Build Your Own ChatGPT: Python AI Assistant Explained

I am Panka Shukla and we start this session. Imagine you want to ask Chad GPT a

question. So what you do? You type explain artificial intelligence in simple words or any other questions. For

example, what is Newton's law? Any kind of question you ask to the Chad GPT. Then what happens within a few seconds?

Chat GPT gives you an answer. Have you ever wondered what actually happens behind the scene? Actually, it

is much simpler than most people think. What is the process? It is like this. You ask a question

and your Python programs takes that question. So your Python program takes that question as input. After it takes

that question, it sends this question to the AI model. [clears throat] What it does? It sends it to the AI model. AI

model returns the response. So what AI model does is it returns the response. After the response is returned to your

Python program, your Python program just prints that response to you. So it just prints this response

to you. So basically your Python programs take the input from the user that means the question and then it

sends that question to the AI model. AI model prepares the response sends it back to your Python program and your

Python program just prints that response. This is what the process is. Now what basically an AI assistant is?

So chat GPT Germany grock these are AI assistant. So what basically an AI assistant is? Is

it is it a software application that can understand your instructions and generate useful responses. For example,

it can answer your questions, right? It can explain the concepts. It can write the emails. It can generate computer

programs. It can summarize the document. It can translate the languages for example from

English to Hindi to Russian. And it also help you brainstorm ideas. So you must be using some kind of AI assistant right

now like chat GPT or Germany what I have told you. So notice something very important. What

is that? The AI assistant responds to your instructions. It does not decide what to do on its

own. That difference becomes very important later on when we build AI agents.

Right now what I am talking about is AI assistant not AI agents. But we have to understand the difference for now. Think

of today's project which we are going to build today as your own personal version of chat GPT. So today we are going to

make our own personal version of chat GPT through Python program. So since this course is about AI agents,

as I already told this course is about what? AI agents. You might wonder why are we building an AI assistant first?

That's a great question. So try to understand. You don't begin with a Formula 1 racing car if you want

to learn driving. What you start with is a normal car. Similarly, before building intelligent

AI agents, you must understand how an AI assistant works. So AI agent is simply an AI assistant plus something with an

additional ability such as planning, using tools, remembering information, taking actions. This is very important.

Taking actions. Your chat GPT does not take any action. It gives you only information. So that basically is AI

assistant. When we will make AI agent, it will also able to take independently few actions.

So what we will build today? We will build a Python program where user can ask a question.

That question we will send to an AI model and then we will get a intelligent response and we will display this

response and continue the conversation. I hope this is clear to you.