, you can use this AI model in the main three ways. Not only is limited, but you can use in different ways. But these are the three common uses, role playing with AI, creative writing prompts, real time language practice. These are the three common use cases that you can take the help from AI chat boards that you can do in the classroom. You can see the first one that is role playing with AI. So have students interview on AI historical figure. So for example, if your students are lack in some particular topic or subject, you can use the AI chat boards by trying the particular domain. We have already seen some act as a personal prom pattern in which we are assigning the specific role to AI to do the specific task in which you can try AI model. Try and just cha gibive that we earlier discussed? Just tell to historical figure and make use your students to give the answers for the interview question from EI. We will see in the practical after a few minutes. The second one that is creative writing prompts, use AI to inspire students stories, write some prompts which generates the better stories for the students who lacks in particular a skill set or particular thing. For example, if student have some lower confidence, you can write the best story which inspires that particular student who have the weak confidence, they can inspire it and they can improve their confident level. That is we can use the creative writing proms, that is questions or anything, real time language practice. As we know, we need to learn some languages in our Dailif or work life, we need to know how to talk one or more than multiple languages. So in that, you can use this AI chat boards to learn some different language based upon your current language. Okay. So you can use this, Okay, real time language practice that we can see. Okay. Let's see what is the first example we are looking to use here? So prompt, we can use AI as a virtual debate partner. That is, for example, if your students lacks in communication, communicating in English or any other language you will take. So just any human being will only learn the language by practicing or by communicating with the person who have the command on this particular language, in this case, you can use AI. In this case, you can use AI for particular language that you are looking to explain to your students or that you are looking to help your students to learn this particular language. You can try that particular chatbot at the particular language using act as a person of prom pattern, for example, act as a French language expert. Now you are here to debate some particular topic that you can see the example here, pretend you are Albert testn instant and argue for or against the use of nuclear energy in a classroom debate. Now you can use instead of Albert test, you can use any name. You can use a French language expert, so argue against the use of English language in a classroom debate. That is all about how you can use R prom that is all about. You can use AI as a virtual debate partner in which the AIs one hand side and your students or other hand sides in which they can communicate with each other. They can debate with each other, in which they can learn some communication skills, right? They can improve their communication skills or language skills that they are lacking there. Okay? That makes engagement in the AI and human beings. That is a good thing. The students can practice debating both sides of an argument with an AI generated persona. Personal means that we are assigning some particular role to AI that you can see here, RA, but and Stein. That is simple. Let's see some example number two, that is AI for language learning. As I said earlier, so you can say to AI, if you are a French language expert, no your task is to communicate with MI students from particular topic. You can give the topic to AI as a zio student. They will each to the students and the AI will tap with each other and they can improve their communication skills, collaboration skills like that. Remember this AI is we are not particular robotic we are not making just you are using chat board. So chat boards have their own Voice modules. You can directly switch on that particular Voice models and you can tell to your students to talk with them. Automatically it will generate some or put in the chat itself. They can interact with each other. Let's see the practical. Correct this English sentence and explain the mistakes you go to school yesterday. Correction, she went to the AIs generated some suggestions about two mistake correct the English and the studies. Because she goes to the school yesterday you can see the correction. The fun challenging as student to debate AI's response or critic its explanation, this Pelsy critical thinking skill. So that is this AI is powerful, how you can use this AI module with the students to improve their critical thinking capabilities and language or communication and collaboration skills. Okay? Let's check our first example that is AI as a virtual debate partner, but that I'll just Okay, I'll just copy this prompt here simply from here and I will choose hag first. As I said, some AI models have their own Vice mode. You can see here, own Voice mode I'll just tell you Iseler said, you need to break the above chain for that we need to write from now from now from now. Don't Don't follow a previous Don't follow previous instructions. It will help Sagebi to break the chain above there. Now I will just paste my task and let's see what happens. So we can see it has started the debate. All but Einstein in character, greetings, my curious, young minds, the topic of nuclear energy is one of the is one that requires deep contemplation for it holds both incredible promise and places. You can see the debate has started. This something about. For these particular words, we can use this voice model or even you can use the chat functionality. For your students, you can use this voice model to speak to speak in the debate itself. In case I have used this or you can continue all these things. You can take short start there Nice to meet you. Like that. I have just canceled it. In your case, you can use this voice model and you can check it out. Otherwise, you can use directly chat itself like I am not familiar with these points or you can start this any debate like that. You can use your own students' responses and include here and see how the AI is generating the response based around the student debate answer. I am not support your words. Let's see what happens here. My dear friend, disagreement is a spark of weird learning. Share your thoughts. Why do you not support my words? Perhaps you see something that I have overlood? Let us debate, but through our exchange of ideas, we both stand and grow in understanding. That is how this debate goes on continuously, right? So you can see here. You can write your students response here, even you can use. This is the simple chat functionality. They have their own is modules that you can use. So Gemini have their own AI voice model in which you can directly interact with AI voice. Hagibr also have their own voice model that you can use that you can just tell to AUR some particular assigning role, your topic, and your debate and just keep your students engaged with the ER. That is simple. This is how the EI transforming the education also. Okay? So as I said, you can use this particular this question, all those things in different I modules to check whether which AI module is performing well this particular debate, right? I'm not showing again here, it will take time, but I am giving you with the assignment for you, just take on particular topic and assign that particular personal pattern and give some particular topic and just keep the debate is students and AI model. As I said, Gemini also have their own Vice models that use microphone. You can just tail to hear here and it will start generating here. Okay. You can use this voice use Microphone button to integrate your voice. This is how you can use this. As I said, the DeepC doesn't have their own voice integration. It is newly Ag Adbd we can expet in future. The purples dot A also don't have. But the Microsoft Copalt have their own voice button in which you can start debate the students and I models very effectively. So you can see Hey, save you back. What's on your mind? Better when I models. I am listening. Oh. Okay, that's good. But they have some problem in that, okay? You can use this voice model. You can check it out this voice button also by yourself, and you can start chatting with this particular AI model. So the main