ChatGPT Masterclass: The Ultimate Guide to Generative AI Tools and Applications

1. 1 Introduction: Hello and welcome to our course on Generative AI. I wonder if Thank you for enrolling and taking the time to learn about this exciting and rapidly evolving field. In this course, we will explore the different aspects of Generative AI, including the OpenAI company and it's Research, Prompt Engineering, GPT and diffusion models and how they can help you with everyday tasks and managing your time more efficiently. Whether you're a full-time working professional student or just someone who is passionate about learning or technology. We will also cover different tools and applications such as ChatGPT, DALLE-2, Midjourney, Bing Chat and Microsoft Designer. To begin with, we will first start by covering the theoretical components so that you can build a good fundamental understanding of the subject. We will then provide practical examples and hands-on exercises so you can apply the concepts you learned and gained valuable skills and knowledge. By the end of the course, you will have a comprehensive understanding of Generative AI and how to use it to enhance your creativity and productivity. We're excited to have you as part of our class. So without further ado, let's get started. Here's a quick look at the course curriculum. We'll start by introducing the concepts of Generative AI. Will talk a little bit about the OpenAI company and it's research will introduce a couple of AI models. Next, we will cover Prompt Engineering and why it's important to provide good prompts and how they can affect the quality of your output and data. In addition, we will cover the benefits of leveraging Generative AI and how they can help you with design, creativity and productivity. After we develop our understanding of Generative AI, we will dig in and start using applications that are built on top of the AI models and APIs through hands-on exercises and practical examples. These include ChatGPT, Midjourney, Bing Chat, Microsoft Designer, and Adobe Firefly. Near the end of the course, we will cover an important topic which is how not to use Generative AI and best practices. And we will wrap up the course with some final words. 2. 2 Generative AI: Okay. So what is generative AI? Okay. Generative AI is a type of artificial intelligence that can generate new content or data, and this new content or data is going to be very similar to the training data it has been provided previously. It has the potential to revolutionize various industries and create new forms of creativity and productivity. There are many applications of generative AI, but will cover the top three that are most popular. The first one is natural language processing, also known as NLP. Generative AI can be used in NLP to generate human like text or speech. GPT three and GPT four are language models that are developed by open AI. They can generate articles, poetry, computer code, and so on. You can even build chat bots or conversational bots on top of those, and you can even have it translate texts from one language to another. Next, we have computer vision. Generative AI can be used in computer vision. It can generate realistic images and videos. DL two is a generative model developed by Open AI the company. It can generate unique images based on text prompts. You can even use it to edit or manipulate images through its features. Lastly, we have robotics. Generative AI can be used in robotics to generate new behaviors or movements for robots. It's reinforcement learning algorithms can generate new actions based on feedback. Just like with anything else, there are challenges and limitations associated with generative AI. One is data bias. These models are only as good as the data they're trained with. If the data they're being trained with is biased, then the generated content will also be biased. Then we have computational resources. These models require significant computational resources for training and generating new data. This makes them less accessible to smaller organizations and individuals. So what happens is that few of the larger organizations who have the power and resources to build this will build them and then offer it to smaller organizations through paid subscription. Lastly, we have overfitting. Sometimes this can become too specialized and unable to generalize two new data. 3. 3 OpenAI company and Research: Alright, now let's take a quick look at the OpenAI company and Research. Openai is an artificial intelligence research laboratory. It was founded in 2015. It first started as a non-profit, but later became for-profit. And it has received massive amount of fundings from giants such as Microsoft. It offers subscription-based services to those who want access to their technology and platform. The company goals and visions are developing safe and beneficial AI technologies. Advancing the state of the Art in AI research and promoting AI education and policy. Openai has some notable achievements that are worth mentioning. In late 2020 to the company announced the development of GPT-3 and made it available to the general public for free in early 2023. It really is ChatGPT for underpaid Subscription Model, which also comes with new features such as plugins. And the development of GPT five is currently underway and may become available for use by the end of 2023. They have made contributions to the field of robotics. They have also made some other code open source such as evils, which is a framework for evaluating large language models. And you can find that on GitHub. In addition, they have many publications and papers to help AI community and the public 4. 4 AI Models: Alright, now let's start introducing some AI models. In this section, we're going to cover two models, GPT four and DALLE-2. Gpd four is Generative pretrain transformer four. And DALLE-2 stands for diverse All scale transformer by a lifelong experience. Two, these are both advanced Generative AI models, and they each have different capabilities, applications, and limitations, which we'll cover next. Gpt four is a deep learning-based Generative AI model developed by OpenAI. It uses transformer architecture and it is trained on a massive amount of text data, which makes it capable of generating high-quality texts in various formats, such as articles, stories, called blogs, and much more. It has the ability to understand natural language. And it can generate human-like responses, which makes it great for human interaction and for applications such as conversational chatbots, similar to ChatGPT. Dalle-2 is also a Generative AI model developed by OpenAI. It uses the diffusion model. It can generate images based on textual input. It also uses a transformer architecture. It is trained on a diverse range of image and text data, and it is capable of generating highly detailed and realistic images. Here's a quick overview of the capabilities of these to AI models. They are capable of generating high-quality text or images based on textual input. They can understand natural language and generate human-like responses. You're good at answering questions and providing relevant information. They can translate takes from one language to another. They're good at summarizing text or Conversations are Chats. And they are good at generating personalized recommendations for users based on the prompts. Here's an overview of the applications of these two models and how end users can leverage them. They're good at generating high-quality content for blogs, websites, and social media. They can enhance Chat Bot interactions for purposes such as customer support and customer service. They're great for automating translations and summarization tasks. They can generate creative writing prompts and assist with writing tasks. They're good at creating personalized recommendations for users. And they can generate highly detailed and realistic images. Of course, there are several limitations associated with these to AI models. First one being biased. As we covered earlier, like ALL AI models, GPD four and DALLE-2 can be biased based on the data they are trained on. Next, we have output quality. While GPT four and DALLE-2 are capable of generating high-quality output. The quality of their output can still be affected by the quality of the input Prompt. So if you're not clear and concise, if you're not clear about your intent and you don't provide enough context that will affect the content that it generates for you. And lastly, we have generalization. While GLUT4 and DALLE-2 can generate highly diverse and creative outputs. They may still struggle to generate outputs that are significantly different from their training data 5. 5 Prompt Engineering: In this section we're going to cover the basics of Prompt Engineering and how he plays an important role when using these AI models and how you can leverage them to create prompts that will give you the best quality output. When using Generative AI models such as GPT, prompts are needed. And these are usually in a form of textual inputs. There are typically two types of Prompts. One is questioned and the other one is instruction. Prompt Engineering is the process of designing Prompts that can be used to guide Generative AI models. It helps them to generate specific content and completing specific tasks. The goal around Prompt Engineering is to create a Prompt that provides enough information to the model to generate the desired output without being too specific or too general. Well-designed prompts can save time and increase productivity by guiding the model to generate the desired output with minimal input from the user. There are different types of prompts, but we'll cover three in this lesson. First we have the tax Prompts. Text prompts are the most common type of Prompt and can be used to generate text-based contents such as articles, stories, or code. Next, we have Image Prompts. Image prompts can be used to generate images based on a specific visual concept or idea. Lastly, we have dialogue prompts. Dialog prompts can be used to generate conversational responses or Chat Bot interactions. Here are some best practices to consider when creating Prompts. Be specific, as specific Prompt will help them model generate more precise output. Be concise. A concise Prompt will help the model focus on the most relevant information and be diverse. A diverse set of prompts will help them model generalize better to new inputs and generate more creative outputs. So in summary, here's why you should care about Prompt Engineering and well-crafted Prompts. Prompt Engineering is a critical part of using Generative AI models effectively. By designing well-crafted prompts, users can guide these models to generate the desired output quickly and efficiently. And with the increasing sophistication of the Generative AI models, Prompt Engineering is becoming an essential skill. In this lesson, I'd like to cover three examples for Prompt Engineering and well-crafted prompts that you could use in real life. So the first one is generating attention-grabbing headlines. So here's an example. Generate a social media posts about the benefits of using our product. Now here we've specified and we're telling the model to generate a post, but it's of type social media. So based on its previous training data, it knows what a social media posts looks like. We're also being very specific of telling get the context and the intent, which is the benefits of using our product. And it can generate us a high-quality output, assuming we have given it enough information previously about our product. Next, we have generating personalized social media content. So here in the second example, our Prompt is generated social media posts about the latest tech news that would be relevant to a user interested in AI. So again, this is a well-crafted Prompt because we've specified the type here as social media, being the Post. The type of posts. It's going to be talking about a latest tech news. But it also narrows down the context to not just any tech news, but tech news that is relevant to somebody who was interested in AI. So it's being very clear, it's being very concise and it's being very specific. And this will lead to a very good output. And the last one here, we have generating social media posts for businesses. So here we have our Prompt is generate a social media posts about a new product launch. That highlights is key features. Again, this is the intent is clear. The context is clear and obviously, having provided IT context and information about the product and its features previously, this will generate a high-quality output 6. 6 Design Creativity Productivity: In this section, we're gonna cover how using Generative AI Tools and Applications can help us with design, creativity, and productivity in our daily lives. Generative AI can enhance creativity and streamline design tasks leading to increase productivity. It can help us by generating new ideas, exploring different design possibilities, and automating repetitive tasks. It can also help us by handling somebody tedious work, allowing us to focus on more high level tasks such as concept development, leading to more efficient use of our time. Prompt Engineering can be used to generate specific design elements such as logos and patterns based on certain specifications. It will save designers a lot of time, and it also helps lead to more consistent and cohesive designs. Here's some tools you can use to utilize Generative AI for your daily tasks. Chatgpt, which is a conversational Chat Bot built on top of the OpenAI APIs and the GPD model to generate content. Next we have Bing Chat, which is built by Microsoft and is another conversational AI tool that can be used for generating contents such as social media posts and blog articles. And lastly, we have Google board. This is very similar to ChatGPT, but powered by Google's most advanced large language model, palm to. The last three tools we covered generate textual output. And in this lesson we'll cover three tools that will generate images. First, we have Midjourney. So Midjourney is a Generative Design tool that uses AI to generate unique visual designs such as posters or book covers based on the user input. Next we have Microsoft Designer, and this one is a design tool again, that uses AI to assist designers and creating accessible designs by providing suggestions and recommendations. And lastly, we have Adobe Firefly. This one is an experimental AI power tool that can be used for generating unique visual designs such as abstract graphics and illustrations. Here are some examples of what you could potentially accomplish with the help of Generative AI in design. So Generative AI can be used to generate unique logos for brands based on certain specifications. It can be used to generate unique patterns for textiles, such as clothing or wallpaper. It can be used to generate unique musical compositions based on certain specification, such as genre or mood. And lastly, it can be used to generate unique visual RPCs, such as abstract paintings or digital illustrations 7. 7 ChatGPT Account Setup: Now that we have walked through the theory behind Generative AI and the Models, it's time to go through some hands-on exercises and gain practical experience to bring it altogether. We'll start by ChatGPT. You don't need a whole lot to set up and get started with ChatGPT. All you need is a device with a browser and an Internet connection. Openai has also released ChatGPT app for iOS so you can use it natively on your phone if you like. Okay, before using ChatGPT, we first need to create an OpenAI Account. In order to do that, just launched the browser and go to openai.com. Once you're here, on the right-hand side, you'll see login and you'll see sign-up. If you already have an account, you can click on Login and putting your username and password and login. If you don't have an account, click on Sign up. And once you click on Sign up, you can put in your email address and go through the registration process. Once that's all done and you verify your email, you can login with your newly created Account and start using ChatGPT 8. 8 ChatGPT Questions and Instructions: Alright, once you gone through the registration process and you login to ChatGPT, this is the user interface that you're presented with. Now let's take a quick tour together. On the left-hand side, this is where you have all the historical data. So one of the nice features about ChatGPT is that every time you start a new Chat, it will return, retained that Chat and all of the information. So if you ever wanted to go back and take a look at specific Prompt that you generated or a specific output as a result of a prompt that you put into ChatGPT. This is the perfect way to actually do that. And you can even continue to continue that conversation with ChatGPT if you want it to. So this is one of the nice features. It keeps a record of all the Chats you've had in the past. And you can also, there's some actions that you can take care. For example, you can clean up your history and the Chats by just simply deleting them. Now on the left-hand side here, because this is my personal account and I use this for personal projects. I've blurted that's why you can't see anything. But once you start chatting with ChatGPT, it will start. Your Conversations will start showing up here in the left-hand menu bar. And on the top here you have plus new Chat button. So whenever you want it to, wherever you want to actually start a new Chat with ChatGPT, all you have to do is come here and click this and it will load a brand new Chat and you can continue on the center here we simply have just some examples and some informational facts. So here's some example prompts that you can use to get started with ChatGPT. You can either just type this in here and they send the message area where you can simply just click this prompts and it will auto fill them for you. And these two columns here, there's just some general informational facts about the capabilities and limitations of ChatGPT that you need to be aware of. Now we're ready to run through some exercises and go through some examples. But before doing so, one thing I would highly recommend is that whenever we go through an exercise, first try to go through it yourself and then get some help from ChatGPT. And get ChatGPT to do that for you. And then compare the two to see how accurate the results are. For our first example, we're going to start with something very simple. When it comes to learning. Chatgpt makes a great teacher and a great instructor. You can literally ask it to explain anything to you in simple terms. You can even give it persona, you can give it. You can make IT expert in a specific subject and you can ask it to teach you things in very simple terms. So it's easy for you to learn specific topic or subject. And sometimes things that would take you days to research. You could learn through ChatGPT in a matter of minutes. Let's say you're a psychology student attending university and you're about to take a course in neuroscience. Now you have no idea what neuroscience is, but you'd like to find out, and this is exactly how you can leverage ChatGPT so that it can explain to you on a high level what the subject is all about. But let's go to the prompts section here and type explain neuroscience. Now one thing we can do is we can stop here and run their Prompt. And ChatGPT will try to gather as much information so that it can explain on a high level what this object is about. Now one thing I would like to bring to your attention is the topic of Prompt Engineering, which is something we touched on in the earlier lessons. And how it can play a pretty crucial role in the quality of the output that ChatGPT will generate for you. Here we simply talk about the water. So we're trying to get ChatGPT to explain to us what neuroscience science is. One thing we can do is we can provide it a little bit more context so we can, for example, talk about the how and who. So the audience this is actually for and how ChatGPT should go about explaining this. And that should, that makes our Prompt clear. It makes it concise. And also it gives ChatGPT more contexts. So it can have go and find more relevant information from the web and it's training data. So let's go ahead and I'll add a little bit more to make this a more well-crafted Prompt. So while we can say is explain neuroscience in simple terms, suitable for a student who is psychology major in second year of university. So here we talked about the what, and here we talked about the house. So in simple terms. And then here we talked about the audience. Who is this four? So this is for a student who is a psychology major in the second year. So we're being very, very specific with our Prompt So let's go ahead and run this. Now. Chatgpt will do its best to try to come up with as much information for us. But one thing I would like you to remember is that this is just going to be a very high level and simple Overview of this subject. But it will help you grasp the basic idea behind it. So as you can see here, ChatGPT has finished explaining this. So you can read through this and get a basic idea of what this object neuroscience is all about. In our last example, we instructed ChatGPT to do something for us and that was to explain a specific subject. In the next example, we're going to ask ChatGPT Questions. So just like similar to how you would go and type a question in a search engine such as Google. You can also do the exact same thing in ChatGPT. So you can ask a question. So for example, let's say you want to enroll in an MBA program. And you would like to know what type of things you would learn about finance when you attend that program. So we can ask that question from ChatGPT. So let's go ahead and do that. What will I learn about the principles of finance in an MBA program? Now one thing we can do again with our Prompt, we can just stop here and run their Prompt. And ChatGPT will give us an answer. But a nice feature about ChatGPT is that you can control the format of your outputs so he can tell it how to, how you want the output to be formatted. So for example, couple of options or you can do it, you can have it format the output in a bullet point format. You can have it in a table format and so on, right? So there's lots of options that you can choose from. So let's go, Hey, let's go ahead and say bullet point format. So I'm just going to instruct it to format the output in bullet points. So let's go ahead and run this command. And this is exactly what ChatGPT is going to present back to us. So you can see that it's highlighting the key features about the principles of finance and what you would actually learn about them in an MBA program. And here's the, here's, here's all everything. As you can see, everything is nicely formatted in a bullet point format. You can also do the same thing in the table format. If you'd like to specify columns, rows, and things like that. And here it says as these are some of the key principles and concepts covered in an MBA that focuses on finance. We can even go one step further. And let's say you want to, one of the bullet points here, call it your interests. And he wanted to get a little bit more detail. While we can do is we can even take this one step further and get ChatGPT to dig in a little bit deeper on some of these topics. So let's say, for example, let's choose this one here. Financial statement analysis, evaluating and interpreting financial statements, balance sheet, income statement and cash flow statements to assess the company's financial health. So this is a good one. So while we can do is let's go ahead and copy this. And let's create a Prompt here. So what we want to say is, I'm going to make ChatGPT and expert in finance been saying as a finance instructor. And now I want it to, so I told ChatGPT who I want it to be. The intent is for it to be an expert in finance. So as a finance instructor, elaborate and provide more details on. And now what I can do is I can put what I copied and I can paste it here in quotes. And now I can actually run this command. So let's go ahead and do that. And this is where ChatGPT will start actually creating, providing up with of exactly what those things are. So you can see Balance Sheet analysis. It gives you more details about just that area Income Statement Analysis. It tells you what that subject is about. Cashflow statements, same thing, you can see the bullet points here. And then financial ratio analysis, a comparative analysis and benchmarking and so on. So now Chat, you can even use ChatGPT to go in more in-depth on some of the outputs that you can get. You can even do one, you can even go one level deeper by doing the exact same thing and say like digging deeper in one of these topics here that's provided from one-to-five here on the specific topics. So as you can see here, ChatGPT can be very powerful tool from a learning perspective. 9. 9 ChatGPT Blog Post: In this exercise, we're going to focus on how ChatGPT can be leveraged by content creators. For this example, we're going to write a blog post on a specific topic using ChatGPT to generate an outline and suggestions for the posts. We will then compare the results and you'll see the ways in which ChatGPT can assist in Content Creation. We will provide to prompts. The first one is to get ChatGPT to help with outlines and ideas. And the second prompt is to get ChatGPT to do more work for us in terms of writing the actual blog. And you can decide which works best for your style of writing. This structure of a typical blog posts consists of four main parts. The title, introduction, body, and conclusion. Now for our first Prompt, the goal here is to get help from ChatGPT to create an outline and to give us suggestions and not to necessarily do all the work for us. For our first Prompt, what I'm going to do is I'm actually going to copy paste it here so you don't have to watch me type word by word so that I don't waste your time. So I picked just a random topic here for this blog post. So let's go through it together so you can see the different components of this Prompt. As a Writing Assistant, I need your guidance to create an outline and suggestions for blog post on the topic of artificial intelligence and health care. Please provide some key points and sub-topics to cover in the blog posts. So here, I'm not necessarily asking it to write the blog post for me and it's entirely, I'm just asking it to give me some suggestions. So let's go ahead and input this Prompt. Okay, ChatGPT has finished creating the output for our prompts. So let's go ahead and take a look together. So as you can see here, first thing, this is the first component of our Blog, which is the title. So you can see ChatGPT has created a nice style rule for our blog posts, revolutionizing healthcare, the power of artificial intelligence. It's a very good title actually, I like this one is catchy and it's sort of conveys the point about what the blog post is going to be. And you can see that the next section that Chat GPT has provided for us is Introduction. So this is great because ChatGPT is not necessarily Writing Blog post for us entirely. It's creating an outline, so it's giving us a title, and it actually gave us an example title here. So this is the outline, this is the introduction. Here you have the after Introduction, you have all the, basically the subtopics that you could talk about this field. And you can see that we have several here. So this is basically the body. This will make up the body of your blog posts. And if you scroll all the way to the bottom, you'll see that there's a conclusion section. Now note that ChatGPT hasn't really created the content for you. It's giving you an outline along with some suggestions. So Introduction, it actually has a written you one paragraph. It hasn't created the content for you, but it's telling you how you could go about doing so. So here it says briefly introduce the concept of artificial intelligence. And it's growing significance in various industries, particularly in health care. So this is where you can actually go and research in this area and then cure rate your own content and add it to your Introduction Part. Another one here is highlight the potential of AI to transform health care. So this is again, another thing you can go and research on the side. And then once you are able to write your own paragraph from the curated data, case studies, whatever it is that you're researching in that topic, you can actually include it there. Then we have the subtopics here in the body. So applications in diagnosis and diagnostics and disease detection. So here, ChatGPT is making a recommendation that in our blog posts we can discuss the role of AI in early disease detection and diagnosis. Here is patient care and treatment. Again, it's giving you tons of ideas and suggestions of some other things that you can actually discuss in the body of your blog. So this is actually pretty fantastic and amazing because this is exactly how you should be using Generative AI. You want it to spark your creativity. So as a writer or whenever you hit a wall, you can actually use Generative AI and ChatGPT to help you get past that. And as you can see here, we have the conclusion section. Again, it gives you a suggestion to Recap the transform. It might have potential of AI in healthcare and its ability to improve patient outcomes and so on. Here, there is a note this is, I find this very helpful as well. It says this outline provides a comprehensive structure for your blog posts. But feel free to modify and expand Each section according to your preference. And these are length of the blog post because we haven't really told it how long be one blog post to be. So that's something we can consider, which you'll see in our second Prompt. And one other thing I really wanted to show you here is you can use ChatGPT. Now that you have this outline, you can use ChatGPT to dig a little bit deeper in any of these topics. Now one example here I wanted to illustrate is the title. So I actually really liked this cidal, but let's say you didn't like this title, but you're having a difficult time to come up with other titles. So this is again where you can leverage ChatGPT. And if you want to keep things very simple, you can simply follow up because ChatGPT now has contexts in terms of what you've asked it and what output it has given you. So now you can say, please generate five more interesting. And I caching titles for this blog post. And ChatGPT will generate five more for you. So as you can see here, we got five more. And if you'd like some of these, you can use them. If not, you can use a piece of each title and create your own and QRA your own. But this is the power of ChatGPT. Chatgpt when it comes to Content Creation. For our second Prompt, I'm going to try and get ChatGPT to write more of the blog post for us. So I'm going to try and leverage it to fill in some other blanks. In the first Prompt, we just asked her for suggestions and an outline, but in the second Prompt, we want to do we want it to do a little bit more work for us and fill in some of the blanks and cure rate some of the content for us wherever it's possible. So again, I'm not going to type everything. I just picked a random topic in sustainable energy. So let me go ahead and paste that here and we'll go through that together. So as a writer, create a compelling and informative blog post on sustainable energy solutions. Please provide a brief introduction to the topic along with the key points and suggestions on statistics, case studies, or actionable steps to make the Blog Post engaging and persuasive. This blog posts should present a thought-provoking piece that promote sustainable energy Practices. This blog posts should be approximately 1,500 words. So note that I'm not just asking for an outline and suggestions. I'm also setting the tone for this blog post and for the audience. And I'm also telling it how long the blog posts should be. So I'm setting some requirements or lengthwise. So let's go ahead and run this Prompt. Ok, ChatGPT has finished the output for our Prompt. Now actually what happened was that the first output that it generated, I wasn't really happy with it because it was very similar to the output of the first Prompt. So all I did was I click this regenerate response and it created a second output for me, and I'm a lot happier with that. So let's go through it together. So you can see here the, it created, again, just like the first round, they created a title for us, which is nice. The one difference between the first Prompt and second problem is that you can see that it actually created the introduction for us. So if we scroll up again to the first Prompt, it didn't really write a paragraph for us. It didn't curate the content. They didn't take your other Content Creation for us. It just gave us ideas on what we could include in the introduction. So these are suggestions or recommendations of what to include in our introduction. So for example, briefly introduce the concept of artificial intelligence and its significance in various industries. Here. In the second Prompt, it actually wrote the introduction for us. So that's one less thing for me to do as a writer or a content creator. So, and it's actually pretty good one, imagine a world where our energy needs are met without harming the planet. Where it clean and renewable source power, our homes, businesses, and transportation. So this is actually pretty good. The body looks very similar, so it's just an outline with recommendations and suggestions of what you need to do. So this is actually very similar to the first prompts, so it's just giving you suggestions, present compelling, compelling statistics on the current state of energy consumption and its impact on the environment and so on. So this is actually very great. This is where you can actually go and research in this area and included in your body of the blog posts. And one other thing is that you can see that it also created a Introduction for us. Again, just like Introduction, which is really nice. Because again, as writers, this is one less thing for us to do. So as individuals, communities, and nations, we hold the power to shape our energy future. So it's very, very powerful. I liked, I liked this paragraph. It's very powerful. It, it is engaging the readers. So again, this is sort of like how content creators and writers can leverage the power of ChatGPT. Now, in, before ending this exercise and to wrap up, I just wanted to highlight a key point. And that's not to blindly copy paste content generated by ChatGPT. This is very, very important and you need to pay very close attention to this. You need to ensure you do your own due diligence and check the facts if you're using generated content from ChatGPT and other Generative AI applications and tools. Make sure you're providing credits where they're due and you need to make sure that you're not infringing any copyright laws 10. 