Create a YouTube AI employee with ChatGPT and Python

Lecture Topics 📖

• Learn how to use the GPT API to make your desired program work as a chatbot.
• I created a GPT-based app that creates playlist videos through conversations with a chatbot.
• If you tell the chatbot what kind of atmosphere you want, it will automatically download the audio, create images that match the song, and make a video.

Demonstration of the program created through this course

Discuss the contents of the book with the author.

Creating a YouTube PD app that does everything from planning, preparing materials, designing, to video editing

• On November 10, 2023, the best-seller published by Easy Publishing!! < Do it! Create AI Employees with Chat GPT & Python > You can watch the book's author and video.
• Where to buy
  • Kyobo Bookstore (December 2023, 4th place in IT Mobile!)
  • Yes24 (December 2023, IT Mobile Top 20 4 weeks!)
  • Aladdin (January 2024, Computer/Mobile Top 100 9 weeks!)

Course Target Audience/Course Purpose 🙆‍♀️

Types of students that knowledge sharers think of

• People who want to utilize ChatGPT's chatbot function in their programs
• People who want to learn how to use OpenAI's API
• People who have learned the basics of Python but want to try a new side project
• People who want to automatically create YouTube playlist videos

Students' Concerns & Solutions

• There are many tutorials on how to use GPT, but it is difficult to find a way to actually create a program that suits my purpose.
• In this lecture, you will learn how to combine the capabilities of GPT with the features (functions) that I created.

Changes in students after attending the lecture

• Learn how to use the GPT API and how to create the chatbot you want using prompts.
• Learn efficient ways to program and their limitations with the help of ChatGPT.
• Create a playlist video by talking to a chatbot I created.

Lecture Features ✨

• There are many lectures that provide fragmented knowledge on how to use GPT, but there are not many lectures that cover the process of controlling hallucination of GPT and combining it with functions that I programmed myself.
• In fact, we will go through the process of creating a chatbot by combining the advantages of GPT with a program that has a clear goal.
• In the process, you will experience the advantages and limitations of GPT, and you will be able to create your own actually useful program by taking advantage of GPT's advantages.

What you'll learn 📚

This course develops the program by developing the function I want and then linking that function with GPT to make it work. The content developed in each section is not limited to independent content, but is connected so that it can be used in the form of a chatbot in the program we are writing.

Section 0. Introduction
In this lecture, we'll walk through the program you'll be building, talk about how it works, and cover setting up your environment before you start the project.

Section 1. GPT API Basics
For those who are new to the GPT API, we will explain how to use it. You will learn how to change the prompt to induce different answers from GPT. And you will learn how to change it so that it can be used in chatbot form.

Section 2. Making it an app
Convert a program that runs in the terminal to an application for Windows or Mac. In the process, you will learn how to collaborate and the limitations of programming with ChatGPT.

Section 3. Using GPT’s Creativity
We explore the uncertainty and creativity of GPT, and develop it into a DJ chatbot that selects music. We also cover how to control the format of GPT's responses to the form you want.

Section 4. Controlling Hallucination in GPT
GPT often gives wrong answers. We learn how to control wrong answers so that we can develop the program to our desired purpose. As an example, GPT checks YouTube to see if the song it selects is a real song, and if it is, it downloads the audio in mp3 format.

Section 5. Image Generation Using the Stable Diffusion Model
Generate images that match the song. Here, we will cover how to generate images using the model you like from Hugging Face. We will summarize the information of the song using GPT, and then use the summarized information to generate images using text-to-image.

Section 6. Image generation using Dalle-2
Generate images that match the song. The method in Section 5 above requires the GPU performance of the computer. However, using Dalle-2, you can generate images regardless of the performance of your computer using OpenAI.

Section 7. Creating a Playlist Video
Finally, create a playlist video. You can use the images you created earlier, or you can create a playlist video using a video you recorded in advance.

Expected Questions Q&A 💬

Before attending the lecture, please write down at least three questions and answers that students may have.
We recommend answers that reflect the personality of the knowledge sharer rather than obvious and formal answers.

Q. How well do I need to know Python?
A. This course does not require a high level of Python skills. If you understand the concepts of variables, loops, functions, and classes, you can follow along.

Q. Can I get the source code?
A. Yes. You can receive the source code as a github repository or compressed file. You can use the commit & push results as shown in the video.

Q. Is this a lecture that non-majors can also take?
A. Yes. This course is designed so that anyone with basic Python knowledge can follow it. If there is anything you have difficulty understanding, please ask questions through Q&A.

Things to note before taking the class 📢

Practice environment

• It was based on Windows, but the code was written so that it can also run on Mac.
• If there are differences between Windows and Mac, we will mention them separately or provide a separate video.
• Image generation using Stable Diffusion may not work depending on your computer specifications. However, we will guide you on how to run it on Google Colab so that you can practice. We will also explain how to use Dalle-2 so that you can create the final result without any problems.
• OpenAI charges a fee to use the OpenAI API, but since you receive $16 in credits when you sign up, you can follow this course without incurring any costs.

Learning Materials

• The code written in this course is provided on github and as a compressed file.
• The videos required for creating playlist videos using videos are provided as OneDrive links.

Player Knowledge and Notes

• You should know the basic syntax of Python.
• Questions are accepted through Q&A, and if there are any changes to the GPT API in the future, we will announce them separately.

Introducing the knowledge sharer ✒️

He is the author of "Do It! Django + Bootstrap Python Web Development Essentials" and has also run courses on Python task automation and Django web development at Inflearn.

I am currently working as a data analyst.

_{The Essentials of Python Web Development> Author}

_{PYCON Presentation : Django + Selenium
Creating a statistical work tool with Python}

_{Seoul National University Regional Systems Engineering
(Agricultural Eng.) Doctor of Engineering}