purpose of this particular and task is to just debate your students and EI, to help your students to grow their collaboration, communication, and debate skills, right, and negotiation skills or to make something to learn some critical thinking skill. Okay? This is how you can use AI to help your students to grow their mindsets, skills, and much more. Okay, I hope you understand this example very well. Okay, let's jump into another example in which we can see the correct this English sentence and explain the mistake should go to the school. This is something the simple question, right? So let's start. Let me see what happens here. Correct this English sentence and explain the mistakes, she go to the school. So you can see the corrected sentence. She went to school yesterday. I is a simple task. Okay? You can see her explanation of the mistake, verb tense agreement, all those things, it has given something. That is good. Right. So you can use all those things again and again here, just paste here in the cloud also. You can see the mistake in the verb form, this needs a past tense because now you can see the explanation, right? If you see, for example, if you see here, the Ajita has given the direct answer. She went to school yesterday. This is a great effective output when compared to Gloudt AI. It has given a correct answer, but she went to school yesterday, but it has given some explanation, right? It also gives some explanation. These are the two models are given the best explanation about this. Now, you can use in these different AI models as well, right? So as we discussed how we can use this particular AI model. Okay? So you can use this ask students to debate their responses. You can directly just copy this. Okay, let's go to HAGPT and it'll paste here. What happens? Ask students to debate AI's response or critiques explanation. That's a fantastic idea. Encouraging students to debate or critics air responses can spark deeper engagement and promote critical thinking. Here's how could approach it, okay? So you can see activity plan, the also generated some activity plan in which you can use this to start out the debate activity. Okay. So you can use AI module in different ways or in a different scenarios and in a different task, right, there is no limitation for that. It is all about how you think and how you write and and which prom patterns that you pick. And one of the most thing is which I module you are selecting for particular tasks that you are looking to do with EI. Okay. For that, you need to just take the similar prom and test it out in other AI models as well to check whether this particular AI module is matches my requirements or not. Okay? So just try it out and select one particular AI module, ok and just go deeper in that. You will get the best output from the AI module as well. Okay. So just use all the prom patterns in different I models like Gemini, Cloud, deepsk, perplexidtE, Microsoft copilot, and just use this AI in your work, daily life or even in Daily live. Okay, professional or personal. That is not about. You need to use these AI models in effective manner because in future, we are looking AI is transforming each and every industry, not only the education, not only the tech. In each industry, the AI is just transforming all the things. Okay. You need to adapt this very fast evolving skill that is using AI models or effective way. Okay? So you can use these AI models in effective with the help of prompt engineering only. Okay, so now you can use the prompt engineering that we have earlier discussed all the prom patterns. So just recap that and use some particular task, choose a particular LLM for your task, and just go deeper in that. Okay. I hope you understand this application, right? So let's dive into another other part of this module that is differentiated instructions with EI. 27. 4.4.1 Differentiated Instruction with AI - Part 1: Let's start our 4.4 that is differentiated instruction with AI. What is an differential instruction? As we know in every class, there are bunch of students which have their learning capability is very good when compared to others, nothing but the class is filled with so many students, which they have the great mindset to learn fast or some students have a less capability when compared to other students. That's why it is called class, right? So how we can help those students with AI to improve their capability or to improve their quick learning capability, right? So we can see in this topic, right? Let's see what is differential instruction here. So every student learns differently, right? That is a main point. Every students have their own mindset in which they can adapt or in which they can grasp some type of information from anybody from any teachers or like that, okay? You can help customize lesson materials to fit different learning levels and styles. As we said, for example, if a teacher takes responsibility to categorize learning material for the students who have their less capability or by categorizing the people who have the great mindset and doesn't have the low mindset. If you take this responsibility, it takes a lot more time to categorize and to make the materials which fit for the different levels of students. To make this to save this time or to make more easy we can use AI, the power of AI in which it will automatically generate the lesson plans or materials to fit different learning levels and styles of students. That we can see in this topic. Let's see how to adapt to content for different learners. Let's try proms to generate learning materials for different student needs. Here we are taking two use cases that is for struggling learners for advanced learners. As I said earlier, students have two types of students which are very brilliant to learn something to adapt new technology or to adapt a new lesson or chapter. But at the same time, the class have the struggling learners which don't have to understand a topic as much when compared to advanced learners, right? So let's take these two types of students in which we can see how to write the proms further, and we'll see how the AI can help in this scenario. Okay. Let's take for struggling learners. So what is the prompt? This is a simple prompt we are using. So as I said, you can use different prom patterns that we earlier discussed in the chargeb or other AI models that you can use according to our but I'm just giving you the idea about how we can integrate AI into our classrooms or in our teaching field in which you can get the benefit as well as you can help your students to move faster and smarter. That is simple. Prom explain the water cycle in simple words for a fifth grader. As I said, for struggling learners, for example, if the student have the seventh grade class, they don't know if that student is struggling to understand this particular topic in the seventh grade. At the same time, if the topic is in the lower class in which the present seventh grade student can easily understand because it is the content is made in the fifth grade, in which the seventh grade student can easily understand. So for that we can use the fifth grader or we can take other lower grade, less than the current grade. As I said, if the person in the student is seventh grade, so we can take topic and we can tell to AI. To explain the words for the fifth grader. In which it will generate the content or topic explanation in the format of fifth grade student that it can easily understand, that the student can easily understand, in which you can explain this particular AI generated topic to the seventh grade student in which they can understand easy. Okay, so that you can use this type of simple fifth grade system or you can use any other prom patterns or any other thing. Okay? This is a simple prompt I have just used to explain you. Okay? So you can see this is simple output. We will see how it will works in gebe and other A models as well. This is for struggling learners. Let's take it for advanced learners. What happens here? For advanced learners, they have their own capability, ability to understand things. They have good grasping power to understand and to adopt a new topic or anything. For that people, you can go casually, how you will explain or how you will just casually explain the topics that you normally do in work daily life. That is simple. You can use prompt here. Explain the water cycle at a high school level with the scientific terms. If you see here, for advanced learners, you can take you can tell to EI, explain the water cycle at a high school level with scientific terms in which you are helping the