10 ChatGPT Emails: For our next exercise, we're going to use ChatGPT to draft e-mails for a few different scenarios. This is very powerful because ChatGPT will save you a tremendous amount of time by helping you use your time more efficiently and increase your productivity. Which is the whole point behind harnessing the power of Generative AI. For our first scenario, we're going to draft an email to follow up with a job application. So pretend that you just applied for a job that you like. And now you want to follow up with the interviewer to express your gratitude and let them know that you're very interested in the position. What we wanna do is to get ChatGPT to do most of the work for us and create the template for the email. And all we have to do is filling the blanks and the gaps. So here's the prompt that's going to help with this e-mail. I'm just going to paste that in here and we'll go through it together. So draft the follow-up e-mail to express gratitude and reiterate your interests in a job position you recently applied for so you can make it more specific and putting the company's name and you can put in the job title. But we're not going to do that. We're going to keep it a little bit more higher level than generic, and I'll show you why. So let's go ahead and run this Prompt. Chatgpt has finished generating the output for our emails. So now let's walk through it and take a look together. So here you can see this is the start of the emails. So Dear Hiring Managers name. So as you can see, ChatGPT has created majority or the e-mail for us, maybe pretty much 90, 95% of the email, which is great. So all we have to do is fill in the blanks. Obviously, you want to read over the email first to make sure everything is relevant and correct. But you can see that here we can just replace that with our interviewer's name or the hiring managers name or whoever this e-mail is actually meant for. Over here you can see that it says, I wanted to take a moment to express my sincere gratitude for considering my application. Here again, it's a placeholder so you can put in the jaw position you applied for. The company name that you applied for. Express your gratitude and appreciation. And then it just basically follow along. And this is, you can see there's gonna be placeholders with the company name, which you want to do. Now what you want to be careful, something I noticed about ChatGPT and Generative AI in general is that sometimes things can give very repetitive. You don't want to say the company's name 102030 times in an e-mail. So this is something just keep something you should keep an eye out. So go through this and once you actually finish filling out all the blanks, go ahead and edit this. Also. One thing, as you can notice that this email is a little bit long for someone who's going to be quite busy with their schedule as a hiring manager. So one other thing you can do is this is about 12345, about 56 paragraphs. And one thing you can do is you can actually go ahead and make this shorter. And because you already, ChatGPT already has context of this email from your previous Prompt. You don't have to be very fancy here. So all you can say is follow up with another Prompts, say, make this email shorter. Now, this is going to put ChatGPT to work. It already knows what email you're actually talking about. And it's just gonna make it shorter. So now you can see that the email is a lot shorter. It still has the same points that you are trying to make in terms of expressing your gratitude and just letting the interviewer know that you're very interested in the position and it's still very similar to the previous Prompt. It has the placeholders that we can replace. And at the end here you can put in your name and contact information. So this is how ChatGPT can help us craft Emails. Next, we're going to as ChatGPT to craft us an email that's going to be four type of event invitation. So imagine you are, want to invite an industry expert to speak, add maybe a conference, or maybe attend the discussion or debate. And this is where you can actually get ChatGPT to help you with creating that email. If you don't know where to start. For this, we're going to use the following Prompt as an example. So draft an email inviting industry experts or influencers to speak at a conference or participate in a panel discussion. So go ahead and run that Prompt. Let's walk through the email together here. So as you can see, ChatGPT has done a pretty decent job in crafting this e-mail for us. So there and then here you can put the person's name. And then here's a Introduction In this paragraph. They are placeholders that again, you can feel just like the previous e-mail and you can talk about what the event is, the date of the event, and then the location. And then here this paragraph talks about sort of like the value that person brings to the event and some of the opportunities available at the event and so on. And again, as you can see, this email looks a little bit longer. This is why you shouldn't just copy paste this straight out of ChatGPT. You should Read through at filling the gaps and then also try to make it shorter and more to the point. I'm sure there's some repetitive things here that you can take out. And then at the end, it thanks the person. And then here again, placeholders that you can replace with your name, your organization, or your title, and then the contact information. For our next scenario, we're going to ask it to generate an email around, say, a product or service Introduction. So imagine you are a company providing service to your customers and you just released a new feature or new product that you're very excited to share with them. And just sort of like as awareness, you wanted to send this e-mail out so customers know that this is available to them for use. So for this, we're going to fall, we're going to use the following Prompt. So compose an email introducing new product or service to potential customers, highlighting its key features and benefits. So let's go ahead and run this Prompt. Alright, let's go through this e-mail together so you can see it has generate the e-mail. But one nice thing is that it has Summarize the features in a nicely bullet point format, so it's very easy to read. Here are some key features or benefits of product X. And basically it just goes through the features here, feature one feature to feature three feature for. And obviously you would modify this to match your criteria, whatever they may be. And then following up. Also important part of the e-mail. It doesn't just talk about the features, but it talks about the benefits to you as a customer. So again, very important and really liked the formatting here in bullet format. You can, as you witnessed earlier, ChatGPT, you can control the type of output degenerates, and it has a lot of different ways you can generate the output. I can do bullet points, it can do graphs, they can do CSV file, and it can do tables. So you can control that in your prompt. In here we kept the very simple as we discussed it previously. But here we have the benefits and the rest of the email with at the end you got your name, your title, and your contact information. So again, very well-crafted e-mail for a few seconds is not bad. This also works. This application works really well. If you also have your brand or you have your clientele or customer base, or just the audience in general, and you'd like to send them updates or newsletters. This works very similarly to this and we can leverage ChatGPT for that as well. As the last example for this lesson, we're gonna get ChatGPT to help us in getting feedback. So for example, let's say you just worked on a project. And now you want to get feedback from either your colleague, co-worker, or if you're in school from your instructor or teacher. For that, we're going to use the following Prompt. Write an email to request feedback from a mentor, teacher, or colleague on a project or piece of work you have done or you have recently completed. So let's go ahead and run this. As you can see, ChatGPT has finished putting together the emails, so we'll skip all of this again there Introduction. But one nice thing is that you can see that ChatGPT actually again, formatted things nicely in a list format. So because this is a feedback requests e-mail, you can see that I kindly request your feedback on the following aspects. And the nice thing is ChatGPT. Not only it's actually filled this out for you and is written the content. But you could just use this as sort of like inspiration of what are the things that you can ask for. So you can treat this as a template. Two are just suggestions or ideas. Chatgpt is telling us that we can ask someone in terms of feedback. We can ask them for overall assessments. So how did we do overall? We can ask about strengths. So we can ask that person to give us the feedback in terms of what things we did well on that project, we can ask for areas of improvements and then any additional insight or thoughts or comments that that person may have. So this is very nicely formatted by ChatGPT. So again, the rest of the email and the conclusion. Now again, just like any other Prompt, you can control the tone, you can control the length. You can even tell you a Prompt. Make sure this email is no longer than 200 words or 500 words or whatever you want it to be, however long you want it to be. But hopefully, these scenarios have demonstrated that it's really easy for you to get help from Generative AI to craft somebody's emails and how much easier it's going to make your life and how much time it's going to save. And hopefully those prompts and scenarios recovered. You can note them down and you can use them in your daily lives. 11. 11 Summarize Books: In this lesson, we're gonna learn how to use ChatGPT to Summarize any book. And this is gonna be a huge help to save you a tremendous amount of time again, to become more efficient. So we're going to cover two different prompts. The first one is going to be very simple, and the second one is going to be a lot more tailored than detailed. And then you'll be able to see the difference. Now, let's walk through the output together. So as you can see here, ChatGPT did a pretty decent job in terms of summarizing this book. Given that the Prompt was very, very simple and not specifically tailored to anything. So you can see that ChatGPT has the Introduction and it talks about what type of book this is. So it's a self-help book that provides comprehensive guide to personal or professional development. So bit of an introduction. And you can see here, obviously the book was seven Habits of Highly Effective People. And interestingly enough, ChatGPT actually created seven bullet points and covered those seven habits in the seven bullet points. So this is actually they readable, very clear, very concise, and you can see at the end, it also has the conclusion and it's just again highlights what the book is about. So provides practical strategies and insights to help individuals cultivate positive habits. So this is great. And you can see again, it already ChatGPT was able to determine the best output, which was the bullet point format. So this is not bad. And considering a book that could be 300, 400, 500 pages, now, you can just take a couple of minutes, read through this and basically get the gist of it and be familiar with the book and the concepts is trying to present to you. For our next exercise, we're going to ask ChatGPT to summarize a book that's uptight fantasy. But for this one, we're going to craft the Prompt In a way that is a lot more customized to our needs. So let me go ahead and paste that here. So here I got this Prompt. So please provide a brief overview of the book, harry Potter and the Sorcerer's Stone, by this author. So we're giving it the title of the book and the author's name. And then it here where your prompt is actually becoming more customized compared to the previous Prompt. That was very simple. So your summary should be kept. Your summary should capture the main characters, central themes, and key plot points in three to four paragraph summary. So we're telling it while we want to see in the output and what the length should be. So here we said don't go over three or four paragraphs. You can even exchange this or replace this with something like 500 words or 1,000 words or something like that. If you want it to, then written in a concise bullet points. So we're specifying the output. Again, you can change this to something else, whatever you need, and then feel free to highlight the most impactful moments, major conflicts, thought-provoking ideas presented in the book. So this is additional information that can be included in the output of what ChatGPT is going to give us. And then again ensured that the summary provides readers with a clear understanding of the Books essence and significance. Remember to maintain the desired length and format specified above while generating the summary. Again, it's enforcing the length and is selling it. Don't go over what the length is. So let's go ahead and run this Prompt. Let's walk through to see what ChatGPT came up with. So as you can see here, it's finished summarizing the book, harry Potter. So this book is the first one in this series. Main character hair, Harry Potter is an orphan. And then he talks about the central theme revolves around the battle between good and evil. And then here it talks about some key plots in the book. And as you can see here, it's not too lengthy, so it's basically obey the rules of our length where we told it don't go over three to four paragraphs. Everything is nicely Readable, readable, and broken down into bullet points. And you can see that we asked it to tell us what the book signifies and that's exactly what it did here the book explores themes of friendship, loyalty, courage, and the power of love. And then it tells us that at the end it tells us that the book is basically setting up this stage for subsequent books to come in the series. So now bad ChatGPT did a pretty decent job. So hopefully these two examples showed you how you can actually use ChatGPT to Summarize any book in any category 12. 12 ChatGPT Summarize Text and Articles: In this lesson, we're going to learn how to use ChatGPT to Summarize any text or any Articles. And there are a couple of ways that you can do that and I'll show you both in this lesson. So first of all, I came across this article. Again, it doesn't really matter what article that is. You can use any article that you're interested in. I just happened to come across this one. So this one is from Stripe improving instant payouts by this author. So you can see that here he talks about the service and how they've improved the service. So one thing we can do in ChatGPT, instead of reading through this entire article, we can just get ChatGPT to summarize this for us. The first way we're going to do this is just simply through texts. Again, this could be really, any texts doesn't have to be this article, it could really be anything. So let's just go ahead and say, Summarize the following text. Let's give it a length in less than 200 words. Okay? And on the new line here now we can just go and copy paste this text. So let's go ahead and copy this. I will copy some more here. And you get the point. You just keep copying. You just keep copying any texts that you want. Summarize. So this is what we can do, very simple Prompt, as you can see, Summarize the following test Text in less than 200. So let's go ahead and run this Prompt and see what ChatGPT comes up with. Okay, let's walk through the output of ChatGPT so you can see that it has Summarize the, Summarize the article that we're interested in, in less than 200 words, which is great, very readable in 34 paragraphs. And here you can see that it just talks about stripe introducing when it introduced a service. It talks about some of the improvements and enhancements it's made to the service. It talks about the expansion in the US and so forth. So then at the end he talks about the feature along with recent improvements and wraps up the article. So this is one way you can Summarize any article or any Text. And next we'll take a look at doing something similar with any article, but in a much simpler way. So another way we can actually Summarize any article that we see on the Internet through ChatGPT is just asking it or providing it with the URL of the article, which is very simple. All you have to do is tell, give ChatGPT is similar prompts. So we can say Summarize the following article. And now what we can do is go back to that article from the address bar. We can grab the URL or the link, and then we can paste that here. And that's all we have to do. You can simply just run this Prompt and ChatGPT will summarize this for you. Okay, as you can see, ChatGPT was actually successfully able to find the article because you can see here is output outlines that the article improving instant payout on this stripe Blog, Blog discusses the recent enhancements made to their service so it was able to successfully find the article, interpret the article, and then Summarize it for us, which is great. Here you can just see he talks about the first improvement, takes about a, talks about another update or enhancement, and then talks about monetization. So as you can see, it has done its job in terms of summarizing it. So again, very easy. A few words you can Summarize any articles that you read or their Internet using ChatGPT 13. 13 ChatGPT Summarize Chats and Conversations: In this next exercise, we're going to learn how to use ChatGPT for Chat and Conversations summarization, whether you have Chat logs or conversational transcripts from a meeting. For example, you can use ChatGPT to generate brief summaries of the Conversations. In real life examples, Let's say you're off on vacation for a week and you come back to a ton of messages or notifications and have very little time to catch on. So this is something you can use your advantage. Another example is let's say there was a meeting that was recorded. You weren't able to attend that meeting, but there was a transcript of that meeting. So instead of going through and watching the entire an hour-long meeting that was recorded by Zoom for example, you can just use the transcript and get ChatGPT to summarize all the conversation for you. So these are just some real life examples. However, if you're a working professional, definitely first check with your company's security and compliance department just to make sure that it is okay for you to input this information into ChatGPT. So what I've done for this exercise is I've actually created a simulated Chat. So this is all made up. I actually got help from ChatGPT to generate this Chat. So it's just a conversation between three people in a meeting. John is the head of sales, Sarah is the head of marketing and MIs head-up Engineering at a company. And they've come together in a meeting to determine the budget allocation for next year. So again, this is all made up. I got ChatGPT degenerate this Chat. And now what are we going to do is we want to get a ChatGPT to actually Summarize this Chat for us. And let's first start with the Prompt. Let's walk through this Prompt together so you can see this is very explicit and it has a lot of detailed within the prompt itself. So please provide a concise summary of the following Chat, capturing the main points, key insights, and any significant decisions or conclusions reached during the conversation. So again, this is something you have to decide. What are you trying to get out of this summary? What are the things that are important to you from the summary of this conversation? These are the things that you can actually tweak in your prompts to match your expectations. Aim for a summary length of 150 words. Here we're setting the length and the size of the summary. Additionally highlight any contrasting perspectives are notable arguments, and then identify any unresolved issues or areas that require further exploration. So this is very important and if you need to know if anybody was blocked in a certain area and you need to help resolve any issues or remove any blockers. This is exactly the type of thing you need to include in your Prompt. And then again, a little bit of reinforcement. Ensure that the summary is well-structured, coherent, and accurately represents the essence of the conversation. So this is a very well-crafted, detailed and explicit Prompt. Now all we have to do, I have colon here and then we're gonna go to the new line and I'm going to paste the Chat. So the Chat is going to be quite lengthy and long. So I'm not going to we're not going to walk through that because there's no point. But I simply just paste that here. And you can see this is just a conversation between those three people in the meeting. So let's go ahead and run this Prompt. Here is the summary that ChatGPT came up with by analyzing that conversation. So you can see in the meeting, John, Sarah, and MR. Discuss the budget allocation for the upcoming year. John emphasize the need to prioritize sales and suggests that increasing the sales force and investing in the marketing campaigns there argued substantial budget to increase marketing, to create brand awareness and MI stress the importance of investing in Engineering and R&D to enhance the products Functionality and staying competitive. So here you can see that each person is fighting for budget for their own department. It says that they debated the priority among sales, marketing and engineering with each person really just emphasizing budget for their own department, which is usually how these meetings go. And here ChatGPT, the nice thing is because of our Prompt and the way we structure than craft there are Prompt. It has a section on contrasting perspective, so it tells us John prioritize immediate sales growth. Sarah emphasized long-term brand-building and MR. Stress, innovation and customer satisfaction through Engineering investment. So again, this is very good and unresolved issues. It basically ChatGPT analyze the conversation and it figured out that the meeting concluded without a decision, the team needs to gather more data and metrics to support their arguments and find a middle ground that considers the needs of all departments. The discussion will continue in the next meeting. So as you can see here, this is very powerful and very amazing because if you, if this meeting took an hour for those three people to talk and not Region conclusion, you are able to read the summary within a couple of minutes and basically find out what the meeting how the meeting went, what the issues were, what issues came up with, issues did not get resolved, which issues did get results. So it's very powerful and you can just quickly, within a couple of minutes, you can go through the summary and figure out what the next steps are. 14. 14 Prepare for Interviews: In the next exercise, we're going to learn how to use HAGPT to help us prepare for interviews. So let's say you've applied for a specific job or role at a specific company, and now you get a call to come in for an interview. Ideally, it would be great to practice interview questions with another person, but now you can actually accomplish this with CHAPT with a few tips and tricks. So let's go ahead and get started here. For this exercise, we're going to go through three different prompts. So the first prompt is going to be very simple and generic. And I'll show you that in a second. The second prompt is going to be a little bit more detailed and tailored. And the third prompt is going to help us make the experience more interactive as if you're there in the room with a person and that person is asking you a question, waiting for you to respond, providing feedback to your responses, and then moving on to ask you the next question. For our first prompt, this is what we're going to use. So again, very simple and high level and generic. I would like your help to prepare for an interview, simulated job interview scenario where you are the interviewer and I'm the interviewee. Ask me ten questions throughout this interview, but do not provide a response. So you can obviously adjust some of these to your liking. You can make the number of questions less or more. You like, and you can ask it to provide a response if you wanted to see what another person, if it wasn't you, what would they say, for example, in this simulation. But because you're trying to practice the interview, you would want to be the one who provides the responses. And that's why I have the prompt tailor this way. So let's go ahead and run this. Okay, let's go through the result of what Cha IPT has produced for us. So you can see that this is the simulated interview results. So interviewer being HachiPT in this case, they're basically saying good evening or good morning, good afternoon. Thank you for coming in today. And now they're starting to ask questions. And as you can see, they're starting nice and slow. Can you tell me a little bit about yourself? And then there's really no responses here provided because we asked not to give us a simulated response. So it's just purely asking us questions. So what interest what interests you about the particular role and the company? How would you describe your strength and weaknesses? Can you give me an example of a challenging situation you encounter at work? So these are sort of more tailored towards behavioral questions. How do you prioritize task and manage time effectively? This is definitely a crucial skill to have in any job. Tell me about a time you had to deal with a difficult co worker or customer. This is a good because it happens a lot and it's a crucial skill to have again. So, as you can see, Chachi VT has curated a list of ten questions for us, and this is perfect because this is generic enough that regardless of what role or company you apply for, some of these questions will come up. So I would say, take your time, go through these questions, and think through your previous experiences, either in school, on projects, other works, working with other teams and come up with answers to help you better prepare for the interview. Okay, for our second prompt, we're going to build it's still going to be static. It's not going to be interactive, but we're going to build up on top of our first prompt and make it more detailed and more tailored. So let's go ahead and I'll paste that in. So let's go through this together. I would like your help to prepare for an interview, simulated job interview scenario. So up to this point, it's all the same, and this is the additional part here. So the title of the job I have applied for is program manager. Again, I just made this up. The company is Microsoft. The industry is technology. I would be working in the Office 365 team as part of this role. So I'm giving it what role I'm applying for, what company it is, what sector it is, and which team I'm going to actually be working in on if I was to get accepted for this role. So let's go ahead and run this prompt. Okay, let's walk through the results here. And as you can see, now the questions that Chat GPT is asking us is a lot more specific and tailored to the role we specified in the prom. So you can see that. Let's go through the questions, a few of these questions together. So can you please start by telling me about your experience and background in program management, right? Because it's a program manager role. So what specific skills and qualifications you possess that make you suitable for program role at program manager role at Microsoft, how familiar you are with the Office 365 Suite. So again, very very specific to this role that we defined. And there's some really, really good projects here. How do you approach prioritizing task and managing resources? That's great because that's something that program manager has to do on a daily basis. And here, this is a good one. Program managers often used to work with multiple stakeholders such as developers, designers, and execs. How do you navigate conflicting priorities and communicate effectively? This is one of the crucial keys and responsibilities of their role. So again, good questions being presented to us by Cha ChiPT to help us prepare for the interview. Obviously you can go ahead and tailor this or edit or modify this prompt to your liking or to your needs based on whatever role it is you're applying for or the company, and you can just modify it and it'll give you the more specific you make it, then the better results you're going to get from the output. But here, as you can see, it was fairly simple, but we did we did give it some key attributes that help us tailor this specific to our needs as a program manager candidate or the role we are actually applying for. So again, take some time, go through these questions and work on the answers because when you go to an interview, they will definitely be asking you, not all of these, but at least some of these questions. So this will be great practice. For our third prompt, we want to make things a little bit interesting, and we want to make the experience more interactive. So imagine you're actually sitting in a conference room or in an online Zoom meeting and you're actually talking to the interviewer. So an interviewer is sitting there right in front of you and they're asking you questions and you're providing a response. So very close to a human, like, interaction. And it makes things more fun. It makes things more interesting, and it makes you sort of think on your feet while you're being asked questions, as opposed to having practice prior. So Uh, in order to do this, I found this I found it a little bit challenging to get Chat GPT to wait for my responses or inputs. But I did a little bit of research and I find a solution online. So let me go ahead and paste the third prompt here. And let's walk through it together. So here I'm asking Chat GPT to be an expert in conducting job interviews. So you are an expert in conducting job interviews for companies in the technology sector. I'm being very specific. Simulate a comprehensive job interview for a candidate applying for the product manager role at Google. Again, I'm defining the role and I'm defining the company, and I've already defined the sector. Are going to follow the following steps. So now I'm giving you a number list of steps to follow. Introduce yourself, ask me a question, wait for my response. After each of my responses provide constructive feedback. Do not ask me more than ten questions. Again, you can tailor this to your liking or to your need. But I found that you also need this line here. So apparently this wait for my response doesn't just do it on its own. So I've added this one last thing here. Now introduce yourself, ask me a question, and wait for my response. So let's go ahead and run this prompt. As you can see, now, Chad GPT is basically introducing themselves. Obviously, there's just some placeholders that you can ignore. But you can see that now, unlike the last two prompts, it didn't just ask all the questions and give us a list of questions, which is potentially what it did. It just gave us a list of questions to help us prepare for the interview on our own. Here, it's actually not doing that. It asks us the first question, but now it's actually waiting for a response. So what we can do is now we can provide a response and see what happens. And what it's supposed to do is provide some feedback, and it's supposed to ask us the next question or follow up questions based on what we said. So it's asking us, can you please tell me about your experience as a product manager and how it aligns with responsibilities of the roles at Google. So let's go ahead and I'm just going to make something up here. As a product manager. I have worked very closely with developers and stakeholders to build new features for our SAS platform. And I believe I would be a good fit at Google. Okay, so let's go ahead and run this and see what HAGBD does. Now, you can see that HAGVT is providing us feedback. It's saying thank you for sharing your experience as product manager. And then over here, it's saying that it's great that you believe. It's saying that your experience demonstrates your ability to work cross functionally and manage multiple stakeholders, which are essential skills of a product manager. It's great that you believe you would be a good fit at Google, as it's important to have alignment with the company's values and culture. So now it's saying now let's move on to the next question. So we can just say something simple like sounds good. And now Chat GPT is going to ask us the next question. So in your experience as a product manager, have you identified and prioritized user needs and requirements to drive product development? Can you please provide an example of a situation? So now it's asking us the next question, and this is where you can actually go ahead and think. You know, it's exactly like being in an interview and the person asking you questions, and you have very limited time to actually answer questions. So you definitely want to note these questions down in practice and prior to your actual interview. But this should demonstrate that you can use HAGPT in very creative ways to help you prepare for job interviews. And this is one very important and fun application of HAGPT. So this exercise is just going to go on, so you can continue to ask questions, and then because we told HAGPT not to ask more than ten questions, I'll probably ask you ten, and then it'll stop, and then that will finish the interactive interview here. 15. 