this is my self-study plan to learn AI, which I made using ChatGPT. This one is for learning Japanese. This one is for learning how to make better YouTube videos. And last but not least, this one is for learning personal finances. These are the four topics, the four study plans that I'm currently following. They're all custom-designed to fit my specific goals. Like for the Japanese one, I don't care about how to speak Japanese or read Japanese. I just want to learn how to understand Japanese so that I can watch anime without subtitles. I know it's pretty specific, but yeah, they're all custom-made at my study level, my skill set, my preferred resources, and my preferred study schedule. Before ChatGPT, each of these study plans would have taken me days, if not weeks, to get to this level. But with ChatGPT, I was able to do each of them in less than 10 minutes after I got the prompt right. Though the prompt itself did take me weeks to actually figure out. But luckily for you, in this video, I'm

going to just show you guys this prompt to make a really optimized custom study plan using chat gbt that works for any subject. Plus, I'm going to give you some bonus prompts and tips and tricks on how to chap BT to make the actual studying process faster and also more fun. A portion of this video is sponsored by HubSpot. All right, let's have that start running while I explain the prompt. So, for this example, I'm going to make a study plan for learning AI. You can change that for any other topic that you want as well. All right, so the prompt: you are an expert learning tutor who is well-versed in Scott's book Ultra Learning. So, that is from this book over here. I have tried a lot of different frameworks from a lot of different books and resources, and by far, the best one for how to create a good study plan using chat gbt, um, this framework in this book works the best. So, I actually covered this um, book in detail. You can check it out in this link over here. So, I'm not going to go

into much of it now, but just trust me on this one, okay? All right, so moving on. So right now, I want to make a study plan for Ultra learning AI based on the principle of metal learning and the why, what, how framework. So that is the framework from the book. Why is why I want to learn it. What is what I will learn. This is divided into concepts, facts, and procedures. How is how will I learn it. This is based on techniques of benchmarking and emphasize SL exclude. Benchmarking means define common ways people learn it by doing research. Emphasize such exclude means for making modifications to align with the goal. My why for learning AI is because I want to build simple AI web applications using LMS and other AI products. The why is extremely important. Make sure to think carefully about why it is that you want to learn this topic and tell Chacha BT. The more specific you are, the better. Like my Japanese one is specific for learning how to understand anime Japanese so I don't have to read subtitles. My personal finance one has

specific goals about buying land and property and for making YouTube videos. I say specifically it's for making video essays to maximize for reviews. I already know how to code in Python. I know how to build Python applications with Jeno, so I understand the components of a web application and how they work, etc. This is where you tell Chacha BT what you already know about that topic so it's able to adjust its skill level for you. So let's create a study plan using this framework. Let's start with the what. For the what in your study plan, we'll break it down into conscious facts and procedures. So it tells me the concepts that I need to learn, such as AI and machine learning basics, natural language processing, pre-trained models and APIs, web application architecture for AI apps, and ethics and bias in AI. And for facts, key libraries and frameworks, data formats, pre-trained model specs, and performance metrics. And for procedures, um, the actual steps I need to take. So all of this is very specific to my goal and looks very reasonable. Um, by the way.

the reason why I didn't give you like one massive prompt, and I said to start with the what is that in this way, it offers you flexibility. Like, for example, maybe I don't care about ethics and bias. I should care about ethics and bias, but I could say something like, I don't care about ethics and bias. He's removed that. But yeah, see, this is how you can kind of iterate on top of it so that you get the concepts, facts, and the procedures that you want to have. All right, let's actually add it back in. Okay, so let's now move on to the how of the framework. All right, we asked Chach BT to do benchmarking, which is through research common methods of learning. So, you and I are not the first people who want to learn how to use AI to build applications. So, instead of trying to come up with something from scratch, we can actually ask Chach BT to research the collective knowledge of people who have come before us, and what they figured out is the best way to learn and

Based on that, we're doing emphasis exclude, so we're emphasizing things that are targeted towards my goals and excluding things that are not useful towards my goal. So, in this case, from doing that research, we know that people use beginner-friendly courses such as Andine Ning's course on Coursera or Fast AI. People use APIs and library documentation. Project-based learning is really important, online communities, and GitHub. And for ethics and bias, looking at case studies and ethical guidelines. So, we're emphasizing project-based learning and API integration of web applications because we want to build a web app using LMM and AI models through APIs. We need to have AI fundamentals, so the focus on large language model specifically. And for bias identification and mitigation, what we're excluding is in-depth theoretical ML/deep learning knowledge. Since I said that I wanted to be applied, I just want to build these applications, not like build my own models and deeply understand them. That's why we're excluding this. And also, we're excluding dataset creation and fine-tuning. These are things that we