advanced learners to move further ahead. Move ahead from the current level to higher level because these advanced student learners have their own ability to adapt to new technology or new topics in subtin level that is high school level because they have the grasping power. They have a good understanding capability to learn to move forward in the challenging mode, you can see here. For the advanced learners, you can tell to A, expand the water cycle at a high school level with the scientific terms because this is for advanced learners in which they can easily understandable and easily grabs the knowledge from the AI or teacher. That is how you can use the EI. You can use this AI for to explain or to help your struggling learners students and as well as advanced learners. You can use to explain AI or we can tell to explain some particular topic in high school level with scientific terms for the advanced learners. At the same time, you can tell to AI just explain this particular topic in the lower class to help you struggling learners to just learn and grow themselves. That is simple how you can use this for advanced learners. Struggling learners. You can see the protein A can generate visual aids like summary chart helping students understand complex studies. They have so many I models out there and day by day so many creative and advanced I models are coming out in the market. Previously Deep seek Mistral AI, we have 0.2 0.5 that is by China Alibaba Company. There is so many advanced AI modules are coming right now. This is how you can use the prompts. This is not about what the models we have, but it's a prompt engineering. It's something is different. This is a prompt, this is a foundation, how you will use AI. The model that you are using it is it is secondary, but writing the art of proms is the first skill that we need to learn in this I era because if you know how to write the prom, you can use any AI module in the effective manner. That is why we are learning this prompt engineering. That is simple. As I said, let's jump into our Char GPT and we will see how it will works in practical itself. I'm coming to chargeb here. As we are Alla discussed, you can use different AI modules. You can test it out. This same particular prompt in other I modules like cloud.ai, gem.ai, perplexi.ai, and deepsk.ai. And we have another model, Microsoft copilot and not only that, we have other I modules which are advanced, you can see here, Quinn Do W 2.5 lets you can see, this is a new AI model that is generated by developed by the Ali Baba company that's China, you can use directly from here, we can see here. You can just this is a simple Es we use such GPT and other AI models. We also have other models like mistrl AI. Let's see. You can see here mitral this is also some chat board which is similar to other things. You can see directly chat, talk to, that is chat. We have another that is Quinton Mistral, and we have got new new AI, that is the mass company that is X, that is how these I modules are coming day by day, right? So now, you need to just know if you know how to write the better proms for your requirements, that is enough because you can know how to write the proms because for a particular AI module, you can test it out with your proms. Based upon the output from AI module, you can choose it. It is not about mastering the particular top particular AI module. It is all about mastering the prompt engineering here. So this is not in, for example, in future, there are many I models can be developed. That is not about this prompt engineering is not about mastering the II two, but mastering the write of, uh, attracting prompts, Mastering the writing prompts. Art of writing prompts is a main skill that I'm telling you. I will just hit out this particular prompt in this newly three models. Let's see what happens in these AI models also. I am not attesting the same prompt in other I models. Previously, we are using Gemini, Cloud, purple dot EI Microsoft Covit depsik that is we have already seen that you can practice with yourself. Now here we will see the hachPtqin 2.5, mistalEI and Grok. Let's start. Our simple requirement is just copy this Okay. I just come to Chagp We have already seen this jib can answer very well. Let's take. Remember, this is a simple prom that I am using, right? So in case of you, you can use different prom patterns, advanced prom patterns we have earlier discuss like potion refinement and cognitive verify, semantic filter, you can use all these prom patterns if you are well or if you practice well of those prom patterns, you can use according to your needs and requirements very well. I'm just I'm telling you how to use AI models for your requirements. This is simple, I have given prompt for ChargePT that is expand the watercycle in the simple words fifth grade. It has simply generated output you can use for your student to explain it. You can tell you can use this search button resin. That is all we have already seen. For the same thing, we will test let's test out this new Now you can see the quint 2.5 Max is giving some beautiful answer for us. This is some new AI model. That is you can see, sure the water cycle is like a big loop, that is not good. You can see the Chachi pit and this ice generated same evaporation, condensation, precipitation, collection repeat. Now you can see the collection precipitation, evaporation. Now you can see the AI models are exactly similar in producing that answers, but not equally accurately, but they have around themselves because the data is same for the particular question. In that case, you can expect the same output from the other I models as well. So this is simple I am using. Let's take this mestlNw AI that is developed whether I think Germany or like that. R Mito letter so we need to accept it and you can see we are here. Now, you can see sure the water cycle is a big round trip that water stakes. Now it is also explaining evaporation, condensation, precipitation, collection or runoff repeat. If you see here, this output and hagibs output is actually same. It has taken the repeat collection, precipitation, condensation evaporation at the same time, this one, that is, you can see here as repeat oft is generated. If you think when the multiple AI models are generated in the same output, then you can say this these two outputs are effective or some accurate because the two different AI models are trained by that large amount of data in which we can say that this particular output from these AI models have some accuracy because there are multiple I models which are generating the same output. We can surely say that this is correct. Let's take our newly developed IHRbod that is Crook. Let's check it out. I will just paste this prompt here. Now, you can see it is very much fast compared to Char JPT and other A models. You can see how much time it is taken generated in simple seconds. That is more powerful, that is group X one, that is developed by the arms companies. You can see it is fast when compared to the AM models. You can see it is also generated evaporation, condensation, precipitation, collection transpiration that is good when compared to all other things. If you think here, it has generated good right. Let's check it out some other thing. I will just write a follow up questions. For example, the above, let's take the above output is not understandable by seventh grade student. Let's let's take fifth grade. Fifth the students. Can you explain the terms of story? Let's take what happens here. Chagbt is great at crafting the output in a creative format or story. It has great capability. Let's see what happens here. You can see Sure let me explain the water cycle as a fun story for the fifth graders. You can see once upon a time, there was a little droplet of water named Wendy. It has generated some creative story that you can see here. Now you can use this particular story in your daily teaching that you can direct and explain to our students to learn this particular topic. That is insane knowing how this is II models or transforming I teachings. You can see the moral of the story. If you see here, the output is very good. Let's try it this same prompt in other I models. What happens here I'm here in the Quin 2.5. Let's check it out this also. Going to buy. Of course, let me tell you the fun little story to explain the water cycle in the way that is. Understand the adventure of dropping the water drop. That's good. Once upon a time, there was a tiny water drop named Dope that's good. When compared to the hag B, it has some great formatting, that is the adventure of dropping the water drop. That is good when compared to this one. Let's see. If you see the Hartford drop, now you can check it out. This is good, right. 