15 ChatGPT Translation: Another powerful application of ChatGPT is using it to translate texts from one language to another. Now, there's many different scenarios that could happen. Let's say you're at work and you need to translate something, and that's not a language that you speak of, then that's very simple to use ChatGPT to actually get that Text, translate it. And let's say you're traveling, you are somewhere that you don't know how to say a specific phrase or speak the language, we can quickly pull out your cell phone and on ChatGPT you can put it into Text, have a translate to another language, and then show it to one of the locals there to help you with that specific need. For this. It's actually very simple to get this accomplished in ChatGPT, super-simple Prompt, nothing complicated needed so far. First Prompt, while we going to do is just provide a very simple phrase here. So let's say translate. How are you? To Spanish? As you can see, ChatGPT was able to successfully translate the text. So the translation of how are you to Spanish is comma status and that is correct. Very simple. And for our second Prompt, where we're going to do now is we're going to give it something a little bit longer. So I paragraph and I basically just got this from the article that we worked in the previous lessons. So this was again from Stripe improving in some payout. So I just grabbed this first one. So let's go ahead back into ChatGPT. And I'm going to paste in the Prompt. And it's gonna be a little bit different, but very similar idea here, translate the following texts to French, and then I just paste that in here. So let's go ahead and run this Prompt. Okay, ChatGPT has finished translating this to French, and this looks correct. And one interesting thing is that one of the reasons I know that ChatGPT is translated successfully because I've worked with translations in the past, is that some of these symbols here, such as the percentage in certain languages when they're translated, There's actually a space in-between. And you can see that exactly in this case, for the French translation, there is between the number and assemble percentage sign. There's a space. Same thing for here. So this is just one more indication that ChatGPT was able to successfully translate this to the language of our desire. And for our third Prompt, what we're gonna do is we're going to do something similar in the previous lesson. We're just going to ask it to translate an article. So let me go ahead and paste the Prompt Here. Translate the following article to Italian. And I'm just passing in the link to the article. This is the same article we used in previous lessons. So very simple. So let's go ahead and run this Prompt. Okay, it looks like ChatGPT was able to successfully translate the article to Italian. You can see the Introduction and then you can see the conclusion if you want it to be sure. We can always between transitions, you can go from one language to another. So just to test things out, we could do is we can just grab this paragraph, which is the conclusion. We can just say. Basically translate the following text to English. Paste that in, run the command. You can see that it was able to successfully translate that to English. So this is how you can use ChatGPT to translate text from one language to another 16. 16 ChatGPT Write Code: In this lesson, we're going to harness the power of Cha GPT to help us write code. Now, this section is meant for software developers and software engineers. So it's going to be quite technical. If you're not a technical person or you just have no interest in this area, feel free to skip this lecture. We're going to cover a couple of different scenarios. We'll start simple, and then we get into a more complex example. For our first prompt, we're going to do something very simple. We're going to ask Chat GPT to write us a function to reverse inklist. And this is a very common interview question as well. Linklist is simply a linear data structure, and we're going to ask hatGPT to do this in Python. You can really choose any language, but for the sake of this example, we're going to use Python. So my prompt, again, is going to be very simple. Write a function to reverse a inklist in Python. So let's go and then run this prompt. Let's go ahead and take a look at the result together. So if we scroll up here, this is amazing. So you can see that CHAPT has actually written the function for us, so we haven't written a single line of code. We got HAHPT to do the work for us, which is the whole point of using the application of generative AI again, to become more productive and use our time more efficiently. So you can see that HAGPT has created the function for us in Python with this class list node and not only this is where things get very interesting. And not only has hatGPT created what we asked it to do for that function to reverse the link list, it also provides us a description. So it says, This function takes the head node of a link list and inputs and returns a newly head node of the reverse link. So it tells us exactly what the code is going to do. So that's one of the great things about hat GPT and writing code for us. And here you can see that it even provided us an example of how we could use it with populated values. So here you can see that it actually did it actually went ahead and provided. It gave it a inklist, so it starts with the headnote of one and then 2345. And then it runs through the code that it wrote. And as it's running through that and reversing the list, you can see that it's printing the results, and this is exactly the output with those example or sample values. So you can see that the original link list was ordered this way, one, two, three, four, five, and the result, which is the reverse link list is now 54321. So this is pretty cool. Again, with one line prompt, we were able to get CHGBT to write all this code for us, explain what the code does, and then show us how it can be used. For next scenario, we're going to do something a little bit more complex. So imagine you are a back end developer, and this is something you would do on a daily basis, but imagine that you need to create a function that makes an API call. So in this scenario, let's say we have an endpoint users and you want to make an API call, that would create users, and the method for that would be post. So let me go ahead and paste in the prompt that I initially wrote here. So let's walk through this together. We want Chat GPT to write a function to make an API call in TypeScript, so I'm telling you what I wanted to do. So write code to make an API call, and which language, which is typescript, the request URL is this, so I've just defined I made this up, but the endpoint here is slash users. The payload should include the property name and email, and it should be passed as arguments to this function. This API call should be asynchronous, so I'm telling you I'm trying to hint that it should use Async and await, and then it should use the fetch library to make an API call. There's tons of libraries out there when you want to make API calls, and there's a reason why I'm using Fetch here. So let's go ahead and run this front first. Alright, it looks like CHAPT was able to create the function for us to make this API call. So let's quickly go through it. So async function and then function name it happened to be make API call. Again, not great, but you could easily change this on your own, or we'll get HAGPT to change it for us in a second, and I'll show you how. But so it created the function with this function name. It basically did exactly what we told it to. So it provided the two properties as arguments to this function. So name and email. Here's the request URL, so it's creating a variable for that and we have the endpoint slash users. It's creating the payload, so the payload will be the name and email that we're passing in as arguments. And then here it basically is doing a Trcche to make the call to the API. So it is using the fetch library to make the API call, and we got a wait here, which is nice. And so it is actually and I guess it happened to detect that this is post, which is kind of like what I wanted this function to be because I wanted to create a function to make an API call to the user's endpoint to create a new user, and it just happened to pick that up. But usually, you could have other methods such as get and delete, for example. So in this case, it's post, which is what I wanted, but I can easily change that if that's not what I'm looking for myself. So, yeah, I did a pretty decent job putting this function together. So try to make the API call. And then, you know, if the response is not okay, I'll throw an error. And then if it happened to catch an error, it'll log the um, issues so that we can troubleshoot later. So this is great. Also, I really like this feature, copy code. You can just click this and copy the code, and you can just paste it in your code, whatever your ID is or code editor that you like to use, and then sort of move on with your task. Now, one other thing is here, again, I really like this feature about CHAPT. It explains exactly what it is. So it says in this example, this function takes two arguments. It constructs the payload, then it uses fetch to make a post. So this is very similar not exactly properly formatted in a technical documentation format, but it's pretty much documentation of what this function does. And then it says this function awaits the response and checks if it's successful. If the response is not okay, an error is thrown. Otherwise, it logs a success message. And here it tells you just an example of how you can call the function, so with some sample values here, and it says here, remember to replace the URL with the actual endpoint that you want to call. Now, one thing is, I think JG BT did pretty decent job in terms of creating this function. But one thing I'd like to illustrate here is that you can still customize this to your preference. So let's say here, I don't really like the name make API call. I want to change that to something that's more reflective of what I want the function to be and what it does. Also, let's say I now changed my mind, I don't want to use the Fetch library, I want to use Axios. So let's go ahead and basically alter that by the following prompt. So what I'm going to say is rename the So rename the function to create user. And here, I'm also going to say, use the Axios library to make the API call. So let's go ahead and now run this command. Right. As you can see, Chat GPT has made the adjustments that we asked it to. So we are using the Axios slip right now imported Axios from Axios and you can see that it has renamed it from Ma API call to Create user, which is great. And obviously, it has rewritten the function, which is shorter and simpler, which is, again, great and very efficient and readable. And not much has changed, so we have the request URL, payload, and then the tryCatch. The only difference here is now it tries to it actually uses Axios to make that post call to that URL with those payloads. And then over here, if you look, it actually even updated the documentation for us. So it says in this updated version, we import the Axios library and use Axios post to method to make the post call, which is exactly what we wanted to do. After our adjustments, the function awaits the response logs of success. And here, again, just a sample code step it on how you can go about calling this function. So create user with some fake name and fake email. This is also something I really like. It says, remember because we changed libraries from fetch to Axios, it says, remember to have Axios installed in your project. So it's telling you exactly what command to run to install that dependency before using your code. So again, this is great. This is how you would use HAEPT to produce code stampets for you throughout your everyday work. 17. 17 ChatGPT for Documentation: In this lesson, we're going to learn how to use ChatGPT to create documentation for our code. Now, there's several ways of doing this. For example, in our last lesson, we use ChatGPT to actually create us a function that would help us create a user by making posts call, an API posts call to a specific endpoint via specific payload. Now, because ChatGPT remembers the context of the Chat and interaction with you, one thing you could do is you could simply continue that conversation here from the previous lesson. So in the last lesson, we got ChatGPT to create this create user function for us. Now all we have to do because it has contexts as what that function is. You can just go ahead and put it in the Prompt and say something like create technical documentation for this function. Now for the purposes of this example, I'm just going to start a new chat just so that I can sort of illustrate. You can do this with any piece of code or code snippet. Now, before doing so, I'm just going to reuse this example. So I'm gonna go ahead and copy the code for this function. Now I'm going to start a new Chat, and this is what my prompt is going to be. So create technical documentation for the function below. Okay, then I'm going to paste. Then let's run this Prompt and see what ChatGPT comes up with. Right, let's walk through the results here so you can see that ChatGPT has created the technical documentation for us. So it starts by the function name create user. And then it basically tells us what it is. So create user function is an async function that sends a post request to a specific endpoint. Here's the syntax or function definition, which is great. Here's the parameters, so the arguments of the function name and email, and it tells us what it returns, which is really nothing in this case. And then here's an example usage, so it tells us how to make that call. So create user function with some sample fake data here. And then the try-catch. So it talks about dependencies in this case, acts years was one of the dependencies that we need to install if you want to use it in our code to make API calls. And then here's some fully detailed and very concise and clear readable Functionality 4 h. So it says function constructs a URL, creates a payload of with those arguments, attempts to send a post request, and then so on. Then at the end here there's no that says respond data can be handled and processed by uncommenting the irrelevant lines of code. So yeah, that's fine. I guess we didn't really even need to copy paste. We could have taken this out and then have our own code in here according to our needs. But I think this is exactly what this is saying on commenting relative lines in the code and modifying it according to specific needs. Yeah, So this is great, this is pretty good. So we go ChatGPT to create technical documentation from our code by just simply providing the function and not much more information. Obviously, we can make things a little bit better by providing more context, but this is pretty good for minimum effort. As you witnessed in the previous scenario, ChatGPT did a pretty decent job in terms of creating technical documentation given the minimal amount of information we provided. Now in the second scenario, one thing I wanna do is I want to go one step further and made things a little bit more technical so that the Documentation ChatGPT produces adheres to the correct and industry standard. So let's go ahead and do that. And what I'm going to do is I'm going to reuse the same function create user, but I'm going to start a new Chat here, and this is what my Prompts going to be. Create, create technical Documentation for the following function. Now, I'm going to give it some industry standard here. So I'm going to say using open API, specification and swagger, which are very well-known in the tech industry at the moment. So let's go ahead and I'm going to again pace in the function. I'm just going to get rid of these comments here because we don't need them. And let's run the command and then see what ChatGPT comes up with. Well, it's pretty impressive. So if we scroll up here, you can see that ChatGPT was able to create exactly what I asked it to ask the two for the API documentation here according to those specs. So create user function and here's the opening APIs with specific and Swagger doc. So you can see that this is the version of Open API right now, three-point, Oh, and here's the title. So basically all you, while you can do is you can also just copy this, save it to a file, and then you can host it wherever it is, you host your documentation files. You can even include this execution in your CI CD pipeline if you really wanted to. And yet access to GBT APIs So there's a lot of applications that you can leverage here with this. But I just wanted to show you how amazing ChatGPT is in terms of accomplishing some of these daily tasks. So you can see that the Documentation is produced and create it. And you can see the path here. So this is the path for that. Here's the URL, here's the summary, create new user. So this is interesting because I never told ChatGPT anything about this function. And it being purpose for creating a new user. All I said was I need a function to make an API call. And interestingly enough, ChatGPT was able to figure out what this function is going to be used for creating a new user, which is what my initial the intent was. So this shows you how powerful ChatGPT is and its capability in interpreting what you want it to do. So this is just very impressive. And then it talks about, there's all these things. So if you're familiar with Swagger, an open API spec, this sort of format. This is a YAML file, so there should be very familiar to you and you can see the types here. So properties name, the type is string, again, description the name of the user. I did not mention any of these. Chatgpt was able to interpret and populate these for me. And then same thing again for email. It's of type string, the format is email and then this is the description. These are both required fields. And you can see some pre-existing, pre-populated responses here, 200,400.500, depending whether your request is okay, bad requests or internal server errors or if the APIs or down. So yeah, this is how you can get ChatGPT to create technical documentation for your code according to specific standards. One last thing I wanted to demonstrate here is that you can alter this if you want it to. So for example, because I want to create a new resource with this function, one of the responses I'm interested in is actually to O1 to 01 is the industry standard or HTTP response for success of entity Creation or resource Creation, I should say. So here, we don't have that in the list. So what I can do is I can provide a prompt to ChatGPT to see if we can alter this. So let's go ahead and say, add one more response to the list of responses. This should be two or one, and the description should be created. So let's go ahead and run this Prompt. There you have it. Chatgpt was able to understand what we're saying and did exactly what we asked it to. So everything is the same here. If you look through this, but the only thing we're really interested in is the responses. And we add it to add the new response in this section, and it was able to successfully do that. So the response code, status code is gonna be to a one ended description is going to be created. This is exactly what we asked to do and this is exactly what it did successfully. It even have a small description here in terms of explaining what it did. So I have added the new response with code 201 and a description and create it to the responses section. So this is amazing 18. 18 ChatGPT for Tests: In this exercise, we're going to learn how to use hat GPT to write different types of tests. Now, in order to speed things up, I'm just going to reuse some of the code that we previously generated in our last exercises. So the function create user that makes an API call. And we're going to run through two different scenarios here, one unit test and one integration test. So let's get started. For the first prompt, we're going to tell Chat EPT to write unit tests for the function. Now, I've already copied the function to the clipboard, so I'm going to write my prompt first here. So, let's say, write unit tests for the following function. And then we want to have some requirements here because these are unit tests and we don't want the test to actually make the call to the API, we have to make sure it's being marked. So we want to specify that the API call should be marked in the tests. And now I also want to specify which frameworks which testing frameworks Chachi PT should use in order to write these tests. So I'm going to choose just these tests. And I'm just going to go in a new line here. I'm going to paste in the function. So this is our prompt here, and then this is the function. So let's run this prompt and see what Cha chiPT comes up with. Once again, HHIPT has done a fantastic job in writing these tests here. So let's quickly walk through the output here to make sure everything looks and seems correct. So you can see that ChIPT here provided an example and how you can write unit tests using Jess. Now, let's quickly look at the code. So here, HHIPT has written a code, snap it here. So this is basically your test file in JavaScript, so you can just create a new file in your project. Using whatever ID or code editor you're using, and you can simply just copy base that in that file and then run the test. But let's quickly run through the test here. So it's mocking Axios and then here's the describe your typical describe it syntax. So describe is the test suite and each its statement is a test case. So here we have the create user. Obviously, if you wanted to, we can rename this when we copy paste over or we can instruct Chad GPT to go ahead and edit this to whatever we like. But let's just leave this for now. And then over here, there's after itch hook that tells Jess to clear all moks after each test has run. So now let's actually look at the test itself. So it says it should make a successful API call. So here, it's trying to write a test, or it has written a test that basically verifies the happy path. So it is creating a mock response here with the Sats 200 and the message. And here it's mocking the post call with some payload here. Now we call the u we call the create user, and that's going to get intercepted. It's actually going to be a mock. It's not going to make the actual call. And here you can see the assertions. So first thing, it's saying that the post call that was made, it was called once, which is correct because we called it once, and then here it verifies the request URL. So it verifies that the API call was made to the right request URL with the right payload. That's what this test is verifying. And then also, it says, expect console D log to have been called. Right? And then the console dot error should not be called because this is a happy path it's going to pass. And so this is all correct and everything is great. And it's written one other test here, so another its statement, and this one is actually testing the failure. And you can see that here again, we are setting up our mocks here. We call the function, so we are purposely going to make this test fail when the call is being intercepted. So we make the call. We simulate a failure using this test and the setup, and then you can see that it says the call to the post it has been called one time, so this is correct. It has been called with this request URL payload, so same as the first test. But note here that it's saying expect the console log not to have been called because if we go back to the code, the original code, we have the try catch here, and the console log, API calls successfully, this is only being called if the call is successful. And in the cat where we catch the error, this is where we call console error. So this is why it's saying that console dot log should not be called and if you go here, the last assertion here says, expect the console error to have been called with this message, which is exactly what should happen in this function in a case of a failure. It hits the catch block and then it will do a console error with that message. So these tests are actually great from coverage perspective. So it wrote one to test the happy path. I wrote one to test the failure. And this is exactly if you're a software developer engineer or an automation engineer or a QA engineer, these are the type of tests that you would write when it comes to the unit test category. For our next prompt, we're actually going to get Chat GPT to write integration tests for the same function. Now, as you may know, integration tests actually make the API calls. So in unit tests, we don't actually make the API calls, we mark them. But in integration tests, we actually want the API calls to be made. So they're real API calls being made to the server on a database that we stood up in our CICD tool. So let's go ahead and provide this prompt so that CHAGPT can get started with writing integration tests for us. So my prompt is going to be, again, somewhat simple. Write integration tests for the following function. The API call should not be mocked, so I'm providing it this requirement to make sure it's a real API call. And then again, I'm going to use Just for this, use the Just framework for writing these tests. Okay, let's go ahead and I'm going to paste in the function. So again, this is our prompt. So let's go ahead and run this and see what ChachPT comes up with. If you take a look at the results of this test, you can see that I think this is really important. So it says to write integration tests for the given function without mocking the API call, you will need to set up a test server. So this is where HGPT actually isn't able to recognize that we want to actually have integration tests. So what we'll be doing is we generate the code here in ChIPT then we'll copy pasted into our project, and then we check that code into the repository, wherever that may be GitHub or whatever, Bitbucket. And then based on that, that's going to kickstart the CI tool, and then we set up our test environment, and then we run the test and that makes a call to the server and database that we stand up in whatever container in the test. Uh, again, depends on what your setup is and how you have things set up for your project. But here, Chat GPT wasn't able to recognize that, so it's saying that it's still using a tool to mock the intercept and mock the actual HTTP request. So it's using Knocke as you can see, again, another popular tool that's being used for testing. And again, this is going to be very similar to because this Mg response is not really a real response, so it's still going to be somewhat of a unit test, not integration test. But the tests themselves are going to be very similar. I mean, the name here says it's integration test, but it's not actually making the call, so it's not a real integration test here. But if you look at the test here, it's still testing one for happy path and one for failure. So again, very similar to our previous test for unit test. One thing we're going to try to do now to see if ChachPT can fix this is we're going to give it a little bit of an instruction here. So in the follow up prompt, let's say, we're going to rephrase this a little bit and say re write the integration tests or actually here, it says that to write integration test for a given function without Mk, you will need to set up a test server. So let's say we have a test server already set up, re write the integration test without Mo without mocking. Let's just see what JAGPT does. All right. Now, if we take a look at the revised code, or in this case, the tests, you can see that actually HAGPT was able to successfully accomplish exactly what we asked it to first time around. So it says, If you already have a test server setup, you can write integration tests without mocking by using actual HGTV request to the test server. So that's exactly what we wanted to do the first time we provided the prompt, but it chose to actually mocking using knock. And now with a follow up prompt, we were able to fix that. So there's one lesson here. You don't have to get everything perfectly right in your first prompt. You start with a simple prompt, and then you can always build up on top of that, or you can customize things further with follow up prompt. So that's really the lesson here. And you don't have to be sort of too concerned or too worried about getting everything perfect with the first prom. You just want to provide enough context so Cha GPT can get started. And that's exactly what happened here. So let's quickly take a look at the revised test. So you can see that it's importing Axios and now here, there's no more library Nock or Just Mok that actually is trying to intercept the HTDB request and mock the response. So here, it says, describe, so this is the test suite name, create user integration test. This is actually pretty good, not a bad name. And here's building the base URL. A let server. So as you can see in the before all hook, it's actually creating a client using Axios, which is great. This is pretty good. So it does that once, and then after all, any cleanup, there's no code here, but if you wanted to clean up resources, this is where you would do that. And now we get to the actual test. So again, very similar to the previous lesson. I wrote two tests, one for successful happy path, successful call and one for failure. Let's take a look at each. So you can see that it's constructing the payload with some fake sample values over here, and then over here, it's actually making the call. So this is a real API call. It's using Axios post to make a call to that endpoint with this payload and here are the assertions. So it's actually asserting on status quo 200 in the happy path, which makes sense. And then here it's asserting on the response data be this case, there's really only one message being returned, so it's a starting on that. So these are great verifications, and this is exactly what you would expect from test to test the functionality of the happy path. So as you can see, Chat GPT is very powerful when it comes to creating test for a specific class or function or really any type of code snippet that you wanted to provide as long as it's reasonable and it makes sense. And basically, now that you have the code, it is really simple because it has done majority of the work for so all you really have to do is you can just grab all of this, right? You can just grab this file. This is a JavaScript file, so you can just simply copy this code or just copy paste the portion that you want here. For example, we just care about the Happy Path test here. It's already including all the setup. So all you have to do is really copy this. Go into your project. If let's say it's a JavaScript project, for example, go to your project, create a new test file, paste this code in and make the necessary adjustments or customizations due to your preference, and you're ready to check in the code and have the test actually run in the CICD pipeline. 19. 