might want to do later on, but since I just want to build an application right now using the APIs, we can keep things simple and exclude that in this current study plan. All right, now we wanted to combine together the why, what, how, and come up with a comprehensive study plan that also has resources and timelines. So, I said combine these things together and make a week-by-week study plan with specific deliverables. For these deliverables, I said make specific assignments and projects with rubrics and specs that you can grade and provide feedback for. Um, this is going to be really helpful so that it doesn't just tell you to like learn something very vague. We wanted to give you specific assignments that have objectives, and later on I'll talk about this as one of the later prompts. You also want ChatGPT to be able to provide you with feedback as you're completing the assignments and going through the study plan, so we want to bake that in while we're making the study plan itself, and we want corresponding relevant resources. I said

that I can study 6 hours a week, with 3 hours on Monday and Wednesdays, and I want resources that are primarily text for video courses because that's just how I learned best. All right, and here we go. Week one: Introduction to AI, APIs, and LMS. The objective is to understand AI basics and how pre-trained models and APIs work. Concepts are going to be AI and machine learning basics, natural language processing, and LM. The facts, so it lists out what the facts are and what the procedure is going to be. It has a deliverable, so it says create a Python script that calls the OpenAI API to generate text based on the prompt. It was very specific about what it is that I actually need to accomplish, and this rubric is helpful because it gives you an understanding of what should be included in the deliverable. And we have our little resources as well. So week two is going to be integrating APIs into Django applications. So week three is working with image recognition models, so not just LLMs, and we have a deliverable.

to create a web app that accepts an image input of a plant and returns information about the species using a pre-trained image recognition API. Very interesting. I don't know why it's about a plant, but seems good. Weave four is your ethics and bias, and weave five is when we build our full-stack AI application. Very excited for that. And week six is going to be for debugging, testing, and refining. Overall, it's going to be 6 weeks, and by the end of it, it looks like we would have built an AI application just like how I wanted to. It tells you exactly what it is that you need to learn, and it has like project-based deliverables that you can apply to skills that you learn, and it even has a rubric, um, so you know what are the components that make up that project, and of course, their resources based upon what other people have used in the past, and they found to be good. Not to T my own horn here, but I'm genuinely pretty proud of this prompting sequence to get to this point. You have no idea how long

it actually took me to get to this point, so yes, just let me have this one, okay? Um, but we're actually not done. There is one more final step. I hope this video has made you realize how useful Chat GPT is in terms of learning things, but actually, that's just the beginning. It's relevant to increasing your productivity across all aspects of your life. HubSpot has a whole bundle of resources designed to help you understand and make the most out of Chat GPT. It will make you super productive. It's very well laid out. It gives you lots of ideas on ways you can incorporate Chat GPT. For example, using Chat GPT to consolidate research materials like different articles, tutorials, and resources, so you can directly ask questions instead of having to flip through all of them yourself or keeping you up to date with industry news and updates. They also have this nice flowchart on when it's appropriate to use Chat GPT to solve a problem or streamline your workflow. My favorite part is that it has over 100

prompts to start playing around with, which is great since, as you can see, prompting is extremely important to getting the most out of chat PT. The best part is that it's completely free. I highly recommend that you download it using this link over here, also linked in the description. Thank you so much, hopsa, for providing free resources to help us leverage the power of AI and for sponsoring this portion of the video. Now, back to the video. Ideally, you want an expert, someone who already knows the content of your study plan, to look over your study plan and just to double-check that things make sense. You can do this if you actually know someone who already knows AI or knows Japanese or is really good at personal finances, and they can look over your plan. But assuming you don't know anybody, this final step is you're going to ask chachu B to now be an expert AI engineer that specializes in developing AI applications, and we're asking this AI engineer to look at the study plan above.

as if it's for the first time and to provide feedback on what they think and also what they think should be changed. So as an AI engineer, here are some strengths: Project-based learning is a good balanced approach, API Centric rubric for deliverables, as ethics and bias consideration. So some suggestions for improvement: maybe you want to have more emphasis on debugging and error handling because building applications often involves dealing with unforeseen edge cases, especially when integrating multiple APIs. And it gives you specific changes like increasing focus on error handling and then adding a mini project in week three where you intentionally introduce errors or unexpected inputs and learn to handle them. This is actually very common practice when you're debugging an application, so very good point here. And also introduce real-world data earlier and more frequent iterations and refactoring. So you can choose to incorporate some of this feedback or you can choose to ignore it. So in this case, I'm going to choose to include some of this feedback. Thank you. Please include

suggestions one and two into the study plan, and voila, there you have it. We have created a custom study plan for learning how to build AI applications. Don't worry if you didn't catch everything; I'll leave a link in the description that has the prompt sequence in it. So, now that we have the study plan, the next step is that we actually have to study. Got to study, got to study. I also want to share with you guys some of the prompts and the little techniques that I use where I can get through the study plans so much faster and like a lot more fun as well. First, one of my favorite ChatGPT study use cases is using ChatGPT to create test questions. A couple of months back, I had to take the knowledge section of my driver's exam, but I didn't really feel like reading the hundreds of pages of rules related to driving. So, instead, I just gave ChatGPT those documents and told her to ask me questions. If I don't know how to answer a question, I would just ask it to

explain it to me, and that was just a much more fun learning experience. You can use this technique even if you're not actually studying for an exam, but in this case, since I was studying for an exam, I also gave it samples of some of the exam questions from the test. So, this is even more effective because, as I learned, I was also actually practicing taking the test multiple times as well. I'm happy to report that I did pass the exam, and it took me 2 hours of study. I also like to use the fan technique with chat BT. This was named after Richard Fan, who has this method for deeply understanding concepts that he's learning. He's a Nobel Prize-winning physicist who is known for really deeply intuitively understanding things. So, step number one is to choose a topic. For example, I want to learn the role of emotions and how it relates to storytelling as part of my study plan in making better YouTube videos. So, step two is to write down what you know and