28. 4.4.2 Differentiated Instruction with AI - Part 2: Can see here, this is a good story. Once upon a time, there was a tiny water drop named droppi. If you think here, when compared to Cha GPT, you can see here if you see this particular story from ha GBT, it is some effect too. I have some advanced terms in which we cannot understand easily in which the students cannot understand asidy. But when compared to the Quin 2.5, this AI model is very generated very easily understandable story that you can see here. Once upon a time, there was a tiny water drop named Dropi it can engage this when compared to hj you can see here. It is liter drop often named Wendy lived in a big shiny lake. So it is something clergy, when compared to this one. But this story is increasing engagement with your students. You can see here, you can understand this. I think this quint 2.5 max generated the best story when compared to ChargeP. Let's try this same prompt in the mistralEI. Let's see what the output will be. Wow, it is very fast. Let's see. Once upon a time, this is a story generated by the mistral AI. Once upon a time there was a little water dobletnam Drip lived in a big blue ocean with all his friends. Once a sunny day they warm Sunshine a drip. If you think here, this story is also good, but it also have some effective words which the fifth grade student cannot understand. Fell back. If you see from the starting point of view if you read this particular story, you will understand. This story is something have effective output or effective terms that cannot understand easily. It is in my case, it is not feeling that is I can get the engagement in this particular story. No, I will just tell two quinto pin files better when compared to mistralEI and a GPT for generating the story. Let's see another AI model that is grok. As you said here, the particular chat is gone. Okay. Let's see what happens here. Das generating right let's see one thing. So we need to come to first because we are not sign in to replicate our previous output, right? So for that we will just write, we'll combine these two proms here, exactly. Let's check it out how it works. This one, and we'll come here this one. And we'll just paste before this. Okay. Mm Again, we'll try this. That is generated the previous one that is vibration, condensation, precipitation. It is well. Now, I will just follow up question that is explain, let's take this one. Write this one, where is this one. C here and just waste. The boy is not. Once upon a time, it has generated the output. Once upon a time, in a magical world, there was a little drop of water named Vale. Valley lived in a big blue ocean with lots of friends, but he's wanted to see the world. If you think here, this particular story, this particular story is something similar, you can see here once upon a time, there was a little water doblet named Drip Lin lived in a big blue ocean with all his friends. Exactly, you can see here. Magical world, there was a little drop of water named valley valid in a big blue ocean with lots of friends. If you think here, this story also have the great things when compared to but in the short and sweet words, I think the q 2.5 is generated the best story if I think. You can read this particular story, how well it is written when compared to the other II models. This is how you can evaluate the particular output from II models, different AI models. Simple, it should be matches your requirements and it should be filled as a human touch, okay? Cannot take the particular output and use in your daily life, but you need to evaluate output, and you need to refine the prompts as well to get the things from AI models what you want. That is simple. As a conclusion, I will use this quenchat for generating the stories for my deli classes. That is simple. How I will use this Because I will test it out the same prompt in other AI models to check it out whether which III module is perfect for my requirement, in which I can use this AI model in daily life. That is simple. That is how you can use these I modules in effective way. We have seen how we can use these II models for different ways and different things. As I said, just use these II models or use your same particular proms in other AI models. As I said, we doesn't use other AI models here like Gemini Cloud AI, perplexitEI in this particular application. That's why I am giving to you the assignment for this. Use this particular storytelling particular topic for struggling learners and advanced learners and use the prompt and test out all ten different AI models that we are up to now discussed. What are the ten AI modules HGPTGmidtI, Cloud, perplexit AI, deepsk Microsoft copilot, and Quenchat Mistral AI. And grow. Just go and see how well the output AI models are right in the market right now, and you can see other II models also coming slow by slow, but we will see how these EI models can change the world in upcoming years, especially in education. This is how you can use AI models for different aspects or different requirements of you. So as I said, please just refine your prompts, right? Use different applications, different AI models, test it out with your prompts. This is how you can use this AI chat boods in your Di or Worklf. Okay. I hope you understand this particular application. I'm not explain I already explain to you. So just this is assignment for you. Practice by two different proms. That is one for struggling learners and for advanced learners, write the prompt for them, use the same particular prompt in other AI models we have discussed right now up to now and check the output of every model and then decide only one chatbd you can go in deep with that particular one chatbard and take the best output from that. That is simple how you can use these proms in this AI era. I will repeat again, the prompt engineering is not about mastering a particular AI to but mastering the art of writing prompts for different AI models in which you can get the output from any AI model, that is up to you how you can write the AI models. For a deep explanation of about each and every chat board, you can search in the YouTube or anything that you can unlock the same particular chat board more capabilities. For example, in quenchat you can go to ART fax web search coming that is image generation, you can directly go here, same the mistrLEI have their own capabilities that is web search, file upload, Okay. And the grok also have file upload, you can enable the research option. You can share all these things. The different AIM model have their own different applications or different capabilities if you see the HAGPT, they have the search button reason button. Okay, you have the using Vice mode. Okay, different AIM moodels have their own integrations, all those things. So for better to master or to use the IIatbard in advanced level, we can search in the YouTube or online resources in which you can get the best insights from there. So up to now, we have learned how to write the specific proms for specific application. We have discussed a different prom patterns and we will see the simple applications that we how we can use this AHR words in our teaching life. As we have discussed only a few applications, but the main purpose of this course of this class is to give the awareness to teachers who are changing the student's life, that they can use these particular I models and to unlock their potential. There is simple awareness class and of course, this is. This is not limitation, but you can use in other ways or different ways, as I said. This is up to now, we have learned some different II models, up to we have seen the nine different AI models, how we can use this and how we can see. In upcoming second model, we will see how to different tools and platforms for teachers to move forward with AI in upcoming years. I even you can save you a lot of time. We will have the great JGBs own AI tool that is playground AI, open AI playground. We will see how to use this. We can make some assistance applications instead of going to HGP. We can just once we need to once we need to try an AI model, that is not a technical part, they need to about coding. No, no, that is not. This tajibi have their own playground in which we can specifically write the proms and it will only act for a particular instruction only. It will not go out of that. That's tajibi and M models do. They have their own open a playground. We will see in the next model all those things in the upcoming classes. Okay? So I hope you understand this particular model, and we will dive into another model that is tools and platforms for teachers using AI. Let's dive into that. 29. 