19 DALLE Introduction: In this section we're going to learn how to use DALLE-2 and how to unleash the power of DALLE-2 to create realistic images and any type of artwork that we desire. We're also going to explore different features in DALLE-2 and how to use them. Now when it comes to the Setup for DALLE-2, you don't really need much. In fact, if from the previous section on ChatGPT, you already created an OpenAI Account, then you can use the same Macallan to use DALLE-2. So there's really nothing else for you to do. But if you haven't already created an account with OpenAI, you can just go ahead and navigate to openai.com. Once you're here, you can select on product. You can select DALLE-2. And this will take you to this page. And then if you already have an account, you can just click login. And if you don't have an account, you can click on sign up register using your email address. And then you can come back to login and login into use DALLE-2. When you log into your OpenAI Account, you're taken to this page and as you can see here, you have several options. The first one is ChatGPT. So if you want to use ChatGPT, you would click this and that will take you through the ChatGPT application page. The second one is DALLE. So if you want to use DALLE-2 to create realistic images, then you would click on this and this would take you through the DALLE application. And then if you want to use the Open API, Open AI API to integrate with your application or business, then this is where you would go. Now, we know we want to use DALLE, so go ahead and click this. And now you're taking to the user interface for the DALLE-2 application. Now let's take a quick tour together and navigate through the different areas of the DALLE-2 application. As you can see in the interface, is actually quite simple. So there's really not much going on here, which is actually great, especially for beginners. But in the center here we have this is the place where you would input your prompts. So this is an input where you can put in any type of Prompt you want, depending on what type of images or artwork you want to generate. On the upper left-hand corner, we have a couple of navigation areas here, so we've got history and collection. If you click on history, history will sort of keep track of all the images that you've created in the past. These are just some other ones that I've generated for Fun and playing around with the application and learning the application by itself. And we also have collections. Collections is, this is where you can actually create public or private collections. And you can save your work to these collections and then share them. And if you go back over here on the right-hand side, on the upper right-hand corner, you also have options. If you want to do things such as buy credits, you can just go here, click by credit. And you can, as you can see here right now, 115 credits is for $15 and you can just click continue and follow the steps and pay for that and then gain more credits to be able to generate more Images. We'll talk about that. We'll talk about credit in a few minutes, but just wanted to quickly introduce the interface. So that's pretty much it. And over here, we'll talk about this again in a little bit. But these are just somebody Example artwork that was generated by users and submitted. And it's a great reference to use when you're thinking about ideas. Or if you want to spark your creativity in terms of both types of images, in which style or class you want to generate. Before we start to generate images with DALLE-2, I would like to spend the quake second talking about credits. So in order to use the alley and in order to input prompts and get images and use those images for your purposes, you need to have credits. Now, the way it works currently as the time of this recording is that when you sign up for an Open AI Account, you'll get some free credits for DALLE-2, play around with the application. But if you want to continuously use it day after day and generate hundreds of images, you'll need to purchase credits. And I found this article which is very helpful here. So this is article from OpenAI and it's at help.openai.com. So this is the article link address. And here it explains how they credits actually work. So what is a valid credit? It says you can use it daily credit for a single request and generating images through a tax prompts. So whenever you actually input a Prompt, you generate an image, it's going to use a credit. And it says that credits are deducted only for requests that returns successful Generation. So if there's a system error or something goes wrong and you don't get an image bag or a set of images back then, it won't. Dark credits. In this section, it talks about what free credits are and when you get them and when they expire. Here it talks a little bit about how to buy Dalley credits, but it's pretty straightforward. You just, as I showed earlier, you just click on the Actions menu or your profile and there's a button called by credits and it's a pretty straightforward process. And over here, I think this is, this section is pretty good because it talks about the difference between the free and PE credits. So it basically just says frequented credit expires one month after they were granted and pay credit expires 12 month after they were purchased. And this is a very important note here as well. Here it says you currently get the same set of rights, including commercial use. This is very important regardless of whether an image was generated through a free or pay credit. So this is actually pretty, pretty good. So if you want it to, you can use free credits and it would have the same sets of rights for commercial use as PE credits. So this is very important. Now, back in Delhi, I wanted to spend a little bit time talking about this section. I think that decision from OpenAI to make this section available was very good and very logical. Now, regardless of what your background or education is, there should be nothing preventing you from generating and creating very high-quality and realistic images based on whatever your imagination or Creativity is or for whatever use case you'd like to him to be. And I myself don't have I'm background in Art. I don't know the terminologies, and I find this section to help me a lot. So for example, when you hover over the, these samples, you can see the Prompt. And for myself, because I don't know the terminologies and I don't come from an Art background. I don't really know what this style is called, for example, right? But if I wanted to generate an image that's similar in style as this one, all I have to hover is to do is hover over. And now I can see the Prompt user used to generate this image. Here. You can see that it says an expressive oil painting of a basketball player dunking. So this, this type, this type of painting is actually, sorry, this type of painting is called oil painting. So this is actually what I wanted to get from this, right? If I scroll down, down here again, I don't know the style for this one. And it says it's a van Gogh's style painting of an American football. They're going down here. I don't know what this one is. So if I wanted to do this, it just says a cartoon of a monkey and space. So if I wanted to generate something along lines of that, I would use the term cartoon, right? And you can see, you can see all of this going down here. This is another one. This one, I don't know, but it says if I hover over, you can easily see that it says an oil pastel drawings. So that's kinda like the type of destroying. So this is, you can go through this collection. I find it very helpful and very useful when I want to figure out what type of Art I want to generate. So this is sort of like what I've referred to. If I am looking for a specific style. One thing I wanted to highlight before we start our exercises with DALLE is the key components to consider in your prompts to get the best output and results. There are four main ones that I always consider. What, the where, or the environment, this style and the type. Let's walk through a Example together to illustrate what I mean. So actually this one is a good one. So the fish, the 3D fish, so as you can see, the image is pretty good. And if we hover over the image, we can see that Prompt that was used to generate it. Now let's see if we can identify the components. So here it says it's 3D render of acute tropical fish. So this is the what, right? In an awkward, um, so this is the where. And then it says on a dark blue background. So here it's specifying the style. And then the last one it says Digital Art. So this is the type. So this is a pretty good Prompt. It includes the main components and as you can see, the results are quite impressive. Now, it doesn't mean that for every image you want to generate or every Prompt that you're going to input, you're going to have, you have to provide those things. You can absolutely just ignore them or not include them in your prompts. And here's an example here, so this is a good one here. So as you can see this cat on the moral cycle, if I hover over it, you can see the prompt for it. So it says a cat riding a motorcycle. So this one is pretty simple, but you can see that D, you just have to know that the less you provide DALLE in terms of your Prompt, they're more you rely on it to figure out the rest when you're rendering the output. And this is a good example of that. So we didn't really give it any style or type. We just said a cat riding a motorcycle. But as you can see, DALLE chose to do this as the type cartoon. So this is exactly what I wanted to illustrate what this example, so it really depends on how creative you want to get and how detailed you want your prompts we, and whether you want to have control over the environment, the colors, the style, and the type of the image you're going to generate with Dolly 20. 20 DALLE Examples and Exercices: It is now time to start inputting some prompts and see what type of images we can generate with DALLE-2. So let's have some FUN. Alright, for our first Prompt, we're going to try and get DALLE-2 generate a realistic image. So this is what my prompt is going to be. Generate a realistic image of a sunset over a beach with palm trees. So with this Prompt, we're going to explore the capabilities of DALLE-2 in generating realistic images of specific scenes. So let's go ahead and run this Prompt. Okay, Now that we have our images, let's quickly walk through some features available through their daily applications. So let's just say out of these four images, I like this one the best here, this third one, again, it really depends on your preference, but this is just an exercise and an example. So let's say I like this one Best. Now, if you hover over this, you will see this action menu, this dot, dot, dot on the upper right-hand corner. So if you click it, you'll be able to see some options here in the Action menu. So the first one is open a new tab, which will do in a second edit Image, which we'll cover in a different video. This is great generate variation. What this is going to do is this is actually going to generate four more variations based on this style of the image. You select that. So instead of putting in another Prompt, this is just, you can just click here, you can click Generate variations. And DALLE will automatically generate for more images that are closer to this style. So not to the other three pictures, but closer to this style that we're selecting at the moment. And lastly, you can just download the image. So this is exactly what you want once you're happy with your image that you generated in Dalley and you want to use it for whatever your use case is, whether you're in school or work or just a personal project, you can just download it and use it. And that's what the DALLE Features and it's very high-quality to very high resolution and high-quality when you're downloading this file is going to be a little bit larger in terms of size. Now, these are the options available. And what we can do is let's go ahead and click this one, opening a new tab. And that should open a new tab in the browser. And we have this one picture here. So it's very similar to this area except this one is just focusing on that one picture, you select it so it's a bit more isolated if that's how you want to think about it. So over here again, you have this picture. And you can also see the picture is also larger so you can see it better. And if there's areas you don't like here, you can choose to edit this or not use it. But over here you have the Download button. So if you click this, it's going to download to your device. Download the image to your device. Again, you can edit the image here, which we'll cover later. If you click variations, it will click for more variations of this type of image and this style, you can click Share and you can click Save. So if I click Save here, this one is we talked about collection earlier. So this is where you can either create a new collection and save it to that new collection. Or if you have Favorites already set up, you can save it to your collections. And if you click, sure, it's going to, You can make it, you can publish it and share it publicly if you want. So these are just some other features available in this view when you open the image in a new tab. So at the end of the day, it really comes down to what your preferences and which view you like to work. And this is Image and it's isolated view and a new tab. Or you can just simply close this and go back and do the exact same things in terms of functionality here. Now, while we're here, I wanted to show you this option here, generate variation. So one thing to notice that when we input this Prompt here, initially we use the credit, right? So as we're generating more images, we're going to use credits. So our Prompt, use the credit. And now if I go ahead and click this, this is also going to use a credit because it's generating for you Images. But I want to do that because I wanted to illustrate this Functionality, which is pretty cool. So just notice this styles of each of these images. And what are we going to do is we're going to select this one because let's say We'd like this one the most, for example, for the sake of this exercise. And what are we going to do is we want to see what DALLE can give us in terms of the same style but different variations of images. So let's go ahead and click that, generate variation and see what DALLE comes up with. As we can see, dahlias finished generating the variations for us. So we actually have four more, which looks great. So on the left here we have the original. So this is what, this was our base and what we started with from the previous round. And these four here to the right are the ones that DALLE created for us in terms of variation. Now if we look at these, some of these are good, some of them are not. So for example, this one, I don't really like this one because the water just looks weird there. No waves here is to 2D and the sun's to white here. Not, not a hint of orange. This one, the water looks a little bit better, but again, this sign is it doesn't have any hint of orange And this one is actually pretty good. I liked the water here. It's the color blue. It looks really nice and there are some waves which makes it more realistic. The sun has a hint of orange and stands out compared to the background here and the sky. The trees are, okay. I actually like the trees better than the original one, but I guess this is what it is. This one not so much. So I would say from the new four, I probably would say this one, at least for me, right? It's a matter of preference. So for me, I like this one, so it's pretty good. Now all they have to do is simply go ahead and download it. And I can use this for my, whatever my use cases or my application is. And again, let's say if this is the one that you like. But in my scenario, let's say I didn't like the trees. I can continue the same process over and over again until I'm happy, right? So I can just go ahead and click Generate variation and then it will generate for more for me based on this image. So now the starting point and the base becomes this one. And no longer, this is no longer the original. Just one thing to keep in mind is that every time you generate, every time you inputted Prompt or every time you generate variations from generated image, you are using credit. So keep that in mind please. For our second exercise, I'm going to use the following Prompt. Create a series of cartoon characters engaged in a lively conversation. So this Prompt focuses on the creative potential of DALLE-2 by asking it to generate a series of cartoon characters in a conversational setting. So let's go ahead and run this Prompt. We now have our results for this Prompt and this one, the first one, not so great. This one here also looks a bit weird. Again, remember this is Generative AI, right? So over time it's going to get better based on feedback and trading data. And it's just going to improve and get better and better over time. So not a huge deal. And if you're wondering why did texts on this looks weird, just something to understand is that right now, like Generative AI does not have the concept of Text or language and he can't really interpret any of that. So that's why things look weird here. The only texts that they can really understand is this one here that you put in the Prompt. But beyond that, it doesn't have any concept of language or Text. And over here like this one is actually the best one. The third one here based on the Prompt that we gave them. So this is something that we could use, for example, if this was for a children's book or some artwork for kids, I think this would be good. This one also looks a little bit weird. So I would say probably the only user one. Use a usable one here is this one. And if you wanted to, we can just make more variations starting with this base. So that's sort of the outputs and the images that degenerate for us. But again, the one point I wanted to highlight here is that by instructing DALLE-2 to create lively conversations, it gives us the opportunity to experiment with the generation of diverse characters, body language, expression, and enhancing our understanding in generating dynamic and engaging visual content. For our next exercise, I want to be a little bit more creative and I'm going to use the following Prompt. Generate a surreal landscape where gravity seems to be reversed. So with this Prompt, it helps us to explore the imaginative aspects of DALLE-2 by asking it for a surreal landscape with reverse gravity, we can dive into the realm of creative and fantastical visualizations. So let's go ahead and run this Prompt. So here are the results of our Prompt and they're quite interesting. So as you can see, you can experiment with generating unconventional landscapes, floating objects, and you can even defy the laws of physics. And this is showcasing the flexibility of DALLE-2 in producing imaginative and unconventional outputs. For this exercise, we're going to try to do something more artistic and we'll see if we can design a cover for a music album. So this is going to be my Prompt. Design an album cover for a futuristic electronic music album. So with this Prompt, it helps us to explore the intersection of DALLE-2 and graphic design. By requesting an album cover for futuristic electronic music album, we can experiment with generating visually striking and technologically influenced imagery. So let's go ahead and run this Prompt. So here are the results for our Prompt. And as you can see, all of these actually have texts denim and as we covered earlier, DALLE-2 doesn't have a concept of language or Text, so these are just sort of nonsense texts that you see here. But the important thing is that this should, there may be parts of somebody's images that you can use or they could be used as inspiration for you to start your own design. So for example, here, if you like this one, you could just download this image here by going through the download Functionality, maybe a cutout this text part here, and replace it with your own tax, which would be quite easy to do in a program such as Photoshop or are they designed tools? This one here, I, it's all, again, nonsense texts, but it's actually pretty cool in terms of reference. So this has a nice font, like, let's say this was your album name. And you can use this type of like, you can use a really nice modern font. It says using sort of like a neon light style. It also has a shadow. So while you can do is you can just use this as a reference and design yours from scratch, but following this same style. Again with these ones, if there's a parts that you like, for example, this one, the bottom part can be used. This one, the top part can be used. And then you can just put your own texts on the bottom using Photoshop or any other program that you like. So again, you can use parts of the image, or if you like, you can just use these as inspiration and follow the style and come up with your own design. And from the results shown here, we can explore the futuristic aesthetics, neon colors, abstract shapes, and other elements commonly associated with electronic music album covers with DALLE-2. And this exercise, I'm going to use the following Prompt. Create a realistic depiction of a classic painting re-imagined in a modern urban setting. This Prompt teaches us the concept of reinterpretation using DALLE-2 by asking for a classic painting green imagine in a modern urban settings, we can experiment with blending traditional and contemporary artistic styles. Now let's go ahead and run the Prompt. The output is quite impressive. Looking at the results, we can see how DALLE-2 can generate realistic renditions of classic paintings while incorporating elements of modern urban environments, allowing us to explore the potential of DALLE-2 in Art and cultural imaginations 21. 21 DALLE Features: In this lesson, I want to cover a couple of features that Dalley provides to us and the end-user. So one is the upload Functionality and what is the edit? So DALLE, he has recently introduced the editor that we can use to edit our images, whether they're through upload or through the regenerated, through prompts. And currently the Editor is in better mode, but there's still some cool things you can do with it. So let's get started. To start with, we're going to try and upload an image. And all you really have to do here is click upload that image and that's pretty much it. That will ask you for a prompt to select the image. And let's say I want to select this one. So this was one that we created in one of the earlier lessons. And this is just the beach with the sunset and palm trees. So here it asks you whether you want to crop or skip it. If you wanted to crop, you can just choose the portion you want and click Crop. In this scenario, I'm just going to leave it as is and say skip cropping. And now this is actually taking me to the editor, Dalley editor. And as you can see, it's quite simple, but there's still some things you can do here. So first of all, if you're happy with the way this is right now, it's right now. 1024 by 1024 pixels. If you're happy with this, just go ahead and click the check mark. And as soon as you click the check mark, these things and the Prompt will become available. So note that this is an edit Prompt, Ok, it's not degeneration Prompt. It will generate things for you in the edit, but it's the edit Prompt for Generation. And over here, you have some options here. So you have the select cursor, you have the pan, you have eraser, you have a Generation frame, and then you can upload more pictures. So this is a very Basic Editor, something that we could do as we can even add more images to this if you want it to. So let's go ahead and click this. Let's select this one for example. And if you want it to, we could combine images, use parts of images if you want it to. So for example, let's say I'm going to make this really small and bring this down here and then hit Okay, and now we have this image incorporated into this image. And one other thing we can do is we can add a Generation frame. So Generation frame is something that we can place here on the existing pictures. And once we placed the Generation frame, we can go ahead and put in a Prompt. And when we put in the Prompt, it's going to generate over or it's going to fill in where our Generation frame is. So for example, if I just put some description or Prompt degenerate an ocean, it's going to fill in this entire generation frame. So that's another capability that DALLE provides to us. But yeah, that's pretty much all there is to it, to the output Functionality, it's super simple. You can use the Upload File Functionality to just upload a single image and edited. Or you can upload multiple pictures and use certain parts of each picture. Or just simply use Generation frame if you want it to. There's even an eraser tool here so you can erase parts of an image and just keep certain parts of the image. And then again, this also becomes sort of like you're Generation frame. So when you go to do the putting your prompts in the edit, input, a will also fill in this portion as well. But we'll cover this part in the edit functionality, but that's pretty much it for uploading files. In this lesson, I would like to demonstrate how you can edit certain parts of an image and how you can use a Prompt to regenerate those parts are the image. So let's start with uploading an image here. I'm going to use our sunset image that we created in the earlier lessons. So I'm just going to leave it as is skip cropping. And there's obviously an endless amount of things you can edit in this photo. But to keep things simple, what I think I'm going to do is I'm just going to say Okay to this. I'm happy with this. And I'm gonna grab the eraser, and I'm just going to erase this sun here. So let's get rid of that and maybe let's just create a blank canvas here. And one thing I would like to say is, now I want, now that I've erase this part of the image, I wanted to replace it with a different son. So now this is where I can use the edit Prompt to generate or regenerate that part of the image. So let's see how Dalley handles this. For my Prompt, I'm going to say, I'm going to try something more creative and interesting. I'm gonna say 3D, render of the sun and then comma Digital Art. So let's go ahead and run this like this uppercase D, Just in case, let's go ahead and run this Prompt Okay, So we have our results. So let's take a look. So the first image, not much is visible, so this is not great. You can click on the right and left arrow to go back and forth. This one, not so bad, but can't really tell if it's 3D or not. This one is still to this or not that great. And this one is just to the bot further away. So again, the results are, this is still in better and the results aren't, aren't too bad. I would say probably, maybe this one is the best. So let's say this is what we want it. And then we can go ahead and click Accept. And once we click except we have access to this and we can go ahead and save this. There's also on new Functionality, Functionality there. And again, we can put in more prompts or you can just come over here and click on Download. And this will download the image for you. One last node on the edit functionality and the DALLE editor is that you don't necessarily have to start with uploading an image before you can actually edit something. You can simply just use the Prompt input here to generate an image. And once you have your output, you can just choose one of those images. And from the Actions menu, you can enter the editor mode and edit that image. So very user-friendly. You can even look at your history over here on the right-hand side and select something that you have already created in the past. So let's do another couple of examples together for PFK-1. So let's go in here again, these are the ones that we created and let's say we really liked this one. So we want to make one more edit. So let's go ahead and click that. And you can see from here, you can just go to Edit from here. Or you can simply, while you're back in this view, you can select and go Edit Image. So multiple entry points to the same functionality here. So let's go ahead and edit this one. Okay, now that we're in the edit mode, we already tried something with the sun. Let's do something a little bit different. So let's grab the eraser tool. And what I'm going to do is I'm just going to erase this part. Okay? And then as soon as I do that, you can see the edit Prompt becoming visible. And it says describe the entire desired image, not just the erase data. So this is actually a very good tip. So here what I would like to say is, let's edit this a bit. So let's say man, surfing in the ocean near a beach with palm trees and as sunset. Okay, So this is a good tip. So let's go ahead and generate this. So here's the new output by DALLE with the edit and regeneration, which actually looks pretty good. So let's start looking, going through this and start looking through the outputs. So this one is not that bad. Not so great either because the human is not looking very real. Let's look at the next image. This one is not bad. It's a bit cartoonish with some green. Some green water or the ocean mixed with blue doesn't go too well. This one is not bad, is really, really small and hard to see. And this one is actually it's not that bad, It's a bit small, but you can see that the person is trying to serve in the ocean. So very interesting. And the cool thing is you can see Dolly was able to maintain the, all the aspects of the original image. So nothing's really changed. What the palm trees with the sun, with the ocean, just the area that you cleared and we erased. And it was able to place to the best of its abilities with the surfer, which is what we described. Okay, let's just do one more for funds. So let's say we like this one better. So let's go ahead here and then we can click, we can either create variations or edit this are shared or do whatever we want with this download and save it to a collection. But let's go ahead and do one more here. So what I'm going to do, I'm going to make more of a drastic change. So what I'll do is I'll just erase the poem every, all the palm trees here. So let me get rid of all of this. I'm gonna get rid of this. So basically no trees are visible and I'm also going to get rid of the sun, will try this one more time to see if he can produce a better result. So really we have the ocean and we have the surfer. And let's see what we can come up with here. Now I'm going to paste in this Prompt for the edit. So just again, getting a little bit more creative. Man surfing in the ocean near a beach with, instead of palm trees, I changed the type of trees to pine trees to see if we can generate that. And a sunset with a 3D sun. So I've also changed that. Just curious to see how DALLE is going to handle this. So let's go ahead and run this Prompt. Alright, so these are the final results and as you can see, it did its best to interpret our Prompt. And some of these actually look quite funny. So this one a try to put in some 3D type pine trees here, the ocean and the surfer or remain the same, the background remains the same, which is good. It was able to maintain those aspects of the image. Here again, we have some palm trees. These trees are not really, we can't really tell what type of trees they are. And this is, yeah, these are close to pine trees. But you can see the sun, the interpretation of DALLE of the 3D sound wasn't that great. Mind you, I can probably do a little bit better in terms of my Prompt. I'm providing a better Prompt to help us understand. And I can also provide feedback in terms of whether this was what I was looking for or not. So yep, but overall, this is just wanted to illustrate the ability and Functionality of the DALLE edit and editor features. So I hope this was helpful and you'll learn something in this lesson. 22. 22 Midjourney Overview: In this section we're going to cover and learn about Midjourney, what it is. While you can accomplish with this tool, we'll go through the sign-up process, will go through different subscription plans that are available to you. Different types of prompts that you can use to create different types of images and artwork. And then we'll cover the basic and advanced features through hands-on exercises. Midjourney is a Generative AI tool that can generate incredible and realistic images and artwork. Any use case that you can possibly imagine Midjourney can help you generate images for. It is built on top of the stable diffusion model and it uses discord as the user interface to the tool. Midjourney BOD is available in discord and you can use prompts to help provide description of the type of images you want to create. Prompts can be very simple or they can be very complex depending on how creative or specific you want to get. Let's take a look at some examples to give you an idea in terms of what you can accomplish with Midjourney. Okay, first thing we wanna do is just launched the browser. And what we wanna do is head over to midjourney.com. So www.midjourney.com here. Once you're here, just click on this button here called showcase on the bottom left-hand corner. So go ahead and click that. So here's the community showcase. This is where people have posted the examples here in terms of what they've generated. Use the prompts and generated in Midjourney. And these are just posted here as examples for you to go through and get some inspiration from. So as you can see here, these are amazing pictures and just like DALLE, this is a great point of reference if you want to come here and take a look at specific photos and get some creative ideas or find the style that you really looking for. And once you have an image that's kinda close it in terms of what you have in your mind. This is where you can just hover over the picture or the image. And then you will be able to see the Prompt user actually wrote to use and generate this image in Midjourney. Okay, so let's just quickly go through a couple of these just to see what the prompts are looked like. This one is quite basic and interesting. So as you can see, the images sort of like 2D posts or types. So if I hover over it, you can see the Prompt Here, minimalist flat design poster. So it's actually talking about the type of image, in this case A1. The person wanted this to be a poster type of artwork. And over here it's talking about the location. It's talking about sort of like the type of photography. So in this case landscape, right? And also interesting enough, the specified the time of the year, summer. So that would influence the image. As you can see, there's lots of light, sunlight on the building here. And that would definitely like the environment is something that would influence or even the weather, for example, would influence the output. Over here, this is a good one. So here we have a puppy. So the prompt for this, it says a cute puppy playing parks sunlight 3D. So as you can see here to all those have been incorporated into the final image here. And then interesting, it says that Disney Pixar studio. So over here you can actually specify in the Prompts, you can specify the type of movie or TV show or even name the actor. And it will add the AI will try to match the expectation in the Prompt as closely as possible by going back and using that data to influence the output, again as closely as possible to what you've specified in your output. If scroll down, this one is a good one. So as you can see, this is a more sci-fi and space sort of image. And if you hover over it, you can see it says aliens that film, movie. So actually this user named exactly the movie they want the image to match in terms of type and style. So this is interesting fluid photography and then close up. Close up again is another type of photography. So close-up, macro, side view, front view, bird's eye. So these are definitely do affect the final. Those are the options and parameters that will affect the final output. And you can easily incorporate those into prompts. If we go a little bit lower here, these are all great, great images, lots of examples and you lots of inspirations from, let's see here we got a sushi one here. This is a close-up magazine quality shot off sushi rolls. So again, this is interesting. So close up again is the type of photography shot and then magazine quality. So kinda like what you would see in a magazine. This is trying to match that. So Midjourney, try to match that in the final result. Scroll down here. Here's another food bonds. So this is Cheesecake. If you hover over, it says a close-up magazine quality shot of a luscious New York style cheesecake. So and then over here you can see food photography. So again, the type of photography is incorporated in here. Scroll down a little bit more. This is a good one, so you can see it says inch photoshoot style close-up. Again, the angle shot and rain drops. So this is talking about sort of like the weather, which you can see in the illustrated for you here. And in the city at night. So the environment and also the time of day can definitely help influence the image that you're looking for. And this is a good one as well. So this one boy is walking with his pet raptor in the leash city locations. So the environment and this Prompt actually illustrates one point that you can actually include the lens type in terms of the type of photoshoot you're looking for. So in this case, it is Canon EOS R7. So if you're familiar with photography and different types of lenses, you can also include those in the Prompt because the AI model has contexts and it has training data. So it knows what. This actually clarifies what type of photoshoot or photography lens the output should closely match. So again, a great example of what you can include in your prompts. 