I'm just writing this down to chat GPT. Step three is to teach it, so in this case, I'm going to be teaching chat GPT about everything that I know. So, for example, during that process, I realize I don't actually understand the relationship between positive emotions and negative emotions. Like, are we supposed to have one positive, one negative, or does it really matter which is when I can do step four, which is identify the gap in my knowledge and try to fill that in. In this case, I can just directly ask Chad GBT to help me fill in that gap so much faster than having to flip through resources using the traditional fan technique. In fact, I realize I can do this even faster when instead of having to write everything down, I can actually use Chad BT's speed functionality and actually say it instead. I want to try using thean technique with you to learn about the relationship between emotions and storytelling. So, Def and techniques first, I'm going to choose a topic, which is the relationship between emotions and technique. Next, I'm going to write

down what I have to know, and then I'm going to teach it, and then finally I'm going to identify gaps. So, actually, I'm not sure about the role of emotions in storytelling. Like, are all the emotions equivalent to each other, or are some emotions more powerful than others? Emotions tend to have a stronger impact on audiences depending on how they're used. For example, emotions like fear, love, and hope are often seen as speaking of Chip's audio function. My next tip is by using the audio function as a way to do things like mock interviews or practicing speeches. It is a game-changer. I'm applying for a data science position, and I have an interview next week. Could you act as the recruiter, the interviewer, and ask me some behavioral questions? Sure, let's get started. Here's the first behavioral question: Tell me about a time when you had to solve a complex problem using data. How did you approach it, and what was the outcome? During that time, I was working for a pretty traditional company as a

consultant. They had a lot of data, but it was scattered all over the place, like in different databases, even Excel files. So, the first thing I had to do with that data was to put everything together. That was definitely the hardest part. I used SQL, I used Python, and then just to establish that source of truth. Um, and then after I did that, I was then able to take that data, um, and then do a dashboard that they wanted me to do. After that, I made sure to help everybody and all the stakeholders understand that this was the source of truth and try to also get the teams to start putting all the data into a single database instead. The outcome was that things were a lot more organized, and also there was now a um dashboard that had the metrics on it. That's a great example of taking initiative and organizing a complex data set. Here's a follow-up question. Another tip you can also ask to provide feedback from my previous question. Can you also help provide feedback um in terms of my answer and how I can improve your answer?

did a great job of showing that you can handle complex, disorganized data and take initiative to improve processes. You demonstrated key skills like using SQL and Python, creating dashboards, and managing stakeholder expectations. That's exactly what data science employers are looking for. Here's some feedback to make your response even stronger: 

1. Structure: It could help to use the STAR (Situation, Task, Action, Result) format to make your answer more organized. This makes it easier for interviewers to follow your thought process. 

2. Highlight Impact: You mention the outcome but could emphasize more measurable results. For example, you could mention any improvement in efficiency or how the dashboard helped the business make better decisions. 

3. Teamwork Collaboration: While you focused on the technical aspects, adding detail about collaboration with team members or stakeholders could strengthen your response. 

Okay, so I have so many other little tips and tricks and prompts for ChatGPT for studying, but I feel like this video is already hideously long, so I'm just going to leave you with one last one, which is when you want to practice something specifically. For example, when I was learning computer science stuff.

something that really tripped me off was dynamic programming. I was just like, dynamic programming boggled my mind, so I needed to drill into that and just like really understand it. So, I asked Chach PT to specifically only give me questions related to dynamic programming. I also asked it to give me different question types and for me to identify which ones would have dynamic programming as a good solution to kind of build that intuition for when dynamic programming is useful or not. Another example, if I'm trying to learn how to draw and I'm just like very bad at drawing noses, I would ask Chach BT if it can give me examples of different types of noses that I can drill and master the art of drawing noses. All right, that is all that I have for you guys today. Thank you so much for watching. Um, please let me know in the comments what you're studying and then also link some of your study plans. I want to see what you guys have as well. Um, and yeah, I will see you guys in the

How To Make The ULTIMATE Self Study Plan With ChatGPT (step by step prompting guide)

Hi! My name is Tina and I'm a ex-Meta data scientist. Now I create content and other internet things! 

This channel is about coding, tech, career, and self-study. I love learning new things and my goal is to always minimize effort and maximize outcome! 

I strive to do things in the most efficient way possible and with the highest reward. I basically refuse to do extra work to achieve my goals c: 