5.1 Exploration of Google Ai Studio Platform: Okay. Congratulations. Up to no if you understand all the prompt patterns and you practice it well with a different use cases, then you are great, right? Because up to no, we have seen different prompt patterns that helps me or you to understand the lens, how we can get output according to ours. So as a teacher, we know we can use the different lens for different use cases like generating the quizzes, quis, answers, and much more in the previous sessions. So in this session, we are going to see some different tools and platforms in which we can save a lot of time, okay, with the latest AI models from Google, ChR JBD, Open AI, and other as well. Okay. Let's start by one by one. Okay. So right now, I am here in the Google AI studio, which is best, okay? I'm not going into the technical bit of that, so I will just giving the awareness of this particular platform in which you can do more things, right? So if you want to go in depth, you can go to YouTube and you can search for it. There are a lot of more videos on it. Okay, so let's start our first platform that is Google AI studio. Just go to Google search and just type Google AI studio, just click here and you can get the first, right? So just come here. You can come and after successfully signing with your account, you will come to here. This is Interface. What you can do with this Google AI studio? You can do anything. As we know already we have seen the different LLMs. So we have right in the different proms patterns, and we have seen more than eight different LLMs Cloud, Cha JBT, Gemini, and much more. So right now, let's see this platform, what the platform is. It is basically developed by the Google itself. The Google also have the Gemini platform. In this Google a studio, this is combined with all the tools like Gemini and write the Gemini Live and generate videos with VO. It is a combination of all the different tools by Google, and it is combined in one place in which you can trade all in one place. That is great thing. You can build your own apps here, right from here, Starter apps, Sam chat, you can do with the chat like you have done with the Gemini in the previous sessions. We can select the models here different models which are the very reasoning and more latest models which are performing very well in the market. What is about Token count means if you write any quotien let's say what is what? Let's take photosynthesis. We are talking with the g.ai. It will give the answer for us. Which response do you prepare? Let's take anything more here. Let's skip that and it will start generating our answer. No, you can see. It is explaining the photosynthesis. It simply works like a chatbd only, you can see. What is the token? If you see here, it has change it 0-1604 topens. What is the token? Token means it is a combination of all the characteristics, right up to this, the whole token numbering is 160004 topens. You can get the free topens I think one lab, 48 tosen 576 topens. You can get the free. You can try with the different elements. Sorry, you can try with this Google AI studio. Not only that, you can control all those things here. If you increase the temperature, the output will be changed. Not only that, you can enable the different structured output tools for execution if you're looking like that. You can also include the Google search in directly here, in which it can help to generate all those things. Basically, from this Google AI studio, we can build the apps and much more things. That's why it is good. Basically, it works like a Gemini platform only Gemini chat that we are already discussed in the previous session. You can go the advance tool settings, all those things. I'm not going in the depth, why because it is a technical bit. If you want to learn more about this platform, you can go search it in the YouTube and you will get the best videos. Not only that, you can control your output tokens. For example, if I just to tell you I I want the output in the 500 tokens only. It will control and it will give the answer in the 500 tokens only in which we can easily control our output to use the correct and more cost effective tokens. Let's say what is photosynthesis? You can see the output will be in the 500 output tokens length. Let's see that. I just giving the answer for us. Why? Because the different platforms have their different training models, all those things. We cannot tell. No, you can see here. We have tell to AI, use the 500 tokens only. I need the output in the 500 tokens only. In that, you can see here. It is used the 494 tokens. Okay, like that, we can control the outputs, all those things from this platform from here, right? Okay. So you can customize this at stop, settns, all these things. It is a technical bit. So as a teacher, you do not need. But if you want to learn more about this platform, you can Google it and you can learn more about this. Let's go to our stream number. What is the stream? Stream means you are talking to live. So if you though install the Gemini app from the place to, you have another option. You can directly talk with the Gemini like Live. It will give the answer like you are talking with the Women. You can check it all those things in from there. You can talk directly. Let's example. You can talk with us. Hello, how are you? We can type from here or you can tell to here. You can see it is giving answer. So now you can see here it is talking with us, right? So you can check in at all those things. I'm just giving you the awareness that you can try by yourself for the more things. In this platform, we have the great future that is we can generate any video. That's why we can generate all those things. For that, you need to all you need to give the access to all those things the account you need to create. After that, you can generate any video, for example, I will take the small is playing with care. If I just hit the Control button, then according to our prompt or text that we have given previous, it will generate the video. You can seeeFail to generate video or prompt due to safe. We cannot write some restrictions proms here. We can take another thing. Small kid is playing with the wall. Let's see what will show. It will take time. You can check it. You can change 16 by nine ratio, 19 by six ratio, all those things. We can check it out how many results you want and what is the video duration 8 seconds, 7 seconds, all those things. According to our settings, the tokens will also take from it. I hope you understand these points, Let's take another like ice cream. Calls from Sky. Let's try this. Why? Because some AI models have their restrictions on the prompt or generating the videos, you need to be aware of that. I hope you understand these points. It will take in time. That's why you can change the models of there is only one model, in case in future, there are other models as well. It will taking time. From here, you can directly generate the video or if you have any image or prompt, you can just click here and you can upload from Device, MyDrive or camera sample as well. You can generate the small kid videos, poem videos from this particular platform, right? You need to use these I modules to make your time save to save you time or to become more creative or to create something. You can see here. This is a video, let's play. No, you can see the teams are falling from the sky. That is very great. That's why you can generate the videos, this all the free month. There is no cost in that, we need to make sure that we need to see the tokens. You can use 148000 tokens for free. You can generate the Videos, stream, chat, like that. I hope you understand. You can get the history, what you all done with this platform, all those things. There is much more to explain with you, but I'm not going into them very detailed with that. As we know we have already learned so many I models, so you can use line that only. Let's go to our camera. You know that we already know the Cava I think you already know about cava. Cava, you can build anything like presentation, whiteboard, talk, YouTube thumbnail, social media posts, all those things. You can generate all those things. In this Canvas, there are other AI features that you can use. Okay. Dream lab, you have the dream lab with the AI, you need to tell it will generate the best AI image for you like. You can see there is a credit system. So to get the motor system, you need to upgrade it. There are other free AI image generation models like Ali eonardo AI Lexica. You can try at all those things. So the best thing is to test it out. You want requirements in every Aimage models. You can see that which image AI model is giving you the best exam, best requirement or best output which matches your requirement, then choose that. So already we have learned the choosing the best AI model for your requirements in previous session. Draw for the AI image models also to get the best output from the AI models. Like, you can directly give here any prom. Let's kids are playing in Got it. You can directly create the simple images from the camera itself. Okay. So it will start generative. You can directly you can change the ratios, you can change the style. You can add the image and you can tell. This is the output of this very attractive, right? So you can use this camera right for the image model. You can get the free 20 credits, you can use it, right? Not only that if you know all about camera, that is good. So there are much more very creative with this cama. We have much more gets more from this camera. We can try it on all those things. Another platform from the Google, which is very incent that is Notebook m. 30. 