23. 23 Midjourney Account Setup Subscription Documentation: Alright, now that you know what type of images and outputs you can generate with Midjourney, it's time to get started. And in order to start, you don't really need much, you just need an email address. And all you have to do is register registration is for free. And you can just head over to midjourney.com again. And while you're here, all you have to do is click the Join the better button. So go ahead and do that and just follow the instructions. If you can. If you already have an account with Discord, you can just go ahead and login. If you don't, then just create one. You will receive an e-mail, you accept the invite, and I already have an account, so I'm just going to login and all you have to do is follow this process. And I'll see you back here in a second. Once you sign up and login to discord, you have to allow authorization for the Midjourney bot to your Discord account. So here all you have to do is click Authorize. And once that's done, we're now logged in and we can start using Midjourney. Once you're here, all you have to do is click on join the discourse to start creating. And this should take you to this chord. Here you just accept the invite and now they should open the discord app. And then we should be able to go ahead and start using the Midjourney and interact with them. Midjourney bought. Alright, we have our Account Setup and we're logged in. Now one thing to note is that there is this section called new commerce rooms. And over here you will see rooms that start with the label newbies. This is where you want to go when you first register an account. Over here, you can actually use Midjourney and provide your prompts to create images and artwork. And you can also see what people are actually generating. What other people out there like myself are, you are going to be generating using Midjourney. And again, it's a pretty cool place to just look at other people's artwork and look at their prompts and just get different types of inspiration. So it doesn't matter which one you click Add. You can just select a random one. I'm just going to click on this one here. And as you can see here now, we have entered the room and you can immediately see the pictures that people have generated. So this motorcycle is an example. Over here. We've got some more Generation, we've got some more Generation. And you can also see the Prompt that people use. So for example, this one, it says motorcycle has Ant-Man Logo. So this is actually pretty cool AK resolution. This dash, dash S is for Seed and we'll talk about that later. But just wanted to let you know that this is where you need to come in in the newbies room to start using this. Now before we start generating some examples and go through some hands-on exercises, I wanted to go over the Subscription plants. And this is straight from the Midjourney Documentation. And they have several Subscription plans that I would like to cover. So you know exactly how much time you have with the free trial and then what are the other what are the other available accounts and what they provide? Now, here, basically, there are three Subscription plans. We have the Basic, we have the standard and we have pro as the time of this recording. Of course, these could change in the future. Also, if you just creating a brand new account, you have free trial. And as you can see here, you have a 0.4 h of lifetime. So again, this doesn't renew every month. This is just a onetime trial that you can use Midjourney and see if it fit your needs. And this is something you want to subscribe to you in terms of plan. But you can see zero-point for our, this is equivalent to approximately 20 or 25-minute of generation time. And just to give you a perspective, it approximately takes about somewhere close to a minute to generate each image from the time you provide Midjourney the prompts. So about 60 s low web, sometimes less, sometimes more. But that means that this should give you roughly about 20 to 25 Images plus or minus a few, give or take. Alright, Next we have the basic plan. So the basic plan is $10 a month. And this one will give you 3.3 h of Generation time per month. Now with the paid plans, you also have the add-on features. So let's say you run out of this, but you only needed a few more images you can use. You can use the top of features, so you can pay $4 for an additional hour. Let's say you didn't need as much as the standard plan. You just need a little bit more on top of your basic plan. You can also use the add-on feature to top up. And one thing I'd like to bring to your attention is that you can pay in two different methods. So you can pay monthly or you can pay annually. If you pay annually for the Subscription, you actually get a discount. So if we scroll up here, you can see that there's actually a description about that. So Midjourney has three Subscription tiers, pay month to month, or for the entire year for a 20% discount. So if you look over here, you can see that the basic plan is $10 a month if you pay month-to-month or if you want to pay for the entire year, then you get it a little bit cheaper, you get it for $8 a month. So something to keep in mind when you're purchasing your plan? The next plan is the standard plan. So this one is $30 a month or if you purchase annual subscription, you get it for a little bit cheaper, so $24 a month. And this one's pretty good. This one gives you 15 h of Generation per month, which is quite a lot of images. So if you're a frequent user of the Midjourney for whatever your use cases or your purposes where there's for school, work projects, personal projects, whatever it is, then this this is a recommended plan here because you get a lot of Generation time here with this. And of course you have the Pro Plan which is $60, again, 48 a month before with the annual discount. And this one gives you 30 h. And again, if you need anything above that, then you can use the add-on feature to top it up. And there's also one other feature with the pea plants here. You can also rate images. So there's a feature too, so that you can read other images in the platform. And that should give you, that should earn, use some free GPU time. So these are sort of what the Midjourney plans are as of the time of this recording. Just a couple of more notes on the subscription plan before we move on. So if you scroll up here, you can see it's very easy. The Instructions to subscribe to a plan, it's very easy. So it says how to subscribe. And all you have to do is use a slash subscribe command to generate a personal link. And it'll take you to the Subscription page, or you can simply go to midjourney.com account. So if I click on this, this will take me to this page here. And over here. This is where you can actually select whichever plan you want. So again, if you want to pay you early, you get it out a little bit of this council, $8 a month build yearly. If you want to go one-to-one, then just click this button here and then it will be a little bit more expensive, so $10 a month here. Now, one other thing is if you scroll down here, you will see that there's some frequently asked questions here and there's an important one here. So these are just a typical questions. If we go down here, this is very important. So it says how does commercial use works? So in terms of copyrights and privileges and ownership. So it says, if you have subscribed at any point, you are free to use the images in just about any way you want. So basically what this is saying is that as long as you have a Midjourney Subscription and then you generated the images using that Subscription. You are free to use it in any way that you'd like. So this is pretty good in terms of ownership and copyright. So very important thing to know. And this is one way to sort of get to this page. Another way, another way that was instructed is we can just simply go here. We can type in subscribe and we just have to click Enter. This will generate a link for us, personal link, and then you click it. And then it will take us to this page as well. Again, just two different ways of going through the same page to pay purchase a subscription. One last note before we get started is that I wanted to bring your attention to this page here. So this is the Documentation page four, Midjourney. And this is very, very helpful, especially if you're getting started. And throughout the course and this section we will refer to this different parts of the documentation. Or you have to do is navigate to docs.midjourney.com. And it will take you to this page. So there's a QuickStart guide. There's lots of very, very helpful and useful information and they should answer some of the questions that you might have. And there's different sections here using this code User Guide. For example, on their user God UGA version. So if you go here, you can see that it tells you that default model version is set to version 5.5. And obviously this will sort of, it was released on May for 2023. And this will keep getting updated with the new new releases and newsletters and release notes and so on. And what we just covered earlier was this in the Subscription that can be found under the subscription plan. So if you go on there Subscription section and you click Subscription plans, that's exactly how you can actually find this on. Again, all the instructions are here on how to subscribe. So very helpful Documentation page and there's lots of useful information. And just like a subscription plan, we'll come back and we'll refer to this from time to time. 24. 24 Midjourney Example 1 Image Generation and Basic Functionality: It's time for us to start having some fund with Midjourney and see what sort of images we can produce. So let's do something funny and entertainment to start the process. So let's say, first thing you wanna do is you want to click on the Discord message part here. Now, you can just start with the Prompt. You have to say imagine this is how Midjourney works in terms of inputting the Prompt. So you can see when you go slash imagine, then you input your Prompt and it says that description says creates images with Midjourney, which is exactly what want to do. So let's go ahead and do that. So imagine and then our Prompt, Let's say something along the lines of cute hotdog with sunglasses. Very curious to see how Midjourney is going to process this one. Now, the type let's say I want is 3D like and let's give it a location. So let's say Hawaii. And then I want this to be somewhere on the beach. And then time of the year, let's say this is gonna be summertime and the weather condition, let's say sunny. So let's go ahead and run this. Now, as I mentioned, Midjourney is going to take about oh, it says that it's not accepted if you have to have to accept the terms and conditions. So let's go ahead and do that except terms of conditions. Please note that the terms and condition Prompt only shows up once when you first create your account and once you agree to wet, it won't show up. Again. Andy can proceed with creating your images with Midjourney. So let's go ahead and input our Prompt. Alright, let's go ahead and put it in our Prompt here. So imagine acute heart with sunglasses 3D. We want this to be in Hawaii, somewhere on the beach. Time of your summer and then we wonder whether it'd be sunny. Now, I'm going to add in one more argument at the end here, and that's dash, dash as space 800. And you can leave this out, but let me just explain what this dash dash stands for and that's stylized. So if we head over to Midjourney Documentation, if we go on their user guide parameters, over here you can find stylized. So go ahead and click that and here it explains what it is. So it says, this Midjourney board has been trained to produce images that favor artistic color composition and form. The dash dash stylized or dash, dash S. So either one you can put in your Prompt, I just use the shorter form here. It says it influences how strongly this training is applied. A low value produces images that closely match the Prompt, but are less artistic. And high values create images that are very artistic but less connected to the Prompt. In my case, because I'm looking for something more creative that's not too close to the reality and a bit unorthodox. I want this to be less connected to the Prompt, and I want it to be more artistic. That's why I tagged dash, dash slash space 1800, excuse me. That over here you can see that the default value is 100 and the integer it accepts the integer of anything in this range. So 0-1 thousand. So as you can see, 100 is sort of on the low-end of the range, meaning by default, it will stay closer to your Prompt and be less artistic. And as I mentioned in my case, because I'm looking for more creativity and I want it to be more artistic and less connected to my Prompt. That's why I put a higher number there. So this is what it is and you can just try this out without the stylized 800. So let's go ahead and run this Prompt now. Okay, Now this is going to take about a minute to generate. So what I'm going to do is I'm going to pause the video here and come back when it's done, so I don't waste your time. One thing I'd like to mention is when you generate prompts and while you're waiting for your image, because it does take awhile to generate. You want to keep an eye out on your image generation process because there's so much noise in this Newbie discord channels that you might lose track. So just keep an eye out. And because this is actually how a lot of people go in these channels and they generate a lot of prompts and Images and it just stacks up. And then it just a lot of noise. So you might have to keep an eye out on your image, your particular image that you try to generate. Now let's go scroll up and be found hour. So this is actually the prompts. So you can see this was our Prompt and we were able to, It was able to generate this Prompt. So we have a couple of options here. Now, let's actually analyze this a little bit so you can click on this. And if you click on this, it shows you the image. But then there's another button here called open-end browsers. So if you click that, It's going to open it in a bigger and better resolution so that you can actually see what the output is. Taking a quick look at the result. First thing we notice is that Midjourney generates for images for every Prompt. So this is to give you some sort of variation and see which one you want to start off with. And by default, as you can see, these are all square shape, but we'll talk about that in the next exercise and how we can change some of these aspects ratios and make them non-square. For example, we can have them in rectangle or other aspects ratios. Now looking at this, given our Prompt, it wasn't our focus was more on the hot dog, but I guess the interpretation of the model and the application was more around the animal dogs. So looking at this, it's actually not bad. So he got most of the things correctly. So for example, in this one there's a dog with the sunglasses on. It's on a beach. You can see the ocean and sky and the background. And you can see that it's pretty sunny. Same thing with this one here on the lower right-hand corner. This left one here. This is actually not bad because it has the picture of the dog. It's pretty much everything with a bit of motion blur at the end there. And there's also a hot dog sandwich in front of the dogs. So this is actually not bad. But you can see that this is sort of like the interpretation of the AI model of our prompts. So you could do, You could actually go ahead and retry this again. And while we can do is you can regenerate. So if we go back to, let's go ahead and close this. And if you go back to discord, one thing you can do here is you can click this button here. So this is basically saying regenerate based on the exact same prompts. So let's go ahead and click that. And now this is going to regenerate for more options for is, it's very similar to us putting the prompt again, but it's just easier and a clickable button. So let's see what it gives us. Okay, we have the result of our regeneration. So if we take a look here, these are actually not bad. These are pretty cute. So these are cute dogs with on the beach during summertime. And this one, again, they're all wearing sunglasses, so this one has a hot dog in front of it. So yeah, this is actually not bad. And this is basically like our first exercise just to get things started. And again, just represents the interpretation of the AI model. And if we go back here in our next exercise, we're going to try some different prompts and then it will cover what these things over here mean and go through some of the features that Midjourney provides 25. 25 Midjourney Example 2 Aspect Ratio: Before proceeding to our next Prompt, one thing I wanted to mention is that one of the nice features about having a paid subscription plan compared to a free trial is that you can actually chat with the Midjourney bought directly. In fact, there's another option where you can create your own discord server and you can invite them Midjourney BAD to that server. But for now, we're keeping things simple because there's so much noise and so many prompts by so many other people in the newbie channels. And it's sort of gets cluttered. You have the option to just chat directly or privately with the journey. But, and that way, things remain a lot more organized and you can easily find things when you scroll up and down. In order to do that, all you have to do is just any Prompt in the channel. All you have to do is right-click Midjourney BAD here, and then just go message. And now you can actually just interact directly with the Midjourney by for this exercise, what we're gonna do is we're going to generate car or a picture of a car using our Prompt and then go through some other features that are available in Midjourney. So let's start by saying imagine. And over here let's create a Tesla car. So let's say Tesla Model Y. You can again put anything. Now, one thing I wanna do is let's try some different settings or tell them Models, some specific things we're looking for. So let's say cityscape. So sort of like your city landscape here is the environment I'm looking for. And then for atmosphere, Let's give it sunset. And then one thing I'm going to tag along here is the aspect ratio. So dash, dash, dash 16 colon nine. And if we head over to the Documentation here, you can see that on their parameters. You can see Aspect Ratio. So if you click on that over here, it's sort of gives you the definition of what this parameter is and what it does. So the aspect or AR, so you can use any of this and you're Prompt. Obviously this one is shorter, changes to Aspect ratio of the generated image. So this actually help us not produced just square images we can put in really anything we want. And aspect ratio is the width to height ratio of an image. It is typically expressed as two numbers separated by colons such as 7443. And then over here it talks about 43 is, there are some common ones, 74169, these are all common aspect ratios. And our overhear it talks about how it translates to pixel. So now let's go back to our bot here, and let's generate this. Let's run this Prompt and see what kind of outputs we can get. The results are now ready. And as you can see here, these are pretty amazing pictures. It did a really great job in terms of matching or the outputs here generating the output that closely match my Prompt. So one thing to quickly notice right off the bat is that these pictures are no longer squares, as you can see, these are actually more rectangle and that's because of the Aspect Ratio parameter be provided as part of our Prompt. And you can see that these are all pretty great. They're all Tesla Model wise. There are, have that city landscape in the picture. And then you can see some great, very nice sunsets here and some other reflections of the light. You can see on the car and on the road and on the street. So I think Midjourney did a pretty decent job generating these four outputs. Now that we have a good output to work with, let's cover some other features available in Midjourney. This one we already covered earlier in the lesson, and it's just a quick way or a shortcut of saying generate for more images that from the initial Prompt that'd be provided. So that's already covered. Now, let's talk about U and V. U stands for upscale and V stands for variation. Now, let's look at the talk about these numbers here. For both U and V, you will see one-two-three-four. And these are just corresponding to the grid here. So this one here, top left, this is number one, top right, this is number two, bottom-left, this is number three, and bottom-right, this is number four. So when you go, when you want to click u1 or upscale picture number one, you're referring to this one. When you click on V4, you're actually referring to this picture here. Now, let's talk about what U and V actually means. So when you click on you or upscale, you are telling Midjourney that you want that exact image. And Midjourney will produce a very high-quality, high-resolution version of that image with refined details. You can then open it in the browser and download the image and use it for your use case, whatever that may be. And you can even upscaled further if you'd like. Now, when you click on V, Midjourney will create for new variations of that picture using The original picture as a base. So if you recall in one of our earlier lessons, this was very similar dysfunctionalities, very similar to the feature in DALLE-2 that we generate four new variations based on the base image which we covered earlier in the course. Now let's try out these features. So looking at the original output out of these four pictures, let's say I like this one the best. So number four. So but I still don't want to use this one just yet. I want to see what Midjourney can generate, offer this one. So one thing I can do is generate four more variations based on this image. So because this is number four in the grid, what I can do is I can click V4. And what that's going to do is going to take this image as a base and generate four more variations based off of that. So let's go ahead and do that. The results are now ready. So if we quickly go and inspect this, they actually look pretty good. And one thing we can do is this is actually higher resolution. So if you click on opening browser, you can see a bigger image and you can even click. It's not fully zoomed in. So if you click and you can see the fully high resolution actual Image Size. So this is actually really, really nice. This one. So this is the first one. Very nice. Car has already parked in the parking. This one is also great. This one is number four, so this is okay. Not really liking this part here on the top. And this is the number three. So this one actually looks very nice. I liked this one. I liked the bridge on the top. And I like to buildings in the background. I like the color or the clouds and the sunset, the light reflecting into the clouds. I also like the sunset and the light reflecting off the car. Very nice. And then also the light on the street. So this is, I would say, probably my favorite one out of these four again, is just a matter of preference, and this is just for learning purposes anyways. So let's say out of these four, I like number three better. And this is now what I want to use for my use case. And now one thing we can do is let's go ahead and close this. And if you go back to this chord, I want to use number three here. So again, remember this was 123.4 and I want number three. So what I wanna do is I want to tell Midjourney to upscale number three, which is going to generate it, generate a very high resolution picture of number three so that I can actually save the image for my use case. So let's go ahead and do that. Click on U3. Okay, so Midjourney finished upskilling number three. And as you can see the results, It's the same image is just bigger and better resolution. So if you click on it, now you can see the image. And if you click open in browser, this will give you the full size. And now all you have to do is simply, you can just right-click and go copy image or save image as whatever you like. And now you have the image and you can use it for your use case. And you can use it however you like. Because I mean, in this scenario, I have a paid subscription and I generated this with a paid subscription. So according to the Midjourney Documentation, I can use this image however, I like 26. 26 Midjourney Example 3 Seed Functionality: Alright, now that we know how to use Midjourney and we're familiar with some of the Basic Functionality is I think it's time to start talking about somebody advanced functionalities. So to start with, we're going to take a look at the Seed Functionality. In order to get started, let's go ahead and generate an image so that we can work with it. So let's just create something very simple. Imagine let's say I want a baby ocean. Ocean with waves, blue and green water each summer. And let's just go ahead and generate this. Okay, so we have the results. And as you can see, this is what the four images that'll give us. I like the first one here. So what I'm gonna do is I'm just going to do apprise one. Okay, now we have the final image here, and I'm pretty happy with this. Now, let's say I just want to change something with this image. For instance, let's say I'm happy with the ocean and the waves, but now I just want to add a person to this picture. So for example, I want to add a surfer that's surfing in these waves and the ocean. But I want the output to still be very similar to this picture that we just generated. And this is where we can actually leverage the Seed Functionality. So let's head over to the Midjourney Documentation. And if you look on their parameters, there is a parameter called seeds. So let's go ahead and click that and see what that one's about. So here there's a description. So it says Midjourney uses a seat number to create a field of visual noise like television static as starting point to generate the initial Image grids. Seed numbers are generated randomly for each image. And I think this is the important part here using the same seat number and Prompt will produce similar ending images. So this is exactly the type of thing I'm looking for, for the purpose of this exercise and what I'm trying to accomplish with my final image. And one, there's some more information here about the seat numbers. This is also pretty important here and it says seat numbers are not static. I should not be relied upon between sessions. Now, if we scroll down a little bit in the Documentation page, you can see that there's some examples here. So they're trying to create an owl picture here. And over here they do that. And then for the second want, they are using the Seed and we'll experiment with that in a second. If you scroll down here now, it tells you how to find the jobs seat number. And it says using the Discord emoji reaction. So you can just see here there's a quick video that shows you how to do that. But basically it's saying that all you have to do is react with an envelope emoji, to a job, and then that should give you the Seed number that we're looking for so that we can use it for our next Prompt. So let's go ahead back into discord and do that. Let's go ahead and react with an envelope to this image to see if we can get that. So all you have to do is click this Add reaction button here. Now let's search for envelope, and then it's this first one here. So let's go ahead and do that. And you can see that we actually got we actually got two things back. We got job id and then we got the Seed. I just care about this Seed one for this exercise. So let's go ahead and copy that. I'm just going to copy that to the clipboard. Okay, Now let's see if we can use this seed number to generate our final image. So remember, I want to keep this as a base image and, or at least something very similar to this image. And just add a person surfing here. So while we're gonna do is we're going to provide a very similar Prompt but with slight modification here. So imagine. And then I'll say ocean with waves. And this is the part I'm just going to add one person surfing, just trying to keep it to one person and now multiple people, and then everything else will remain the same. Then at the end, I'm just going to do dash, dash Seed, and then I'm just going to paste that number. So this is very, pretty much the same as the first Prompt that be provided to get this image. Above. All I've added here is this part here and one person surfing. And then over here I tagged this parameter dashed I Seed, and then this is the seed of our original base image. So let's go ahead and run this Prompt. Ok, so the results are in, and this is not too bad. So we have four pictures with ocean and waves. And there's one surfer in each of them. Now I believe the bottom two here look closer to the image that we started with these, this one and this one, not so much, but these two, especially this number three here, it looks very close to this one. And I think maybe it might have gotten confused when I provided the word beach here because in this one you can see a little bit of the beach here and the sand. But in none of these we can actually see that and logically it makes sense because when you're surfing, you're further out and you can't really, you're not close to the beat. So maybe what we can do one last try is we can actually take that out and maybe simplify it a little bit and rely more on the Models. So Let's go ahead and do that. Let's produce a new image. So let's say imagine. And then we'll just provide the exact same Prompt here. So we'll say ocean with waves. Then blue and green water will just say summer and we'll leave it at that. So no beach in this one. So sorry, I spelled width here. And then Ocean with waves, blue and green water and then summer. So let's go ahead and run this Prompt. Okay, here's the result of the new one. And let's say I like number four here, so nice clear waves. And then let's just see if we can actually go ahead and add surfer to this one. So first let's do, let's apprise number four. So we'll click U4. They should create a new image and then we can grab the Seed from that new image. Now, let's find the seat number for this. So go ahead and react with an envelope to this one. Okay, great. So now we have the Seed number. Let's go ahead and copy that. And then for my next Prompt, I'm going to keep things very similar and I'm just going to add the surfer. So let's say imagine. Then we'll say ocean with waves. Here. I'm just going to say one surfer summer. And then we'll do Seed. And then this is where we'll paste the seat number. So everything is exactly the same except I just added the one server here and then Use the Seed number of the new MHC. So let's go ahead and run this. We have our results and if he quickly inspect this, we can see that probably out of all of this, I would say probably this one here. So number two is the one that matches the wave, the direction of the wave, and coloring and the texture and the waves of the base image. So if you go back here, probably, I would say number two in this result matches the one here. So this shows you how you can actually use the Seed Functionality. So hopefully this has helped you with that. If you recall, back to the earlier lessons, this Seed feature is very similar to dallies, erase and fill, degenerate and fill feature because in reality you could actually you generate an image, then you can use an eraser to erase parts of the image. And then you can provide a prompt that similar with slightly something different to make a new image off of that base image. So again, you could accomplish, this is very similar in terms of concepts. So you can accomplish something like that with Midjourney using the dash, dash Seed Functionality 27. 