ChatGPT Tutorial – For College Students and Coders

In November 2022, a company named OpenAI released a chatbot named ChatGPT. This chatbot works on Generative AI, artificial intelligence that generates text or other information when provided with a prompt.

Now ChatGPT is limited to text only, which means it generates textual information for us, but there are other AI generators in the market. More tools are available in the market, such as image generators which can easily convert text into images. Some apps can even generate videos for you, but ChatGPT is the most popular generative AI in the industry.

There is a small metric to measure the popularity of ChatGPT, its number of registered users. Within five days, there were one million registered users of ChatGPT. If we compare this figure with Instagram, which, in social media terms, is considered to be a very addictive platform, then it took two-and-a-half months for Instagram to reach one million users and ChatGPT only five days. 

Now we can use ChatGPT in multiple ways, inside coding, and in many general areas. We can use it to write articles, or we can use it to write a summary of a research paper. We can use it to write songs, poems, and stories. If there is any problematic topic like quantum computing that we want to understand easily from someone, then we can use it for that.

We can use it to help with interview preparation. If we have to make a machine learning model for ourselves, we can use it to generate some dummy data. We can also use it to generate code, and in fact, we can use it to solve our college assignments. 

But whenever we use ChatGPT, we have to take special care of one thing: ChatGPT is imperfect. ChatGPT provides a disclaimer acknowledging that the information it generates has no guaranteed correctness. In other words, the generated text may contain errors, biases, or harm.

Therefore, it is recommended for skilled individuals who already possess knowledge in a specific field to use ChatGPT as a tool, as it can help streamline their work and simplify various tasks.

Therefore, as a beginner or learner, if you aim to explore or learn something entirely new, ChatGPT may not be suitable. This is because we need to learn about a specific subject to know whether the information generated by ChatGPT is correct or incorrect.

So as a coder, as a programmer, we will get the maximum benefit of using this tool only when, whatever programming language we have or technology we have, it is already good enough so that the things we need to do redundantly can be simplified.

In this blog, we will explore ten things that can be done using ChatGPT. The first of which is to generate code. Now, to use ChatGPT, first sign up, and for that, visit chat.openAI.com by using the link. 

There will be two options: one for login and one for sign-up. 

If you are a new user, you must sign up first. Once signed up, you will get basic options to create an account using any email address.

If you have already signed up, then log in. After login, you will see an interface showing examples, capabilities, limitations, and whatever we discuss. In this space, you can write all the prompts and queries you want to explore with ChatGPT. So here, you will write your messages.

Now we will learn how to generate small codes using ChatGPT. For instance, the smileys on the internet. We will use ChatGPT to create smiley faces using HTML and CSS.

Generate Code

We will write the prompt – “Generate HTML and CSS code for a smiley face.”

ChatGPT has generated the code for the smiley face and explained what it has done. Copy and paste the generated code into a file in Visual Studio Code, preferably an index.html file.

Save the file.

When you run the index.html file, the generated smiley face code will look like this.

If you have knowledge of HTML and CSS and want to create a smiley face, there’s a more straightforward approach than starting from scratch. You can generate a basic template and then make adjustments to it. However, if you’re a beginner, it’s challenging to determine whether the generated code is accurate.

As we know HTML and CSS, we can fix it. How can we do it? 

Their eyes are fine. We can make some changes to the mouth. We can make adjustments to the generated code, especially for the mouth. Specifically, we will set the rotation to 0 degrees, and the shape will be like this: Furthermore, we can make the mouth circle slightly more circular and adjust its width to 30 pixels. As a result, it will appear similar to this. Additionally, we can increase its left distance, for instance, to 35 pixels, by moving from left to right.

Save.

Refresh.

Following these steps will achieve a structure that resembles a smiley face. If we focus on the shape of the mouth, we can further enhance the smiley face effect, i.e., an art, by giving it the shape of a line.

The core concept here is that ChatGPT allows us to generate code. Still, if we already know the programming language, we can easily modify any code that might be slightly incorrect according to our preferences.

Write Resume

If you want to make a resume for college admission or apply for a job as a fresher in any company, then ChatGPT can help you with Writing a Resume. 

Just key in all the details you know, and then it can generate a basic resume template for you. Now, we are going to prompt it. We will write the prompt, “I am a B.Tech computer science student. I have done multiple projects using machine learning. I want to apply to a company for a fresher data analyst role. Generate a resume for me,” and we will enter it.

It has generated this response, inside which we have basic details, objectives, and education. In skills, because I have already said that I have done machine learning and want to apply for data analysis, it includes the tools and technologies used for machine learning. It has included databases because that is also a part of the job, and different general things should be included in a good resume. If we want to talk about projects, a section has been made for them for work in progress, and if there are certificates, then for personal projects. So we got a basic template containing only textual information about a resume, from which we can start.