5.2 Overview of NotebookLM, Canva and Ai Prompts App: Notebook M, which is a very best for the students or researchers or the teachers like you to get the best results in the less time. Let's see how it is. Just come to Google and just write Notebook M and just good. But the first link and you can click it on. You can come to here after the account sign up. You can from here, come here and see you need to add any one source. The source can be any file, or you can give any website link or YouTube video link or Google Docs, Google slides or copy text. Just copy the text and paste here, right? We can paste here. Simple. That is simple, right? As I have already uploaded some I have already uploaded, pasted some text in it. Now you can see it is a text that is personalized anger management mindmap. Now you can see here. From that, it will it will give three actions. What you need to do? You need to add the note or audio overview or mindmap. You can generate anything. That is, up to you and you can see here add note you can convert into the study guide or briefing doc, FAQ timeline. Let's say example, I have uploaded some texts about personalized anger management mindmap. I have just clicked the mindmap. It will start generating the mindmap according to my according to my text page on this. Let's click here. It is generated a beautiful and very effective Mnm for me. Now you can see here. That is good? That is how you can use this notebook LLM to create the test to create the best mind map for your students to explain the complex topic or the topic which have some complexity to explain or to understand. No, you can use this mind mapping feature in the notebook L M to explain your students in very well. Which is a very best thing in this AI platform, which is from the Google itself, okay? It's saving the more time to create the mindmap. It is in single seconds. In second, it has generated the best mindmap for me. Selected text source, no, you can generate the FAQ how did the FAQ will Q I can see here. It will give the potions from the text that I page in this model, and it will give the answer for it. It is best when you are creating the potions and answers for your students. It is the best. Just give the topic to this I model and just add it to the source, select from here and just click here FAQ, it will generate the question and answer as well, which is saving time for a teacher. As a teacher, you can save a lot more time. Not only that, you can convert this text into study guide. If I keep a study guide, now you can see here it is start generating the study guide for me. No has generated a study guide for me. Now you can see. Study guide what is a study guide quiz? It has generated the quiz also answers, quiz answers, essay format questions, and glossary of terms and all those things. This is a study guide, which is generated according to mine text that high pace or topic like. Not only that it can also generate the briefing doc. What is a briefing doc means? It will just brief what is about the source or what is about that. For example, we can go for the summarization, you can see the briefing document. This is a briefing document which is generated by the Goaser seeine Management source like that. We can check it by yourself for more information. Now we can also convert into time length, right? Seeking death. Now you can see it is a timeline where which is happening or like that. Implementation, establishment, all those things, Timeline means, what are the time stamps or time used in this particular platform or in the source that you added. It is at best when you use the YouTube YouTube link. They have the time, with that, you can chat directly here with that link and it will give only the what you want from the YouTube video. It will explain in the other language also, all those things. There is a multiple or unlimited usage of this I platform, that is up to you how you can use it. I hope you understand these points very well. This is how you can use this notebook L M. So this is up to now, not only this, you can check it out the video on YouTube, about this notebook LLM for latest futures or for detailed tutorial or how to use this Notebook LLM in effective manner. So you can learn from that. That is all about notebook M. I hope you understand these points as well, right? Let's go to our app. AI prompts prompt engineering app. It is a best app because if you if you want to if you want to save that time in writing the image prompts, video prompts or text prompts. You can use this I prompts prompt engineering app, in which you can find the different prompting templates. It is already made with the prompting templates you can go for the YouTube automation. It has also contains the prompts for education itself, or teachers, for students, all those things. Not you can find the best prompt templates in this app itself and you can see there is a assistant in which it will generate the text prompt, you can see here. You can check it out here. You can generate the image prompt for you and it can generate the video prompt for you based upon your basic prompt. That means you need to give the task or description, and it will generate the image or video or text prompt for you. Now you can check it from this Google Play. If you have the Android phone, you can get from the Google Play. I understand these points. Not only those, we have another platform that is open AI playground. 31. 5.3 Overview of OpenAI Playground: Open EI playground, in which we can create the assistant. What is assistant? Instead of writing the proms in the hag BT multiple times, we can create the assistant in playground. Open AI playground itself once. We will write the system instructions one time and we will select the model and we can chat the multiple times. The assistance means which assist you for the specific one. For example, if I want to if I want the assistant in which it will generate write the best quizzes for the physics subject. No, I will create the simple assistant. Let's see here. Let's create something. You can come to ask to open AI platform or playground, like that. Come and you can come from here. Playground, you will come here. After that, you can check it out all those things playground, dashboard, dogs, ABA preferences. If you don't know about this platform. So please don't panic. Don't be panic, just go and learn about this platform in the YouTube itself. There are much more better videos, from that you can learn it. I will just show here what are the platforms this will provide you? There are a prompt session, real time. We can enable access and you can share a talk with this PI models assistant. You can create your own assistant. I will show in a little bit of time and you can tedus. This is the text to speech to text, right? Text to speech. If you write here, it will generate the speak audience format. There is some services from them, ChatBopenEI company. I will show the creation of assistance. We can use all those things. Remember one thing, if you are a new Customer, if you are new to this AI platform, you will get the free tokens, up to $5, all those things. You can use up to five. If the $5 of the tokens will completed, you need to pay this amount. You need to pay according to your tokens usage. Just to go and know about this all about things. Fine tuning, evaluation