27 Midjourney Example 4 Combining Images: In this lesson, we're going to explore a couple of different ways of Combining Images. So let's say you have a few pictures on your computer or your device, and you want to blend them together. And we can actually accomplish this through several different ways in Midjourney. So let's get started. Now, in my case, what I've done is I've created two images in Midjourney, and they're both images of the cars. And what I thought it would be interesting to do is to try and blend two different cars together to see if we can create a new concept car, which is a pretty cool idea. So what I've done here is I've created and saved an image of an Audi Q4 e-tron, which is an electric SUV. That's going to be our image one and our for our image to, I've created a Tesla Model three sports, which is an electric sedan. And the goal here is to try to combine the two to create a new concept car. In order for us to accomplish this, we first need to upload our images from our computer or device, whatever device it is we're using into Midjourney and it's actually very easy to do. So all you have to do is click this Plus button and upload the file. And here I'm going to select both of these. So Audi Q4, e-tron, and Tesla Model three. So I'll go ahead and select both. I'll upload them, and then just simply click Enter. Now that we've finished uploading the file. So this code, we just have to grab the image links so that we can combine them using Midjourney. And it's actually very easy by default when you upload an image, scored puts them on CDN. So we just have to go and figure out what that CDN is. And all you have to do is select on this image here, Open it in the browser, and then you can simply right-click and go copy Image link. Alternatively, you can just grab it from the address bar, whatever is your preference. So I've gone ahead and grab the link for that and I copied it to the clipboard. So all I have to do is go ahead and type in imagine. Then I'm just going to paste the link and then space. And now I want to put in the link for the second one. So again, just go ahead and click on the Tesla Model three here Open and in the browser, right-click Copy Image link. And now we can close this and then we can just after the space here, we can go ahead and paste that. And now what we're trying to do is we're trying to blend the two cars together to create a concept car. And as you can see, this is actually a pretty simple Prompt. There's not much you have to do. Just provide the two images via their links and lead a Midjourney takeover. So let's go ahead and run this Prompt. Alright, we have the results and if we take a look, they actually look pretty good. The final car is actually still at Tesla, but we can definitely notice the modifications from the result of the blends. So if you take a look here, if we go back, you can see that in the original model pre-test slide here you can see, for example, the rims here. They are not the same as the output and the rims on the new concept car are actually more closer to the one from the Audi. And it's really the combination between the Tesla rim and the I'll do Aldi RAM, but the emphasis was on the waiting was heavier on the Audi on the final result. If you take a look here, also the front lights, you can take a look here that they are different. So on the test life you take, if you take a look at these lights, and if we go back here, you can see that this is more representative of the hybrid of the two because the light on the Aldi are quite thin here and the ones on the Tesla are quite wide. So the final product or the final blend, is actually sort of like the hybrid of mode. So you can definitely notice traits and characteristics from both cars here in the final concept car, which has actually pretty cool. There is another method to Combining to images in Midjourney and it's a feature that they've created for users and it's called blend. Now, let's go ahead over to head over to the documentation for Midjourney. Let's expand the user guide, the commands section. And then over here you can find blend. So go ahead and click that. And here it tells you what blend is and how to use it. So it says the blend command allows you to upload two to five images. So we can basically combine up to five images. Now whether you're using the blend command or the image URL or link via the previous method we covered You can do anywhere 2-5 images using either approach, which is fine and that's as of the time of this recording, they may increase that in the future. So two to five images quickly and then looks at the concepts and aesthetics of each Image and merges them into a novel new image. So it just tells you how to use it. Very simple blend. And it says is the same as using multiple Image prompts with imagine. But the interface is optimize for easy use on mobile devices. So either way your preference, you can choose either method to combine images or blend images. Here they have some options and talk about Options. And here they tell you how you actually can do that. And it's very easy. You just type in the blend Prompt and you can start uploading your images. So let's head back to this chord and see what we can accomplish with this feature. So for this lesson, I've actually created to Images and I'd like to blend them. So the first one is just the character or a man in the formal dress and wearing a suit and tie. And for the second one, I actually have created Iron Man. And I want to use the blend Functionality from Midjourney to combine these two images. So let's go ahead and do that. Now in Midjourney, while we all we have to do is just typing slash plant. And we can just select it. And as soon as you selected, it automatically gives you two placeholders for image 1.2. If you want it to do more, you just, you can actually, once you upload the first two, it will allow you to upload more images, as you can see here. So let's go ahead and do that for Image one, I'm going to pass in the this man here. And for Image Tool, I'm going to pass in Iron Man and I'm interested to see what the results would look like when Combining these two images. And let's go ahead and run this Prompt. As soon as I click here though, I just wanted to show that if you wanted to add more options, as soon as you click in the Prompt, it opens this dialogue box and then it shows you if you wanted to upload image 34.5, you can. All you have to do is click here. It'll give you another place holder. And then you can go ahead and upload more images if you'd like. So let's go ahead and run this Prompt. Alright, there results are now fully rendered out and as you can see, they look quite interesting. So you can see parts of the Iron Man armor is now integrated, are incorporated in this man's suit. You can also see that his facial like, he's become more cartoonish like or comic like. And his facial expression has now changed and is more serious, which matches as seriousness or the facial expression of the Iron Man suit here. So this is quite interesting in terms of results. And this should basically covered the Functionality blend and how you have two different ways of Combining Images. One node I would like to make here when it comes to blend or the Combining of Image functionalities in Midjourney is that you don't necessarily have to blend two objects together to have one final hybrid of the two, you could use the combined Functionality Midjourney to simply place an object in another picture. So here is what we're going to do in this lesson. Now. I've created a picture of a desert here with some hills and lots of sand and a lot of sun. And what we're going to try to aim to do as part of this exercise, is we're going to try and create this Tesla that we created earlier and combine it with the desert picture to see if we can place the Tesla car in the desert and see how that turns out. Okay, let's go ahead and grab the address link for this picture here. So just go ahead and copy image link of the desert. And then we'll close this. We'll say imagine, paste that in and then space. And now let's do the same thing for the Tesla car here. So let's grab this one and then paste that in after this space. Let's just know that any Prompts that said, let's leave it simple. So just to URLs separated by space. And then let's see how Midjourney actually is going to interpret that and what the final outlook is going to look like. Okay, let's look at the final render results here. And as you can see, it actually looks pretty good. So it was able to accomplish exactly what we wanted with minimal efforts. So you can see how powerful Midjourney can be here. So all we did is we just said, imagine passed into URLs of the 21 was the environment and the location. The other one was an object which was a car and a with a very simple Prompt and no modifications or configurations, it was able to interpret exactly what we want it. So here it plays the Tesla car in the desert and now this one's driving. And from here you can generate more variations. Use the Seed Functionality to actually make slight modifications while keeping the base the same. Or you could use more bland Functionality. So you can even bring in more objects here into this picture, whichever you like based on your preference. But over here you can see that you can, basically the point of the lesson is you can actually place one object into another environment or location using the combination of Images 28. 28 Midjourney Example 5 Image Weight: In this lesson, we're going to learn about a feature called Image Weight. So let's head over to the Midjourney Documentation here. And under user guides we want to expand this section called advanced prompts. And over here click on Image prompts. And if you scroll down here, here there's a section that covers Image Weight parameter. And basically what this is is that it says it used the image, use the image with weight parameter dash, dash IW to adjust the importance of the image versus the tax portion of the Prompt. And here it tells you what the available versions are. I'm currently on version 5.1, so this is, I can set this, the default is actually one, but I can set it anywhere from 0.522. And this is sort of like the range from min to max. And this is where we want to signify the importance of the image with the rest of the tax prompts. So let's go back to Midjourney and this gourd and see this inaction through an example. We're going to use a similar image here that we generated previously in 1 h lesson. So over here I have a picture of a desert with hills. And what I would like to do is I want to place car in here and then play around with the Image Weight parameters. So let's go ahead and do that. So I'm going to use the picture of this desert here. And let's type in imagine. In our Prompt. I previously showed you how you can find the image link. But there's another quick way of actually doing that and that's simply by dragging and dropping. So all I have to do is drag and drop that into the Prompt. And as you can see, it has that. It's able to populate the URL Image URL for me. So now let's go ahead and proceed with the rest of the Prompt. Now I just want to say Tesla model three sports. Then I'm going to tag along the IW or Image Weight parameter. And then let's start saying 0.5. So we're going to assign a low value for the Image Weight. So this is going to actually put more. It's going to signify the car tesla, and then it's going to reduce the significance of the desert. And what we're gonna do is we're going to try with this first, and then we're going to try with a higher number and see if we can notice the difference. So let's go ahead and run this. Alright, looking at the results here, I'd say this looks pretty good. So not so much. There's one there's not much represent, representation of a desert here, but number 12.3 actually look pretty good. So you can see that desert in the background and you can see the full car in the output or the final image. So now let's go ahead and try with a higher number. So signifying the desert over the actual car Tesla here. So let's go ahead and do imagine. Now let's go find our picture. Okay, so same picture, just going to populate the URL by drag-and-drop functionality. And then we'll say the same thing here. So space tesla, model three, sport. And then for the Image Weight, we're going to say 1.5. So last time we tried with 0.5, this time we're gonna do 1.5. So V are saying that there's going to be a significant way on the picture are the desert over the car tesla. So let's see what that comes up with. Looking at the result here, you can see that their friends for sure. So you can see that actually number four doesn't even have a car in it. I mean, that's how strong the emphasis was on desert, that there's not even a car here. Number three, I think Midjourney screwed up a little bit in terms of rendering the final output. So the car isn't that great? But the point here is you can see the emphasis on the desert. So a lot of this picture is just sort of dedicated to the desert part than the car part. Same thing with number one. You can see that The car is further and a lot smaller in the final picture. And you can see there is a heavy emphasis on the desert sand, the hills, and the background, mountains and everything. So this is how you can actually use the Image way, property and parameter to specify more significance or less significance in your mixture of your image and tax prompts 29. 29 Midjourney Example 6 Prompt Weight: In this lesson, we're going to learn about a feature called Prompt Weight. Now just like Image Weight, how you can change the significance of your image compared to the rest of the texts Prompt with Prompt Weight, you can change the significant of a portion of your tax Prompt versus another portion of your tax Prompt. So if we head over to the Midjourney Documentation under User Guide, let's go to advanced prompts, and let's go to multi prompts here. And if we scroll, scroll down over here, it tells you what the multipronged basics are and how you can actually use them. But if you go down here, there's a section called Prompt weights and this is what we're interested in. And simply the description is when a double colon is used to separate a Prompt into different parts, you can add a number immediately after the double colon to assign the relative importance to that part of that Prompt. So let's go ahead and put this into practice. For this exercise, I've gone ahead and generated this image that we saw earlier in our lessons. And what I would like to do is I want to use this as a base and play around with waiting. So what I want to do at first, I want to start by placing a heavy emphasis in our Prompt on the car tesla, and then less emphasis on the desert, which is the environment. And then we'll do that again, but reverse the numbers. So we'll place a lot of emphasis on the desert and less emphasis on the car Tesla, by using the Prompt Weight. So first thing first, let's go ahead and get the seat number for this. So here's our seat number. Let's go ahead and copy that. And now we can go into our prompts here. Imagine. And this is where we want to say Tesla. And immediately after tesla, this is where we actually want to assign the way. And this is how we want. We can accomplish this by using the double colon here. So I'm gonna give it a higher weight. So I'm going to say 1.5. And then I'm going to say driving in the desert. And for the desert immediately I'm gonna do a double colon and I'm gonna give it a very small weight, so 0.5. And then at the end, I'm gonna give it the seat number. Okay, Let's go ahead and run this to see what our results look like. Alright, we have our results. So if you take a look here, some of these actually look pretty good. So this one not so much it as a person and some weird stuff going on. So this one's out. This one I don't really like the coloring here doesn't really signify the desert. These two TO number two and number four on the quadrant look pretty good. So you can see that there's a heavy emphasis on the car and you can see a big chunk or the picture is taken up by the car which is dashed line here is actually driving. So we can see sort of like the sand and everything that's on the dust that's left behind by the car driving. So this is actually pretty good. Now, let's go back to Midjourney, and let's do this same thing, but now let's reverse the numbers. So let's just say imagine. And then we'll say Tesla. Now, immediately here we'll say, we'll give it a double Prompt, double colon for the Prompt, and we'll say 0.5 instead of 1.5, driving in the desert. And then to the Desert, we're going to actually bump it up to 1.5 instead of 0.5. And let's see how that's going to turn out. And then I'll give it the same Seed. Now let's render this and see what the results look like. Okay, let's inspect the results here for this last latest Prompt. And as you can see, this is kinda what we were expecting because you can see that Midjourney hasn't really put a lot of weight or significance on the car portion or Tesla portion, as you can see it try to fulfill the other side which was driving on the road. And it's got that pretty perfect. So but in picture number one, number two, and number four, the car is any men, Tesla, the focus is really the desert and driving a car driving on the desert, not necessarily a Tesla. Number three though is this is okay. This is a Tesla driving in the desert on the road. Why do you can see that still there's, the car itself is very smaller in proportion to the entire image. But because we define Tesla as a, in terms of less significance in our texts Prompt, the AI model focus more around driving on the desert, which you can see is illustrated perfectly on Image 12.4. So this is obviously, this is just one example, but this is how you can apply the Prompt Weight in your prompts to put specific and fest emphasis. Or putting more, heavier or less and less emphasis on the specific parts of your tax Prompt. 30. 30 Midjourney Prompt Consideration: In this lesson, I want to spend a little bit time covering some of the important aspects of what you should be thinking about and what you could potentially include in your prompts to achieve the best results. So if we head over to the Midjourney Documentation here under the Getting Started section, there's a section called next step. So go ahead and expand that, and then click on prompts. Now, over here there's some basic descriptions of what a prompt is and what did somebody different components of a Prompt and the structure. But we've seen a lot of this already. So I would like to bring your attention to this part here at the very end of this page. And it's, it's, there's a paragraph here that says, think about what details matter. So there's a lot of things that you could potentially include in your Prompt. Obviously, the less the simpler your prompt is, and the less you include your relying more of the model, figuring things out for you, and more creativity. But if you want to achieve the best results given your use case, you might want to consider including somebody's aspects in your prompts. So for example, you need to specify the subject. So it could be a person, animal, character, location, object, the medium. So it could be a photo, painting, illustration, sculpture, doodle. It could be anything that would distinguish the type of output you're going to create and also the type of output. It's the illustration, right? And what it's going to look like the environment is gonna be indoors, outdoors on the moon, on a different planet, right? The lighting, There's different aspects to lighting. So soft, ambient or gas, neon, right? The color, you can define things that black and white. You can define them as bright, you can define them as vibrate. Then there's the mood. So you can select the mood of the picture or the portrait you're trying to create, right? And then composition, um, and then you have the different types of photography compositions here. So we've got portrait headshot, which is gonna be very up-close, close up, bird's eye view. I've even some other things that are not mentioned here though, and I've seen in a lot of prompts are things like movies. You can put the name of the movies in your prompts. If you want this style to be very similar, you can name an actor, right? To make things more closer to that characteristic, you can use an actual actor or the person's name to create and modify certain things in that picture for that actor, for example, their clothes, their smile, and so on. And you can even put the sort of, the, one thing I've seen is the camera lens. So there's a lot of popularity camera lenses out there. And if you're looking for a specific shot using a CAM, specific camera lens, for example, a cannon one. Then you can even put in the lens and the modal number in your Prompt and AI will try to go and figure, figure it out based on all the data it has. So there's a lot to think about. So one thing is go through this list and always refer back to this as you're constructing your prompts. And now next, next thing we're gonna do is just do a few more pictures considering all these things so that we can tie it all back together. Now that we have an idea of what things we can play around with in terms of the details. Let's go ahead and just do a few more prompts and experiment with somebody's dimensions and attributes. So first thing I'm going to do is let's do, imagine that. Let's say extremely close up portrait of a woman with beautiful ice. So let's run this and see what it comes out with. Okay, We have the result and as you can see, they look pretty good. So here we got four very close-up portrait of then I. And you can see they all look very nice. And this is exactly the type of thing we're looking for. Something very close up, right? In our prompts. So extremely close up. And if you want it to now you can optimize one of these and use it for your case. Or you can generate more variations based on one or just completely regenerate for new options. Now, for our next one, let's do something a little bit different. So let's say, imagine portrait of a woman in New York City, fashion show. Let's say it's indoors. Ambient lighting, black and white color. And let's say calm mood. Okay, so let's go ahead and run this and see what we get. Careless. Take a look at the results here and the idea look extremely good. So you can see it actually was able to match our Prompt pretty closely. So we have a woman here in the portrait. There's blurring the background in all four pictures, which I really like. There's ambient lighting and then there's the person is actually staring Staring into the camera, in this case looking at us. And it's also black and white, which is the color we determined. So the results look pretty good and pretty close to our Prompt here. For our next problem, Let's try a different view here. So let's say imagine, I'll say bird's eye view of a horse running in a field. And then I'll say Disney. Pixar Animation should be outdoors, and then vibrant colors. So let's go ahead and run this. Right if you have the final render result here. So these look pretty good. I would say maybe this one, number two is the closest to bird's eye view, definitely not. Number four. I would say potentially number 12.3 are kinda closer to the bird's eye view, but most of all number two. But as you can see, it has done its best to try to make this into animation. Horse running in a field that's closely matches the any movies or shows that were created, or animation that were created by their Pixar Animation by these needs. So again, results look pretty good. Okay, Let's go ahead and do one more here. So imagine. Then let's say astronauts, lady and every wanna do neon colors. And here I'm going to include the Canon lens. So let's do a popular one. So Canon EF 50 millimeter lens. And then let's play around with somebody. Some conditions here. So I'm going to say raindrops on her space suit is really interesting. So let's run this Prompt and see what it's going to give us. Alright, let's take a look here. Again, results look pretty good. So I would say this one is really good. Number two, number three is also very good, matches the prompts here. And this is just some other ways of playing around with some of their parameters and arguments that you could include in your prompts. Now, feel free to make a list for yourself of water at some of the things that you could include in your prompts and be more specific to achieve the best results. If you ever need help. Or a point of reference, I would say your best bet is just referring to the Documentation here. This is a really, really good list here that you can always refer back to as a reference and use in your prompts. And it should help you achieve the best results 31. 31 Midjourney Settings: Alright, let's spend some time together and go through the Midjourney Settings because I think there's some important ones that you need to be aware of. So the way to access Midjourney Settings is very simple. All you have to do is type in slash Settings and you can see it already autocomplete stuff for you. So go ahead and press Enter. And now you can see all the settings here. Now, there's a lot here and we won't go through all of them, but we'll go through the most popular ones. So first of all, here you can see all the Midjourney versions available currently. Midjourney version 5.1 is available, and that's what I selected. If you want, you can select older versions if that's what you desire. But for me personally, I always want to have mindset to latest because the latest has more training data and also it has new features which could be beneficial to use. We also have, over here we have remixed mode, so re-mix mode is quite interesting. If we go, Let's go ahead to the Midjourney Documentation here. And let's go to User Guide. Let's open advanced prompts. And then here we have re-mix. So let's go through the description together. It says use remix mode to change Prompts, parameters, model versions or aspect ratios between variations, re-mix. We'll take the general composition of your starting Image and use it as part of the new job. So this is very important and this is how you can actually use re-mix. And you can see here, it's showing you an example. So it starts by, here's, here's the first starting points. So it says Line Art stack of pumpkins. And this is what it generates. And when do we have remixed mode? You got this button available. So you click that. And then over here for the remixed Prompt, you can just say pile of cartoon owls. And then this is the result here. But you can see that if you look at this pile of cartoon owls, It's very similar to this one here, which was our starting Image, which was a stack of Pumpkin, Pumpkins. This is how you can actually alter your prompts using re-mix to achieve the same structure with a different object in this example. And then here's some other examples that you can go through and play around with in remixed mode. And again, it's very easy. All you have to do is type in slides setting and then turn on remix. Okay, Now let's discuss this setting here called fast mode. So what fast mode is is that with my subscription, I get this option. And what this means is that every time I provide a prompt in Midjourney and I ask it to generate a picture for me or an image for me. It's always going to use GPU time, but it's going to actually make it really fast. So it's going to do as fast as possible to generate that image for me and use as much GPU power to do that. Now, one thing you can do is you can simply turn this off. And if you turn this off, it's going to turn on something called the relaxed mode. So let's head over to the Midjourney Documentation. Over here there's a section called Subscription, so click that. And then here you can, there's a section called fast and relaxed mode. So click on that. And over here there's a description. It says Midjourney uses powerful graphic processing units or GPUs to interpret and process each Prompt. When you purchase a subscription to Midjourney, you are purchasing time on these GPUs. So very important to understand. Essentially you're paying for hardware is while you're paying for in your plan. And it says different subscription plans have different amounts of monthly GPUs. Obviously, the more you pay, the more you're going to get. And over here, it says, how many GPU minutes do my generation's cost? And it goes into some details. But I want to bring your attention to this section called fast versus relaxed mode. So here it says, subscribers to the standard and Pro plan can create an unlimited number of images each month and relaxed mode, relaxed mode will not cause any of your GPU time. But jobs will be placed in acute based on how much you've used on the system. So this is very important because what relaxed mode allows you to do is it allows you to experiment with concepts without getting deducted from your time that comes with your plants. So it won't charge you anything. It does take longer to give you the image, but you can, in theory, create unlimited amount of images for free, right? So you want to use the actual fast mode or fast setting for when you have your final results and you want to sort of upscale that or make variations of that, right? So you want to, you want to, you've already has some, have something you're happy with. But if you just have some thoughts and you don't really know, or if you're experimenting with different parameters are Settings and you're not sure what you're looking for to begin with, and you just have thoughts that you want to experiment with. Relaxed mode is the better way to go because you can try to provide, you can, You can provide as many prompts as you can or as you want, and it won't charge you anything. And once you're happy with something specific, then you can and you're happy with your Prompt and the structure of your Prompt and the attributes of your prompts. Then you can switch the fast mode and then you can use it to generate the final results of scale it and use it for your use case. Also one other neat feature in or command in Midjourney is whenever you want to figure out what your usage is, depending on your plan. So whether you are in a free trial or one of the other three paid plans, it's really easy to figure out how much of that you've already used and how much you have remaining all you have to do in Discord typing slash info. And if you just click Enter, you can see that how much you've actually used and how much you actually have remaining. So here you can see that I have the basic land. It's on fast mode, it's public and fast time remaining. I got 19, 4 min out of the 200 remaining, I still have about I've used about roughly 3% of my plan for this month. And so far, I've created 92 images and I've consumed about 1.06 h. And for this particular plan, I have about 3.33 h in total. So yeah, this, this is just a neat little feature or trick that shows you what you have. And then also there's some information here about jobs that are currently queued or running. And she's just give you a quick snapshot in terms of what you're usages. 32. 32 Midjourney Recap Exercise: Alright, earlier in this lesson, we try to have a Prompt where the Prompts said acute hot dog wearing sunglasses and the bead somewhere in Hawaii. And as you can see, Midjourney wasn't able to interpret what I wanted to do. So over here we generated two different variations. So this is the first one, as you can see, and this is the second one. And as you can see in all the eight pictures, the the dog is the main the dog is the main focus. I actually, it's not what I was looking for. I actually wanted a hot dog character like acute hot dog character like a cartoon or something like that, wearing sunglasses sitting on the beach. But Midjourney just focused on the dark bar, the animal part. And of course in this one, for example, they capture both the dog and the hot dog. But it's not what I was looking for. So what does this mean? Does this mean we should give up? No, it just means that we should try different things. So we just need to think outside of the box a little bit. And we should be able to see if we can use some other techniques we learned in this section to go back and tackle that problem and see if we can get the results that we desire. Okay, let me explain my thought process in terms of how I think we could potentially tackle this problem, again, is just a matter of trial and error. We could really achieve anything. We just have to take the right approach. So what I'm thinking is that first we have a Prompt where we generate sort of like a character in the format of a sausage, like a cartoon character, animated character that is a sausage and wearing sunglasses, then we can generate a separate Prompt. So in our second Prompt, we're going to generate just the beach in terms of again, cartoon beach or animation beach or something that's animated. And we can use the combined Functionality in the image prompts to bring the two together. So we're going to take the sausage character and we're going to place it onto the beach and then see what the final result actually looks like. Okay, for this Prompt, I'm going to generate the sausage character first. So this is what my Prompts going to be, a close-up portrait of a sausage wearing sunglasses. Now I'm going to throw this in Pixar animation studios so that the model, the AI model knows what I'm looking for. And then I'm also going to throw this in there. And this is Sausage Party is a movie that came out in 2016. It's an animated movie which is sort of includes adult humor. And there's a lot of different types of sausages and different characters in the movie, which I think with fit this case perfectly. So let's go ahead and run this Prompt. I forgot to include the imagine, sorry. So let's do that again. Okay, So we have the results here, and these actually look pretty good. So first picture is out, this is again a picture of a dog wearing sunglasses. So it didn't really again understand, but number 23.4, look pretty good. These are sort of exactly what I'm looking for. I'm thinking probably number three is best what I would like to have. So that's what I'm going to upper upscale. So let's go ahead and I'm just going to grab number three and upscale number three. Now that we have the picture of our sausage or hot dog, now let's go ahead and create the picture of our beach. So this is what I'm going to put in. So imagine and then I'll put in this prompts. So I'm just simply saying beach, summer, sunny. And then I'm going to say Pixar animation studios again to give it that cartoonish or animation like fields. So let's go ahead and run this. Okay, Let's quickly inspect the result and these look pretty good. So I think number two is a little bit crowded here for us to place this oxygen. So number one and number four are ideal in terms of the beach being close. This one, the number three, it looks great. You see more of the water, but it's just a little bit further away. Either one, it doesn't really matter which one we choose, but I think the ideal candidates will be either number one or number four. So in this, in our case, let's go ahead and choose number one here. So I'm going to upscale number one. Okay, now we have both of our images stand alone. So we have the picture of our hotdog and we have the picture of our beach. And now simply we can use the technique of Image Prompts. That basic technique we saw on how to place an object in one image into another to try and incorporate the hot dog here into the beach. So let's go ahead and do that. Let's start with imagine. And why we can do is we can just simply drag and drop. So we get address of the image link for the sausage. So go ahead and press Space. And now what we wanna do is drag and drop the address of the beach here. So we got the, we got the first image, we got a space here, and then we got the second image. So let's go ahead and run this Prompt and see what the results look like. Now, if you look at the results here, these are pretty great. So this is exactly what I envisioned when we first started this exercise. So as you can see, this matches my Prompt a lot closer now than the other one where I just got the dog wearing sunglasses. So this is actually what I wanted. And the other two variations that regenerated, That's not what I was looking for, but we were able to use the techniques we'll learn to bring things together and achieve the results we want it. So the whole point of this exercise was to show you that you won't always get the result you expect given your input or Prompt. But that doesn't mean that you can't achieve the results that you desire. So you just have to learn the tool, the application, different techniques and features. And you need to be able to think outside of the box a little bit to bring some of these designs and Creativity together. 33. 33 Bing Chat Overview: In this section, we're going to learn about Microsoft Bing Chat. Now, Microsoft Bing Chat is very similar to ChatGPT, except it's built by Microsoft. Now, in order to get Setup, it's really quite easy. You really need two things. One is the Microsoft Edge browser, and you also need an email address to sign up for the Bing Chat. In order to get started with interacting with Microsoft Bing Chat, first, just open the Edge browser, navigate to www.bing.com. And there are several ways of actually accessing Bing Chat here. And there are different entry points. So one is up top here, it says Chat, so you can click that over here. It's as try it. And this is where you can actually, it will take you to the Bing Chat page if you click this. And also over here, there is a button, blue button with a B on it. And if you hover over it or if you click it, you can see that it will Open a side panel where you can start using the Bing Chat here it says ask me anything and you can put in anything you'd like. Again, it's very similar to ChatGPT. You can put in prompts in terms of Instructions are Questions, and it will give you answers. So that's another entry point. And there are three different ways of getting there with this page. Now, you need an account. So right now I'm already I already have an account and I'm already logged in. If you're not logged in, it will say sign in here. So go ahead and click Sign-in. If you haven't Outlook account, you're already good, you can just use that to sign in and start using Bing Chat. If not, then you'll just have to create one and then use that to login. And after that, we're ready to Use Bing Chat. I have my account, I'm logged in. So let's go ahead and click this Chat button here. And this should take us to the Bing Chat interface and you can see everything is new here. It says Welcome to the new bank. Here's some, let's quickly walk through the user interface. So it's actually quite very straightforward, just like ChatGPT, where you have an input where he can put in your Prompt or Questions and then it will give you answers. So here are some sort of Example prompts that you could potentially provide. So what are some of the meals I can make for my pinky taller, what are the pros and cons of the top three selling beds, vacuums, and so on. So you could ask it questions, you can get more creative. Inspiration is like this one here. And Bing Chat currently runs on GPT four. So and also on top of ChatGPT, one of the advantages to Bing Chat at this point in time is that it can also search the web for results. So for the most accurate and up-to-date result, ChatGPT currently cannot do that, at least not the free version, but Bing Chat. Not only it's running on the latest version of GPT, which is GPD for at this time, but it can also search the web. I'd give you more accurate results. So that's one of the good things about using Bing Chat. And over here you really have three different conversational styles. And you need to adjust this based on what you're trying to accomplish using Bing Chat. So we have creative, so this is when you want to test out concepts or you want to get something. If you want to get inspiration for creativity, or if you're trying to explore design concepts or ideas, for example, balance is sort of in-between creative and precise. And if you want something precise than you would want to choose the precise option. For example, if you're looking for a food recipe, you probably don't want to be creative. You want precise results. You want precise portions of a specific ingredient to put together a meal. So these are sort of what these things mean. And over here, this is basically the most important part, which is the Prompt. And you can see you can type up to 2000 characters in here. And you can also click the microphone if you wanted to sort of speak and it would turn that into tax. But this is where you can actually put in your questions and Prompts. Also, whenever you are done with a specific topic and you go your results from Bing Chat and you want to start something new, all you have to do is hover over this button here it says new topic and you can start a brand new Chat and on a different topic or a different thought that you're trying to explore. Now that we're familiar with the user interface of Bing Chat, let's walk through several examples. Now because Bing Chat and ChatGPT are very similar as in they're both Prompt based. We're now going to go through a lot of details because we did go through a very comprehensive coverage with ChatGPT. And you could pretty much accomplish the same thing with Bing Chat. We won't go through everything again with West Bank. Now, one thing is I do want us to run through some examples just so that you can see how you can use Bing Chat in your daily lives and how you can take advantage of this? 34. 34 Bing Chat Examples and Exercices: For our first example, let's start with something simple and finance. So let's go ahead and ask this. By the way, I'm going to leave this unbalanced for the duration of this course. So let's go ahead and ask it. First. I wanted to look up some information for me, so I wanted to actually do some research and then I'm going to ask it my question or give it the instruction. So I'm going to say look up the last earnings call. So these are the quarterly calls for companies that are publicly traded on the stock market and on the exchange. So look up the last earnings call for Microsoft. And then I'm going to say, Tell me about the revenue growth in simple terms. Okay? So it's realized what we're actually trying to do and now it's going to give us the answer. Okay, so now we have our results from Bing Chat. And immediately you can notice some differences with this compared to ChatGPT. So we do have our results here. But one thing is, as I hover, first of all, actually on the bottom here you can see that we, Bing Chat was able to search the web for the latest information. And it included all the sources that it got its information from. So there's five here. And the nice thing is if I click of any others is going to take me to that article or two, that news article or that website that it got its information from. Also, if you hover over some of these, you can see that there's a hyperlink. So in here it says the middle range of the middle of the range will add 55.35 billion implies 6.7% growth. Now, if I click this, this is going to open a new tab and it's going to take me to that article from CNBC that included this information. So this is very, very helpful. And if you wanted to dig in more and go more in depth, you could actually go through the original article and even research further if that's what you want to do. So this is one of the differences between ChatGPT and because Bing child can surf the Internet, the nice thing is it automatically includes the references for you. Now, another difference that you might want to consider is that Bing Chat also presents some follow up ideas or suggestions and recommendations. So here it says what, where the earnings per share. So in our original Prompt, we asked about revenue growth. Now it's asking us, do you want to know about the earnings or the profits in this case for that specific company, right? Then it says, how does this compare to other tech companies, or what are some of Microsoft biggest revenue streams? So this is a good one, for example, again, depending on what you're looking for, if I click this, it's just going to automatically populate that Prompt and send it to ChatGPT. Oops, excuse me, I meant to say Bing Chat. And now Bing Chat is going to again research more and then give me the results here. So you can see here Microsoft, largest product-based revenue streams or Office products and services, azure Cloud Services, Windows Server products. And it gives me it gives you the breakdown as well from their revenue perspective in dollar amount and in percentage amount. So this is very good. And again, it provides the sources where it got information from. Okay, we have our results here, and these are basically Bing Chat got the results from Time Magazine and it was also able to recognize a year here, even though I didn't say anything, I just referenced this year and it was able to change my Prompt to talk influential people 2023. So that was its search criteria when it's looking, browsing the web for results and information. And it gave me this list here, which is great. And I have 100 people. This is basically a list of hundred, top hundred influential people love 2023. And over here again, nice thing is it has the source, right? And then some follow-up suggestions. So for example, let's say you are doing an essay or topic, or you're writing a book on a specific thing or a person, right? You can actually use this method. And if you're doing an article or news article on a specific person, for example, and that person is in deltas list. This is sort of the follow-up Prompt can help with Bing Chat. So you might not necessarily get the best result if you have a very long Prompt asking, get tons of questions about specific things in a specific sequence. But if you have shorter prompts, but you follow up with other shorter Prompts able achieve the best results. So over here, we asked for this. Now we want to sort of narrow things down a little bit. And it already gives me a suggestion that says, can you tell me more about Jennifer and this other person? You could either just click on this or you can type in any of the people that you see in the top 100 lists, right? So you can really type the Prompt yourself, or you can simply just click on this. And it will actually give you more information on that specific person or people or whoever that you are actually looking for. And right now, it was able to recognize the two people that we asked. And now it's giving us more information and we can even ask it to further, expand even further. And over here, you have some emotions, right? So this is to help the AI model and the algorithm. So here if you say like, then it means that you like the answer. If you didn't like you click this. And then you can also copy this To Clipboard if you want to paste that in a document format or a text editor or whatever. And over here you can export and share. For our next exercise, let's go ahead and start a new topic here. And the point of this exercise is we wanna do something very similar where we can actually ask Bing Chat to give us the output in a specific format. So in this exercise, we're going to try to get it to give us the results and then put it in a table format with some specific criteria. So for the Prompt, I'm going to ask it. Let's say Look up the top three electric vehicles. And I'm going to say format, the output in table format. Or I can just reward this a little bit. The output should be in table format. The table should include four columns. The column headers should include. So I'm going to say the car model, the price, the pros and cons. So let's see what Bing Chat comes up with. Alright, as you can see, Bing Chat was able to do this pretty well. So it created a table here. And of course I can copy this table or downloaded or export it and use it in different places such as Microsoft Excel for example, if or put it in a presentation. But you can see that a creator the table. So it got the information, it recognized what we're asking for. It outputted the information inward into a table format. They table includes four columns, which is exactly what we did. And then it labeled it according to our instruction. And the three top cars, electric vehicles it found was Tesla Model S, Ford Mustang Mark II, and Audi e-tron. So, and it has the price for each. And it looks like here, the Ford Mustang is actually the cheapest option out of all the three. And then it goes into describing what the pros and cons of each car is. So this should hopefully demonstrate how you can specifically format your outputs with Bing Chat. For our next exercise, we're going to get Bing Chat to help us with Content Creation. So in this example, let's do a blog post on a mobile phone. So let's go ahead and tell Bing Chat to help us with creating a blog post here. So for Google Pixel seven, so let's say generate an idea for a blog post on Google. Pixel seven, detailing the new features and the tech specifications. Textbox. So let's go ahead and run this. Okay, so Bing Chat gave us some ideas here, three ideas. So it says a comprehensive review of the latest features and textbox, so we could include that in our blog posts. So we just have to do a better research. And here's some pointers here equal to include details on phones Design, display, camera, battery life and performance It says five exciting new features of the Google Pixel. So there's some things here like the new camera, the new processor, built-in VPN and other Features again, you could include these details in our blog posts. And then it says we could also do a comparison between this new phone and other smartphones and see how it stands out. So this is pretty good. Now, I would like to do a follow up Prompt. So we got the information that we need in order to help start writing our blog post. But now let's say I want to create some images for those. Now previously we've covered Midjourney, and this is where now you can leverage programs or applications such as ChatGPT and Bing Chat to help you create those images. So here, now that Bing Chat has context as to what I'm trying to do in this case, create a blog post with on the phone Google Pixel seven. Now it has that context built-in, so I can continue with my follow-up Prompt. Now what I can say is generate three mid journey Image prompts that I could include in this blog post. So let's see what it comes up with. Okay, So these are the three Prompts that it came up with four Midjourney. So the first one is not bad. A picture of a person using Google Pixel 17 as camera to take a photo of a beautiful sunset. So this is not bad. Number two and number three are very specific to the phones feature and I don't think Midjourney will be able to interpret that. So what I'm going to do is I'm going to generate another, I'm gonna give it another Prompt to Generative three more pictures, but I'm going to keep it a little bit more specific. So I'm going to say generate three Image prompts for mid journey, highlighting Google Pixel seven. Design. So let's go ahead and run this. So it gave us three more Rock prompts here for Midjourney. So the first one, a picture of Google's seven and modern design. So this is a good one at a picture of the Google Pixel seven camera, lens and Flash. Okay, this is not bad. And then a picture of Google Pixel seven is high refresh rate display. Not too sure about this one. We could try it in Midjourney and see what it comes up with. Now, what I'm going to do is I'll pause the video. I'll go in Midjourney. I'll put in these Prompts, generate the pictures, and then I'll share them back with you. Alright, I've gone ahead and put it all three prompts that we discussed earlier into Midjourney. And I have to say the results are quite impressive. So let's take a look at the first one. The first one, the Prompt was a picture of a person using the Google Pixel seven camera to take a photo of a beautiful sunset. So as you can see here, this actually looks pretty nice. So if you want it to in your blog, if you're specifically talking about the camera and the quality and the resolution, this is definitely one of the pictures that you could include in there. Now, for our second Prompt, it was a picture of Google Pixels seven, slick and modern design. So this is a good one. If we enlarge this, you can see that these are all pretty good here, depending on which one you like. This one here, number four, it looks pretty good. And one and number three also looks pretty good. Again, based on your preference, you can use any of these in your article Post or blog post or whatever it is you're doing. And again, if you're not happy with this, we covered different techniques in Midjourney section where you can create new variations and play around with other Settings. And for our last one, the last Prompt was a picture of the Google Pixels, camera, lens and flash. So this one's actually really, really nice. So over here, I would say number two and number four, specialty number four, they look very beautiful in terms of the image and the concept. And it is really truly highlighting the flash and the lens of the camera. So as you can see, Bing Chat is very powerful and it can help you with creating Image Prompts to be used in applications such as dolly or Midjourney. Bing Chat is a very powerful tool and just like ChatGPT, you can use it to accomplish many things. We could use it for Content Creation, we can use it for inspiration. You can use it for creativity and design ideas. You can use it at work to increase your productivity. You can use it to get ideas for graphic design, our even prompts for generating images using other AI applications. It's basically endless possibilities and you can use your imagination for really anything you'd like to accomplish. It can even, you can even use it to solve math problems or math formulas. Now, just like ChatGPT, one thing I would like to ask is please fact check the results you get from Bing Chat because sometimes they're results it gives you in the output format is not quite accurate. So you don't want to necessarily just copy paste everything you get from Bing Chat and use that directly on your project or whatever it is you're doing. Please do your own due diligence, make sure the information is accurate. So you may have to do a little bit of more research on your part. And you just have to use this as a starting point. So this is just like very similar to ChatGPT. You want to be very careful because sometimes, again, the output is not exactly correct or precise or even true. So you want to avoid things that are being made up. You don't want to use facts that are made up. And this is why you want to make sure you double-check all the facts and you do your own due diligence 35. 35 Microsoft Designer Introduction: In this section of the course, we're going to learn about a tool called Microsoft Designer. Microsoft Designer is another AI tool to help us with our creativity and design when it comes to graphic design. So first things first, you just have to launch the browser. And I have the Edge browser open here. And we just want to navigate to Designer microsoft.com. And again, you just need an account that's really all you need is currently free to use. And if you don't have one, you can create one. I already have an account and I'm logged in. So as you can see here, this is the user interface. But before we start getting in, I just wanted to quickly explain the tool. And Microsoft Designer is simply a graphic design AI tool, just like DALLE-2 and Midjourney to help you with your ideas and creativity and help you produce creative designs. It's also a fantastic tool because it's a complete suite. So it's not just the Prompt based graphic design tool. It, you can use it first to generate your images and graphics, and then you can use the built-in editor to further edit and modify specific components in your design to help you finalize that design or the output from all the way from the starting point to the end. So it's very, it becomes very similar to Canva or fake bar if you have used those applications in the past. When you first navigate to the Microsoft Designer page, this is the user interface that you're presented with. So over here we have the Prompt inputs. So this is where you could actually input the Prompt In terms of what you're trying to generate for graphics. It says describe the design you'd like to create. And over here there's a placeholder texts that's giving you an example. So an Instagram post about my cosmetic product launch on July 1st. On the bottom here you've got a couple of buttons. One says Add Image. So once you create your graphics with Microsoft Designer, you can use the Add Image Functionality yourself to actually incorporate that as part of your design. So it couldn't, it doesn't have to be all AI generated. You could bring in other images that you created yourself or from other sources and include those as part of the final design. Generate mH. So this will generate the image based on the Prompt. And on the right-hand side, again, we got Examples. And as always, you can use these as inspiration. So just some other ones that have been used by previous users and posts that he has Example. So over here you can see it says a picnic club monthly social meetup event, April 2023, Central Park. So it's sort of like they've designed a poster for this specific event for advertising. Over here we've got some food. If we scroll down here, because one for ramen, so it says delicious RAM and recipe. And check out my cooking blog. And it says a Facebook post showcasing a Raman recipe. So it's specialized for Facebook posts just like this one here and Instagram post. So it's important to include what type of social media posts you're trying to make and which platform. And then it will try to, the AI model will try to adapt and give you that result that will best suit your needs. If you scroll down here, there's one for graduation, and obviously you can customize it to the name you like for the purpose you want Overview have over here we got the dancing glass here we have some of animals and so on. So these are very good examples. And just like any other applications we've seen so far, when you hover over the example, it will tell you exactly what the prompt that Text problem was that the user used to generate this output. Or you can simply just click on it and it will populate this area and it will generate the image. Obviously, the results may be different than what's displayed here. 36. 36 Microsoft Designer Examples and Exercices: To get started, Let's start with a very simple prompts. So over here, I've put something together, very simple. So let's go ahead and try this out. It says a picture of a person holding a healthy snack with a caption that includes the recipe and neutral nutritional information. So put that in and then let's go ahead and generate this. Okay, so here's the result of what Microsoft Designer generated for us, and they all look pretty good. The first one is, it says a healthy snack recipe, and then there's a picture of a granola bar here. This one, someone eating healthy salad. And on the bottom digit texts that says healthy snack recipe label here or here we have a woman again eating something healthy without some healthy snack recipes and so on. So results look pretty good. Now, if you don't like the, any of this while you could do is you can simply press generated again. And the model will try to generate some more new variations for you. So let's just say for the sake of this example and the exercise, we like this one here. So go ahead and click on that. And then once you do that, you get two options on the bottom, as you can see, you can just download this and it will give you the picture. Or you can click on Customize design. And when you click customize, customize design, this is where it will take you to the edit mode and you can use the editor. So let's go ahead and click on Customize design. Okay, now we have been directed to the editor mode in Microsoft Designer. And as you can see here, you start to have, you got your main image in the center and you have access to some tools here. On the left you got this menu bar templates, My Media visuals, Text, and brand kit. On the top here you've got the preview of your Prompt. Here you got your Zoom, undo, redo some Basic Functionality. If you just wanted to totally scrap this and start from scratch, it would just click on new design. Or if you're happy with this, exactly as it is, you can just click Download and then follow the instructions to get the download file and the format that you're looking for. Now on the right-hand side, one thing I'd like to bring your attention to is the, some ideas here. So for example, let's say you are not happy with any of this. These are some other things, some of the other templates that it can present to you that you can potentially choose from to replace this one if you're not happy with the structure of how things are placed on the graphic. On top of that, you can make edits of your own to this one. For example, if you don't like this text here because parts of the parts of this is actually going on the granola bar. And the word here can really be seen that well, you can modify that. How are you like? So if you click on it, as soon as you click, you'd see the edit options here. So you can change the font. You can change the size, you can make it bold, italic or underlying. You can do other styles such as strikethrough. You can align it loves the line, center line, and what have you. And then you can change the position and the opacity. And on top of that, you can choose to rotate it or resize it. So in this case, let's say I wanted to resize that. I can even grab it and move it a little bit to the bottom here. And you can see the automatic aligned tool letting me know when it's perfectly aligned with the top one. So over here, I can see that it's letting me reread, basically reformat this so that the whole thing can be seen. And if I wanted everything in one line, what I could simply do is again, click on this and change the size to something smaller, right? And then everything will fit into one line here. So just a very simple example while you could accomplish here. Now when you click on the whole picture here, you can see you will get, you're presented with this menu on the top. And this is where this is the background. So this is where you can choose the color of your backyard background. You can replace it with a different image if you'd like. You can crop it. You can apply effects to it. So for example, if you click on Effects, Let's go ahead and do that. Here's some filters, your typical normal warm punch. So let's do warm here. And as soon as I apply that, you can see that over here, the picture changes to apply that filter. Obviously, I can adjust the intensity here, so let's do less it, the less filter, the more it presents the original picture and the more it will move away from the original picture. So this is sort of add to 50% line, right? Obviously you can go and experiment with this. You can change the brightness, contrast, saturation and all of these things. And you can even apply blur. Here. We can do the blur. And you can also, if there's your picture that you're trying to bring in has a background that you don't want to have the background because it's not the main portion. You can, there is removed background Functionality that you can leverage over here. So now let's go back here. Let's undo all the changes we just made. Now, let's experiment with somebody's idea. So let's say this was the original, suggests that image that you're starting off as, as a base. And you're not really happy about how things are organized. Of course, you can choose to move this around because you have access to all of these components on the page. And you can move this around however you like. So for example, right now I can move this text, I can move this text, I can resize them, I can change the content I can add more texts. And Eve over here, I can, There's a, if I click on this background and I click detach background, I can actually just completely remove this background and replace it with a different Logo here, right? So that's what this Replace option is four. So let's go ahead and undo that. And if you want it to, you, what you can do is you can explore somebody other ideas here with different structures. So for example, let's say we want to try this one here. So go ahead and click that, and now you get something that's completely different. I actually liked this one. This one is pretty good and it's perfect for sort of like a poster or a cover or as social media ad. If you're trying to advertise something or promote something, healthy snack recipes, the colors are very nice. They compliment each other. They're not too harsh on the eyes. So this one is actually pretty good. I liked this one better than our original one. Obviously there's other styles here. So for example, this is another one. This seems very close to the posts that you see on Pinterest platform. If you've ever Use the Platform Pinterest, this looks very similar to that. But overall, I think this is a really good one here. There are a lot of different edits you can make to this graphic, including adding your own text if you want it. So for example, here you have the design, the final design looks pretty good. But in addition, you want to include your own website or the website or the URL to your blog posts so that as part of this ad or promotion, people can actually see it whatever you post this on social media, for example. So it's very easy to accomplish. You can see again this menu, all of these are very self-explanatory, but over here you've got the Text tool. So you can just click on it. And you can put something in here and let the AI generate the texts for you. Or you can simply do this over here. So you're presented with some pre templated Text styles that you can choose. So for example, this one's nice, This one's cool. Grand opening, this one's good. And then you can just explore somebody's right. But whatever it is that you like, you can choose that and then you can use it. We'll add it to your final image and then you can customize it. So let's say I like this whole for example. So let's go ahead and click that. And as soon as you click that, it will add this placeholder to your final image. So let's go ahead and actually customize that. So first thing, I want to make this smaller. So go ahead and select the text here. And from 60, we want to reduce this down to something. 