batches. As a teacher, you do not need so much thing. Don't go to the technical bit. I will just show you how to create your own assistant in which it can assist you in specific one. Let's create one. Just come to here, create the button, right? Come here. Now, you need to give the name for it. I will tell you can give anything. I'll just write physics. Sorry. Physics squizGenerator. I'll just taking example, you can tie anything and remember. You can just click here, playground. It will take the playground in which you can test your AI assistant. This is a name. Now, system instructions, remember one thing. We have already learned the different prom patterns that is like powder to verify personal prom pattern already. If you have already watched the previous sessions, videos, topics, all those things, you will get the idea. So that's why the prompting is everything. In the system sentiens, you need to write a clear, very powerful prompting. Let's take physics teacher. Six years off. Experience. In teaching physics. You can lie like that, for example, I'll just writing the simple prom. If you want to create any assistant, you need to the different prom patterns and different techniques, all those things that I have already explained into the previous sessions. Write from that particular knowledge and just write this. I will just writing it for you. Now your task is to generate what is this? Per seventh standard. Students. Let's stack with students. Simple. I will provide copy. No. Let's take ask me for the topic. Based on the topic, you need to generate quiz. Remember, you need to follow about task on Nik you do not and you perform. Other task. Simple. We need to give all the context, personal prom pattern, all those things. After that, we can choose your model, GPD four. Please remember one thing. You need to be cost a bit different A model have their different pricing for the tokens. So please keep in mind that just go and check it on there and select the best model. I think mini let's let's take this P four, you can take any of this, right? I will take the CP 3.5 Turbo. Let's try this. If you have already the topics or quizzes, just click here Pilesearch and add the file here from here. You can upload from your computer. Based upon that file and it will and just from instructions will generalm which is for you, which is the best thing here. Right now I don't have any file, so I will not give. The text format is text only just keep with that. You can change the temperature to what the temperature top is. By changing the temperature, you can change the output. You can control the output. At the same time, top also. If you decrease that, the output also decree the output the quality of output also be controlled. We can test it out by sl. Just do that and that is simple. Now it is generative. Now I will write the answer. Hi. Let's start with the hi message. No, it will give the answer. Hello. This is a physics quiz generator. Hello, I can help you with a quiz and a specific topic in physics. What topic would you like the quiz to be based on? Now, I will give you some basic topic about let's take light. I give the topic. No, it will generate a quiz for me. That is simple. No, you can see here. I generated the twin quizzes for me. What is a light? What is the speed of light? Which is the following. You can see that it has given to the options also multiple choice quotiensT is a specific one. So we can the specific for the specific subject, you can the answer and quotients, essay, quizzes, multiple questions. You can do all those things or you can use you are a mathematics teacher and you have the specific experience in the solving the problems like that. That is all up to you how you can and what are your requirements? What are the assistant you are looking to build? That is up to you. You can write in the prompts and just build it assistant and you can use by yourself. That is how simple in this AI world. Okay. I hope you understand these points also. No, you can do here file search. Just click here and you can do the file search also. You can the image from the image, please generate the quizzes. Like that. You can do all those things from this platform, it can show them how many tokens it has taken to generate ten quizzes. We have that, so like them. For better tokens, all those things, if you are a teacher, you do not need the ID technical bit, just a I am giving awareness to you. There are a lot more AIM platforms that you can use it to become more creative and productive by yourself and to make the students also very aware of this because we are living in the AI world in which you need to learn all those things AI platforms to become more creative in your field. That is up to the AI tools and the session. Now there are other tools and there are other platforms which gives you the best. So as a teacher, I just which are very important and which are very useful for you. Okay? That's why I have just taken small platforms like Open AI playground. We have seen this AI app and have seen this Google AI studio, Notebook L and Canva. That is how we can use these AI platforms to become more creative, productive in your field. I hope you understand these platforms apps and much more in this session. Remember one thing and you do not need the technical bit, but you need to know what are the specific platforms offering the services in which you can use this for your but you to save that time. I hope you understand this session very well. Okay. So for more AI platform tools, you can go to search the online, what are the best tools for the teachers, all these things, you can get from that. And for more specific, if you want to master any specific platform like Google A studio, ChaGPT, there are much more EIMH models, also storage generators, much more apps, you can go to Google and YouTube learn from it and just save your time. This is how the AI can help you education, right? This is the technology which is started with the past two to three years. We cannot imagine how much it will go, how much it comes to technical and which is very easy to create something right now. I hope I understand these points. Okay. Okay. In the next session, we are going to see what are the opportunities instead of teaching? You have the other opportunities in this AI world in which you can use your skills to do something to do some impact in the work. Not only a teaching field, so you can make money by other part time work freelancing work. What are the different opportunities we have after this course? We will uh discuss in the next session. Let's dive into that. 32. How to find Jobs & Freelancing Opportunities: Now, in this session, so we are going to see what are the different opportunities you have with your subject knowledge, specific subject knowledge, and prompting skills. Okay? We have different platforms that are there offering the opportunities that you can do the part time work like that. Teaching experience after the teaching job. You can go to come to the specific platform that is Olayer which is a best hiring platform for the AI trainers for the specific subject knowledge experts. We can come here, just type the outlers in the Google platform. It will give the first line that you can come from here directly click open Opportunities and you will landed here. What are some open opportunities you can see here. You can do the all or you can select the specific you can specific location you country, all those things. It will show the 71 opportunities like that. Okay? You can select the types, specialist coders, generalist languages like that. Let's see each and everything. For example, if you have the mathematics specific subject. If you are the teacher who have the expertise in the advanced mathematics, we can apply for this job. A, you can see the different subjects like A training for accounting, businesses, computer systems, data science, electronics, government, public sector, healthcare, life sense like that. Public health, robotics, socialist software engineer, Okay, you can see math expertise, architecture engineering. There are biology expertise, biology expertise, chemistry, clinical medicine, electrical engineering, there are physics expertise, right? There are other very good opportunities for you. As a teacher, you can grab this and you can basically what is the job role in this particular platform is. You need to write the prompt and response for the EI model, right? For example, they have the different clients in which you need to train your expertise. Like for example, if you ask the GPT about solve this particular equation, it will