14 is good. So let's do 14 here and then let's bring it down. And now I just want to replace the text with my website. So let's say in this case www.example.com, Let's say this is my website. And now I have it in the picture. I can sort of make this size a little bit smaller or bigger. And what I can do as I can even change the position of this and potentially bring it bottom or to wherever you feel like is the right place in your graphic to include it. So this is how you can add text. This is how you can edit text. And again, adds a lot of different content or information to your ad or to your graphic. Once you're in edit mode, it doesn't mean that you cannot use the AI anymore to generate it. In fact, you can definitely enhance your graphic and Image with further modifications that are AI generated. For example, let's, let's, for the purpose of this exercise, Let's say that this granola bar, I am not happy with this. So instead I want to have like a protein bar instead, and I want to add that to my poster here for my purposes. So all you have to do on the left menu bar here, click on visual. Now, when you click on visual, it opens a panel, it opens a side panel and by default, it gives you some recommendations that you can use. So because it knows you're looking for healthy snacks. So it tries to find those things for you. And by default, it presents a list to you, so somebody's actually look pretty good. So if I liked this bowl of trail mix, for example, or fruit, I can just click on it and add it to my poster. If you scroll down, there is other things. For example, there's filled shapes, so you could add those two to your image to cover something or to create a background for a text or whatever it is you'd like to do. Here you have photos that are spraying photo and so on. You can just use any others. And of course, you can search for other things to see what are the other pre, pre available images are there for you to use. But one thing I'd like to sort of for you to focus on is this Generate button. This is, is to generate new images based on Prompt, because these images already existing and he can choose that for your purposes. But when you click Generate, this is where you can actually put in a Prompt To generate something that's more closely relevant to your posterior. Once you click Generate, now let's go ahead and produce the image we were looking for. So in this case, we want it to generate a picture or graphic of a protein bar and place it on our poster. So let's go ahead and do that. And then click degenerate here. Okay, It took some time, but Microsoft Designer was able to generate these images and these actually look pretty good. So if we were pretty, we're now presented by three graphics for the protein bar. And you can choose whichever you like. Let's say, I like this one better for example. So let's go ahead and click that. Obviously this is too big for our image, so I want to reduce this by a bit. And one cool thing here, one cool feature is that because if I place this here, it doesn't look very, it looks very abnormal because the background of the protein bar doesn't really match the background of the granola bar. So while we can do is there's a functional feature or functionality available here called Remove Background. So once you select the image of the protein bar, you can go ahead and select Remove background and it'll try to do its best to remove the background. So now as you can see, I can just bring that down and it's sort of nicely blends into the background that I had previously, which is more of a pinkish color here. And it fits the background a little bit better compared to before, and it nicely blend so it creates that seamless illusion that they're all part of this same picture. So this is how you can remove that. You can remove the background via that Functionality. Now if you notice, when we added this protein bar to our design on the right-hand side, all of the ideas are now updated to include that. So if you take a look here, these will have all been updated to include that as part of the new design, which is pretty cool because it just happens automatic. There's really nothing for you to do. It's effortless. So you can explore these ideas further. And for example, let's say this one, you can, this one is actually not bad. It rearranges things on the final graphic or the poster. Now, over here you can see that you see some residue here, over here and here, and maybe here. And this is just as show that the Remove Background feature did its best, but it's still left some parts, Like left in the it wasn't able to do it perfectly. And it kept some other residues and it's left on the picture. But that's not a problem because you can download that picture on your computer. You can modify yourself in a Image Editor and then upload it back in. So it's not a big problem at all. I think it try it did a pretty good job. You did a pretty decent job here, maybe 98%, 99% perfect in terms of removing the background. Anyways, going back to the ideas here, you can explore these and see which ones you like. This one's pretty good. And then there's all these ones that you can potentially explore and just choose the one that you think best illustrates the point. Now, this was the generate TAP was where you could actually generate, have AI generate Image for you based on that Prompt. But there's other things available here. So there are other tabs here, for example, photos. So if you put in photos, it will. Again, these are existing photos in the platform that you are free to use. You can search for whatever it is that you'd like to do. So for example, protein bar, Let's see what it comes up with. And these are the ones that are available for you to use years at one other person eating a protein bar here is one here, here's a hand holding it. So this is where, this is exactly what we have here in our poster. So the granola bar, so there's lots of available options here. Then there's graphics. So these are more sort of 2D or sketches that we can choose from. Again, you can search for what you're looking for and there's some videos available for you to use and then you just have to search for a specific Thank. Now, this is everything you have access to in the platform. But also what you can do is you can upload images from your computer or your device. So you're not limited to in terms of what's available to you in just the platform, you can literally import any picture. And the way to do that is very easy. All you have to do is click My Media. And here you can choose the device so you could do it from your laptop or your computer, your phone, google Drive, Dropbox, Google Photos, whatever you like. And in this case, let's just use an example here. So I'm just going to select this picture we created earlier in the course in Delhi. And it adds it here to the collection what we previously had. And once the upload is done, you can simply just click it and it will add it to our image. Obviously, in this case, this is, it doesn't make any logical sense for us to add this picture to this poster. I just wanted to demonstrate the functionality that you can upload any picture that you'd like from your device. Now, let's say you're pretty happy with the current picture, and this is your final design. And now you want to export it out of the Microsoft Designer platform and use it in other places such as social media. So very easy to do again, all you have to do is click this Download button. And once you click this, you have to select the type. So you can do PNG, which is an image format. You can do JPEG, which is the Image format. Or you could do PDF. You can select any of this. You can. There's some other options here. So for example, and make the background transparent if you're taking this image and putting it or overlaying it on top of another image somewhere else, that would be a useful option. You can select, remove watermark, so there's really nothing on there, right? So I would keep that option. And then you can simply just click download or you can copy as Image and then pasted it goes into the clipboard and you can paste it somewhere else. But if you want me to have the physical file, you can just click it, click Download, and then that will download it to your whatever your devices. And over here it finished downloading. And we're good. And you can also send that to your phone if you'd like here. So if you click Download, there's another option here called sent to fall. So this is how you can simply export your image out of the platform and then Use a, use it in other places. 37. 37 Adobe Firefly Introduction: In this section we're going to learn about an experimental tool called Adobe Firefly. So this is called Adobe Firefly, which is Embed Ahmad at the moment. And the way you can access it is simply open your browser and navigate to Firefly that adobe.com. Now because this is new and experimental, you currently have to add yourself to the waitlist and wait for them to add you to the platform so that you can start using. I had to wait for a few weeks and now I have access. And I have to say I'm pretty impressed with what this tool can accomplish. And in addition, this is a standalone tool like Midjourney and DALLE. But Adobe is working in incorporating these into other applications. So right now, this is actually, it says here that Firefly is coming to your favorite apps, but it's currently, it's integrated and incorporated in Photoshop. So if you have Photoshop, you will see a version of this and you can do some pretty cool stuff with it, like Generative filled where you erase a part of an image and then you can run a prompt to replace it with something different. Once you've created your account and you have, you gain access to the platform. You simply just after navigate to Firefly that adobe.com sign-in with your account and you're presented with this interface. And again, we're just going to do a quick tour here. And there are some guides to help you get started here. This one here, for example, text to image. This is very similar to what we've learned so far with DALLE-2, Midjourney and Microsoft Designer, simply, you provide a tax Prompt and try to do its best to give you an image output. There's another feature here called Generative filled. So this is pretty cool. This is says that you can use a brush to remove objects or paint in the new ones from text descriptions. So if you recall back to earlier lessons, this is exactly what we could accomplish in the DALLE-2 edit mode. Over here, you can apply styles or textures to Text with texts prompts. So you can do some really cool things there. Generative recolor. So you can actually change color variations. Here. You can turn it 3D objects or models into images. Again, very cool. Here you can extend Images and change the aspect ratio of your images. And here is, again, just some sample prompts that you can use. If you hover over this, it will tell you exactly the picture. It will tell you exactly what the prompt is. And you can just simply click this button and try Prompt and it will input it and run it and you'll see the results. Here's another one, Here's another one. So you can go through all of these and explore different prompts here and pictures. Again, this is really helpful. And you should definitely go through these and use it as inspiration. And also, you'll learn about the things you can actually include in your Prompt and the structure of your prompts to get the best results. And if we scroll a little bit further down here, these are some other things that the Adobe team is currently working on in terms of introducing new features into the Adobe Firefly application. Very, very cool stuff here. So if you scroll down here, there's this one. Personalized results are generated images based on your own objects or styles. Text two vectors. So this is pretty cool. Tx2 pattern. Here we got Text to brush, sketch to me. So this is pretty cool. Use case something on a piece of paper and it can turn it into an image. And then we've got Text, two templates. So lots of new features the team is currently working on to implement. So can't wait to see somebody's inaction 38. 38 Adobe Firefly Examples: Okay, so let's go through some hands-on exercises here and provide some prompts to see what this tool can generate for us. So let's start simple. We'll start with the text to image. So go ahead and click Generate over here and there's a description and you just have to, actually, I'm just going to skip all of this. And yeah, so here we are taking to this page and you have your big input here, which is where you would put in your Prompt. But the one thing I wanted to demonstrate, again, there's tons of endless examples here in terms of what you could potentially generate. And again, if there's something that you like that matches your style or what you're looking for. All you have to do is hover over it and you'll be able to see that Prompt. So this all here looks pretty cool. So yeah, I urge you to go through this and explore some of these. And they should serve as a good place to start and get inspiration from. For our next three Prompts, we're going to try something different just so that we keep things interesting and move, we move away from some of the previous examples I'm prompts that we covered earlier in the, in the course. And we'll try to enter the world of fantasy. We'll try to create some science fictional type stuff. And we'll cover the realm of space and see what Adobe Firefly can generate in terms of output for us. So there should be super interesting. So for our first Prompt, we're gonna do something simple. So we'll say a planet with three sons in the sky. So let's go ahead and run this Prompt and see what it comes up with. Okay, we're now looking at the results that Adobe Firefly generated for us. And I have to say, I'm pretty impressed here. So again, just like other tools, Image Generative AI tools that we saw, it gives us four variations here that we can start off with. And these all look pretty good and they match the, match the text Prompt. I gave it very closely. And I have to say, coming into Adobe Firefly after having experimented with DALLE-2 and Midjourney and Microsoft Designer, the user interface for Adobe Firefly is very intuitive, which is quite impressive. It goes to speak to their efforts into making the user experience very, very easy and very helpful in order to, for us to get started. And It's, it's quite impressive in terms of what it can generate. And if you hover over one of these images, it's pretty self-explanatory. So for example, it says rate this result. So I'm quite happy with this image. So I'm gonna give it a thumbs up. And obviously that's going to help the AI model in the future, right? So that's one, if I don't like it, I'm gonna give it a thumbs down. So for example, this third picture is not as great as the first picture or even the fourth picture. So I'm not happy with this one. I'm going to give it thumbs down because that's not what I'm looking for. Again, this is just giving feedback and they're going to use the feedback to train the AI model better, which is something I encourage everyone to do. Now, looking at all of these four things are all great. I like, I really like number one and number four here, if you hover over this, so we covered these features here. Now, if you go here, it says show similar. So this will give you or show some variations. Here we have degenerated filled. So this is where you can erase part of the picture and then sort of fill it with something different. Here we have the options menu. So if you click this, it says submit to Firefly gallery. And that's kinda like this showcase we saw earlier with Midjourney. Here you're choosing to open to share your Prompt and the results of your Prompt with the rest of the users and Adobe Firefly to help them get inspired by your design. And then some other options here like Copy to Clipboard and clipboard. And if you're happy here, you can just go ahead and download the picture and use it straightaway. Okay. If there are pictures that one of the pictures that you like, well, you can do, which is what I did in this lesson was, let's say you like this picture better and not the other three or just this one the best. If you click this one here, shows similar. It's going to change the other three pictures. And it's going to use this as a base and generate variations based on the other ones. So that's exactly what I did here. I click this, I'm going to click it again. And then it's going to change these three to something similar, just a different variations of this image. So what, that's what this shows similar does, again, it's very close to the generate variations in other Image Generation tools that we've explored up to this point so far. So that's what that's going to do. Now, let's Explore the Generative fill feature. So go ahead and click that. Once you clicked on Generative fill, it's going to take you to this page and it's a very simple, as you can see, very simple and intuitive editor. And this is where you can actually manipulate parts of the image. If you're happy with certain parts, you're not happy with certain parts. You can actually use the generator field to remove or add certain components to the image to make it match your desired results. So in this case, let's say I'm pretty happy with here we have the planet, we got the sun's here. We got some mountains below the atmosphere. So let's say the only thing, one thing I really would like to do is add a space station in the planet or in the orbit. So we can easily achieved that. Now over here, I got the add tool, select that, so go ahead and select that. And let's say this area. So this is what we wanna do. So go ahead and erase at this part here. Okay, so I erase this part. And as soon as I erase this part, note that on the bottom, we, the user now can see this Prompt input. And you can see there's a note here that says if you leave, if you leave the if you leave this Prompt blank, it will generate, it will fill it itself based on the surroundings. But in this case, I know I want the space station. So let's go ahead and type that in and see what Adobe Firefly can come up with. So let's say Space Station. Because I erase this part, it knows where to place it. So let's go ahead and click Generate and see what it can do. Alright, so Adobe Firefly was able to come up with some variations for me. So let's quickly walk through them. So this one, it actually plays something here. So it's actually cool because it's not exactly what I'm looking forward. This is more of a rocket but, or a spacecraft rather than Space Station. But you can see it placed it exactly where I asked it to. And also the surrounding look pretty good because it blends into the picture. There's no weird board, weird borders, or any inconsistencies that would stand out. This one here, this one is another option, so this is another rocket, here's another one, and here's another one. So lots of different options. Now if you want, you can click, if you're happy with these, you can click on keep. Or if you're not happy with this, you can just select Cancel. And I'm going to select Cancel and I'm gonna do something maybe a little bit different. So I'm going to say inter National Space Station ISS. And then let's see what it comes up with. Let's see if we can do something better. Okay, So we have four more results here. So again, these are more closer to rockets and spacecraft is not so much the station. But I think that's okay. I think the tool is still fairly new Anna, experimental and they're making it better and it's still learning. Here you can even click the More button and issued. It just more means generate four more variations for me. So that's what it's currently doing. So this one is not too bad. This one is more of a satellite, so you can experiment with this and see what you like. Let's say, I'm happy with this satellite here. I'm just going to keep it as is. And if you're happy with the final results, you can simply, you can even use the pen tool here. And you can move, move the image around if you like, and so on. But if you're happy, happy with the results, you can just click on Keep. And now it's going to keep that satellite in there in that location that you want it to. I wanted to insert the fill in. And you can go ahead and click on Download on the top right-hand corner to download the final image. Alright, for our next Prompt, we're going to try something different. So let's say Space Station or bidding black hole. So let's go ahead and generate this. We have the render results now, and they all look pretty good. So this one is kinda nice. This is great. This one is cool. This one is really nice. I like this one, and this one's also not bad. Yeah, let's say, I like this one the best year, so I'm gonna give it a thumbs up. And if you want it to, we can even share that with submitted their Firefly Gallery, which is I think it's good to help other people out as well. If you want it to make more variations of this, we will click this shows similar and it would change the other three images to something that style. But let's say we're happy with this. If you wanted to, we can go ahead. That's just download this image and start using it. But in this exercise, I kinda wanted us to explore some of the options here. So again, these are very, Adobe has done a great job in terms of making these self-explanatory and intuitive. So here it's very easy. It says Aspect Ratio. So you saw how we were able to change Aspect Ratio in Midjourney using the parameters. It's a very similar idea, except here everything is already ready for you to click on. So all you have to do is choose a preset one here, you got landscape for 233 to four squared, which is one-on-one, 16 to nine, which is something we saw in Midjourney and so on. So you can change these aspects ratios to whatever you like. So as soon as you click it, it will automatically change that. And now it's no longer a square, but it's going to be a rectangle. So it's doing all the redoing their rendering for you. And it's also creating some new variations for you with the new Aspect Ratio. Just note that when you select the aspect ratio is going to redo the design as well here, as you saw with this exercise. Also note at the bottom here, there is this filter applied called ART and that's referring to the content-type. So you could do none. You could do photo, you could do graphic, or you could do Art by default, is currently set to Art. And this is the type of images it creates based on that setting. If you want, you can simply change that to photo. So let's go ahead and click that. And that's going to generate some variations with the content type of as we go, I'm actually noticing that these variations are getting better and better. I really liked this one like this is really nice and very close to kinda what I envisioned when I first put in the prompts. So I'm gonna go ahead and give that at thumbs up there. It's very nice. This one's kinda cool. These are, these are all very nice and perfect for exploring the tool and design concepts here. And over here we have this style is we got popular, we got movements, we got themes without techniques, we got effects. So you can see popular right now is selected. And you got these ones. You got Digital Art, you got synth wave, you got palette knife, you got chaotic, you got neon. And just one thing to note is that you could also input these types into Midjourney. The AI model will be able to interpret that. So you will see a lot of people up if you're looking at examples, you'll see people mentioning Digital Art in DALLE, for example, or synth wave or even in Midjourney. And that's kinda like where somebody's are coming from. So you're not restricted to include these in just as part of your Prompt In Adobe Firefly, you're going to include these in other Generative Image AI models as well, or tools or applications. So let's go ahead and select one of these just to see what it looks like. So let's do synth wave. So let's go ahead and as soon as you select that, you have to rerun your Prompt again. So it says degenerate button becomes available. So go ahead and generate that and now issued, create new variations based on this style synth wave. As you can take a look from the results, this style is a lot different than what we initially had on the screen. And this is actually pretty cool. And that's what St. may produce as if the style is selected. Let's do one more or less, do neon here, I'm going to unselect same wave, I'm going to do neon. Let's generate this again, just to play around with one more setting here under the popular section to see what the results look like. Looks, results look pretty good. And yeah, this is why, this is probably one of my favorite ones. I really liked this one. This one is also great. And yeah, actually they're all pretty good. So that's kinda like demonstrating the whole style of neon when it comes to Generation. Now, there's some other settings here, color and tone. So you can go ahead and select different things here. For example, you can do vibrant colors. If you select vibrant colors, you're going to see different things. So let's do a combination here. Because it just takes forever to go through all different every single one of these one-by-one for lighting. Let's do something like dramatic lighting that should be interesting and composition. It says. So these are some of their photography competitions like close up, wide angle, bird's eye view, macro photography, shot from above. Let's do, let's do close-up. And then let's, as you can see, as we adding or filters, they are all getting added here. And if you want it to remove one of these, all you have to do is click the X button beside that filter and it will remove it from your prompts. So let's go ahead and run generate. Yeah, we can definitely tell the difference now. These two are definitely cool close-ups in terms of the view and the results look great. And if you're not happy with this and you would just want it to generate four more variations over and over again. You just keep half to click the Refresh button. And it'll be able to do that for you. For our last problem, let's do a little bit more creative and fantasy likes are we going to say a dragon flying through and asteroid field? And then let's see what it comes up with. Looking at the results. They look okay. I really don't like any of them. So what I'm gonna do is I'm just going to click Refresh to do for it to generate four more for me. These ones look a little bit better. I particularly like this one. This is actually pretty cool and very close to what I was looking for when I first put in the Prompt. And again, you can explore what these options and apply different type of filters to these to get the results you're looking for. So as you can see, you can literally try anything that you can imagine via these tax prompts here and all the filters available or other features such as the Generative fill 39. 39 Adobe Firefly Text Effects: We learned about the texts to Image Prompt on the Adobe Firefly application. This one Generative filled. So this is gonna be very similar to what we explored here. There's some really cool features here. So for example, Generative re-color. Some of these are still in development and design, so they're not even available at this time, for example, 3D to, 3D to Image or extended image. But I think this one is particularly nice as well just because some of the other Applications can't really handle texts really well as we covered in the earlier lessons. But I think this is one worth going over and getting familiar with in terms of Text Effects. And this is where it can apply styles to textures, two different apply styles or textures, two texts which the tax Prompt. So let's go ahead and run through some examples. Once you're on the Text Effects page, all you really have to do is describe the in the Prompt here, it says inter-text. All you have to do is describe the tax effect you want to generate so you don't include the letters or anything like that here, you just talk about the effect or the texture of how you want the output to look like. So let's go through some examples together. First one, let's say on the beach there's a lot of seashells. So that should be an interesting one. Let's do that. So that's just typing seashells and then see what it comes up with. Once you're on the Adobe Firefly Text Effects page, all you really have to do is it's a two-part prompts on the left-hand side, you just need to tell it what text you want or what the text is going to be in the final product. And then the second part of the prompt is just going to be the, you describing the effects that you want to generate on that text. So what kind of texture or effect you want to apply on that? So here let's keep things simple. So I'm just going to say Firefly. And for this one, Let's go through several examples. So on the beach there are a lot of seashells, so this should be an interesting one, let's say seashells. And now let's see what it comes up with. And we have our results here. So this looks quite interesting. And if you're happy with it, you can give it a thumbs up. But I think this is pretty good. You can also try different variations here. So if you're not happy with this one, click this one. And then L, we render based on this tile here. And you can see it's a little bit different and you can experiment with the other two as well. On the right-hand side, you got some sampled prompts. You can try flowers. So your texts with flowers, we did seashells, but you could do flowers, you could do steak, you could do driftwood, wires, balloon, a lot of things. Anything you can, any object you can possibly imagine, you can click on View All to see more samples. Over here we got the text effect fit so you got tight, medium and loose. You can choose your font. Again, there's only a couple of mentioned here, but you can click on View all and change the font. So here I can select poplar, and it's going to re-render that to match that font. You can see that it can, it will change. And again on the bottom you can select the color if you want. Right now the background and sort of the color is transparent because they're assuming that you're going to take this and put it over layered on top of another image so it not having background makes that process a lot easier. And again, this is pretty straightforward and self-explanatory, very cool feature. And in order to experiment with different textures and style of Text Effects if you want it for your posters, for example, or your fliers. And very intuitive. And again, if you wanted to, if you're happy with this result, Let's say this is your final effect and you're happy with all the filters you've applied. All you can really have to do is just click on the Download button and you can download the final image. And again, there's some options here that you can submit to Firefly gallery if you're happy or just copy to the clipboard and paste it somewhere else? 40. 40 Best Practices: Let's go through some best practices. In particular, how not to use Generative AI. Now, one of them most important thing and one of the things that I would like you to take away from this course, and probably the most important is please do not copy paste everything you see from the output of applications such as ChatGPT or Bing Chat. Because you, there's been cases where I've seen ChatGPT just makes up facts on its own which are not true. So it's very important that wherever you get an output, You fact checked everything to ensure that it's correct and it's precise and it's accurate. Also, if you're trying to take the easy way out, for example, if you're trying to get ChatGPT to write an essay for you or create a content for you. There are a lot of things that are easily detectable. As you've seen, ChatGPT gets really repetitive when it's creating content. Especially the longer the content, the more repeats certain words. There's a lot of different AI detection tools out there. So the market is going to get flooded with a lot of content that is generated using such tools as ChatGPT, I'm Bing Chat. So what's gonna happen is a lot of these platforms and applications are going to put in AI detection tools in place so that they can filter out the low-quality content and focus more on the higher-quality. So again, please do not copy paste everything you see here, and what you should do is you should use it as a starting point to get ideas and inspiration. And in terms of how you want to structure your work, whether it's a blog post or a social posts such as Instagram or Facebook. What the composition is going to look like for your blog, for example, and so on. And as we discussed earlier, you really need to be aware of data bias because the output is going to be based on the training data. And this is something that the providers and the builders of these applications also need to be fully aware of because if the training data is bias while the output data is also going to be biased. So something to be mindful of, so that whenever these models are getting trained and needs to get, they need to get trained on diverse data and not bias data. Also do not use to create derived, use these tools to create misleading or false information and always be aware of potential legal issues. For example, do not infringe someone else's intellectual property and rights using these tools? 41. 41 Thank You: We have now reached the end of the course. So I just wanted to say congratulations for finishing the course and thank you for enrolling. It's been an amazing journey. And thanks again for taking the time to go through the lectures. By now, you should have the necessary skill set to use the Generative AI tools to help inspire you and help increase your productivity and life. I really hope that you have enjoyed the educational content and discourse and can take the learnings and apply it to your everyday life. And I wish you all the best in the future