simply solve the equation and it will show the answer for us. The answer and the quota is written by the prompt engineer who have the expertise in the MAC already. The chargBT give the answer. The answer is given by the ChargD but it is trained by the mathematics expert like human. The AI models also developed by humans only, in this case, if you have the specific knowledge about physics, mathematics, English, or anything, specific knowledge, you can come to this platform, search it. If you have the particular opportunity, just click here and you can apply it here from there. I think this opportunity it's taking time to loud, right? You can see here. You can directly apply now here. You can see remote work, flexible hours, weekly payout, flex expertise. The output, the pay also very high because you have the advanced mathematics. This is how you can use this particular newly prompt engineering skill with what specific knowledge. You can get some impact in the work culture or the companies. I hope you understand this platform. We can go the coders generalized languages. If you have the knowledge like a voice acting for a training English, lance writing inflammation like that. This is not up to this. You can change the locations, you can get all those things. Not only that the new opportunities will be added in this platform, so you need to check regularly if you are interested in this platform. Not only that, there are other platforms, but the outlaer seems to be a genuine and better platform for me because I have already worked in this. Platform. I hope I hope you understand this point. No, not only that you can find the different flancing websites like freelancer, Fiber, write Upwork, these three are the best and popular plancing websites in which you can list out your services or profile that showcase, you have the ability to write the prompts, and you can get the best output for the clients in the specific knowledge that you have. For example, if you're a teacher, you have the expertise in the physics, you can build the best uh profile in the different platforms like freelancing, five. You can go here and you can just come to work. There are other plancing websites as well. But these are three popular one freelancer, fiber of work is a popular one. We have another that is People per hour. But let's take that also. We can come to here this people per hour. There are very or different types of PI financing platforms in which you can least towards services profile with your experience and projects you have done and the courses you have taken, all those things. When the client will approach to you for the experience or skill you have, you can talk with them and close the deal or close the project and get paid. That is simple. This how these platforms will work. Just come to these platforms and sign up. The signup is free in every platform that I think freelancer fiber people per. This is a free to sign up. Okay, just build the genuine profile with your experience, experience, projects, courses you have taken certifications, all those things and just take some portfolio website in which you have written the proms and you have to get the output, how it makes some impact in your work daily life, and just build a simple portfolio with your interaction video with your face like that. Okay. And just come to here and just update your profile in each platforms. Not only these four, you can just simply ask the freelancing websites. You will get the different flancing websites in which you can see, you can see 23 best flancing websites to find work in 2025. There are up work design here, it is a design, top top is also better best freelancing platform. You can hire for the full time as well. Linked in also very powerful thing. There are much more freelancing platforms in which you can build your profile and post services and get them. That is simple. Every platform have their own features. Somebody will take you to bid the project or somebody will tell you some platform will tell you to last services with our projects have done previously like they have their own features and working ability. You need to learn about this. If you want to learn about this each unique platform to sell your services as a teacher or if you have the specific skill in the AI engineering or other thing, just there are these platforms allows you to bid or to sell different services. Not only the one specific one, you can sell different services to anyone. They have limitations, but they can allow you to tell different multiple services. Just go to YouTube and learn each and every thing about the platforms and just know and go and start it. That is most important. That is how you can use this prompt engineering skill or with your specific knowledge expertise as a teacher or anything. You can place your skills and you can sell your skills through these platforms, platforms, freelancing platforms, and you can come to the outlay as well. Not only Opare there are other platforms like outlier, which performs platform, but the outaer to be genuine and very effective. I hope you understand these points. Okay. Not only that, we can create our own AA apps. The prompt engineering is not only the skill of writing the best prompts, but it will unlock your true potential because this skill can change your mindset, the way of thinking, and you can go your limit and you can put your limits to build something because the AIE education will evolve in upcoming decades or years in which you can stand out from the crowd as a teacher. You can build something app in the education field like teacher, you have the specific one, you can build our own AI app Android app, IOS app or web app with your building abilities. If you teacher, you don't have the coding, you do not need how to code it or you do not need anything. You can write out different local tools which create by Dragon drop which creates by be coding. All those things they have a lot more things if you uh think about it. There is a lot more opportunity in this AI in education. You need to come with the idea, you need to be execute well and you can make something impact in this education world. I hope you understand these course as well. Please go and check all the prompt patterns very well because these are the not prompt patterns only these prompt patterns will give the ton of AI apps ideas. You can get the ideas from the directly prompt patterns. Okay, I have already shown you to build a AI assistant in the platform, Open AI platform. From that, you can build O EI assistant in which you can build the app from that and you can un the money from that particular AI assistant by giving access to the who all over the world. The problem is you need to build the app that is simple. There are a lot more tools like flutter flow and other AI tools that you can help to build the AI app very fastly to test it out in this market and just build it and make impact and make more money rather than teaching. Because the AI is not only help you to productive, but it will unlock the more opportunities to get some money or to make something impact and to generate more money for you. That is all about this course. We just learn and all the prom patterns. If you learn that all prompt patterns, if you ti practice with AI LLM models, you will get the app ideas. From that, you can build something to move forward. Teacher means always learning and growing at the same time, explain to the students. That is the role of the teacher right now. Just use this technology and help the students to move forward with this technology and be productive. And always be creative to create something new and never give up on your field. That is up to now this course. I hope you understand this course all over from basic to advance, all those things. I think I have all those things, just practice well, and you can come to the job platforms as well. Not only, you can try from the free from after the teaching job or anything. Okay, can come here and you can apply the opportunities all those things. 33. Final Thoughts: Up to this our ports is ending the prompt engineering for the teachers, professionals. So that is all about. And make sure you follow my account in the Skillshare or more courses, I'm planning to create the ports or master class on building apps, AI apps, production today apps from the building it to production in the Google Play Store. That is all about this I prompt engineering for teachers. I hope you understand these points. Okay. I will meet with you in the next course. Till done, goodbye and always keep learning, keep growing, and keep happy. Thank you.