Practice for a job interview

The next thing for which we can use ChatGPT is to “practice for a job interview.” That means, let’s say, that I have applied for a software engineering position at Google as a fresher, and my DSA’s data structure and algorithms round is about to come, but right now, I don’t have any friends with whom I can sit and practice. I want someone to ask me questions about DSA for practice, so in this way, we can write a little thing. I can write something like, “Ask me three data structure and algorithm interview coding questions for a software engineer position.” 

So I will enter it.

And it asked me three DSA-based questions: “Given a string of characters, write a function to determine if it is a palindrome.” We will use the stack data structure. He has already mentioned this condition. Using the binary search algorithm to search for an element in a sorted list, it gave me three questions. 

I can also level up or down the questions; if I want some easy questions to be asked, then I can add here three accessible data structures, algorithms, and interview coding questions.

So now the general question he is asking me related to DSA is a very easy-level question compared to these. And if we want, we can also increase the level. Instead of easy, if I write complex here after 3, more questions will be asked of me after leveling up a bit.

So after leveling up a bit, more questions will be asked of me. So earlier, when the stack was being used, it asked me the question of Dynamic programming (DP) and binary and asked questions using graph theory. In this way, we can practice for DSA and theory questions.

I can ask the same thing, but instead of data structures and algorithms, I can write object-oriented programming interview questions for a software engineer position. 

So it has asked fundamental oops-related questions, class-related questions, and inheritance-related questions. This is another way we can use it as a tool.

JavaScript Timer

Next, we will generate a “JavaScript timer.” I am about to write HTML, CSS, and JavaScript code to create a timer on the screen. 

So a code has been generated for us, which we will copy and run. So something like this: a timer is running on my screen that is counting backward; if I refresh it, it will restart again, generating a timer of 1 minute. We can make some changes to it, like, if I want, I can start the timer at 2 minutes or give it 120 seconds. So I restarted it, and now it will begin in two minutes. Some of the basic code or basic things within well-known programming languages have been generated from ChatGPT.

Let’s take another example of the data structure algorithm. So I can generate my algorithms using ChatGPT. Let’s take a straightforward example of Bubble Sort. Generate code for Bubble Sort. So it has developed the Javascript code for Bubble Sort.

Translate Code into another language

And the next thing that we can do as programmers is translate this code into any other programming language. As we copy that code, we will go down and translate this code to Java. It cracked the entire code into Java for us.

Find Bug

And one more thing that we can do is use ChatGPT to detect bugs inside the code, i.e., errors or mistakes. If we want, we can copy this code and write to see its mistakes. What will we do by copy-pasting the same code? After removing the three swapping steps inside this statement of mine, we will now submit the chart to GPT, so let’s see if it can find the mistake inside this code. So it detected the error in my code; the swapping step was missing in our code. So what did it notice? If the function’s name is bubble sort, there should have been bubble sort here. It means the swapping step should have been included. It has generated a new code that provides for the swapping steps.

Generate an Optimized Version of the Code

Now there is another way by which we can use ChatGPT for code generation: to generate more optimized code. I can write it as “generate C++ code for swapping two numbers.” This is a straightforward code we learn quickly in programming, so what did it do? It will generate this code in a very usual way. This method is the most common way to swap two numbers, but we will provide the same input to create the code with better time complexity. If you have heard the word time complexity for the first time, it means more optimized code that runs faster in your system for more significant inputs. Here, we will write to generate faster code and write “generate C++ code for swapping two numbers with the best time complexity.” 

If we have to understand the basic meaning of best time complexity, it is to generate more optimized code. This code has been developed for us, and one thing you must have noticed is that these three lines are different from the previous three lines; the rest of the code is the same. Because what did it do now? The fastest way is to use the XOR operation in C++ and Java. It has changed a little in the form of solving two numbers. This way, we can use ChatGPT to produce a more optimized version of the same code.

Explain Code

Now the next thing we can use is to explain any piece of code. For if we pick up any code in C# that we do not understand because I have not learned C# Yet, to explain it, we can write “explain this code to me,” and then by pressing shift and enter, we can write my code in the following line. I can run it by pasting it. So it has already been detected that a C# code is used to send email using the SMTP protocol. They told me the usage of my variables separately. And in the last, it has also been said that the code you have pasted is an example; to run it, you will need actual email credentials using the specified SMTP server. So he has written a complete explanation and given me the same code. And in fact, it can detect errors even inside this code. If I paste this and remove this comma from inside, I say, “Find the error in this code.” Even then, it will detect the error from inside this code because it is a very general code, so it noticed the error, inserted the comma along with it, and gave me the correct code back. Now the codes that we are generating or the code that we are giving it are very general, and the maximum use of the chat GPT should be only for generating those available codes because if you give it a code that comes with too many complex or too many nuances, then the chances of error will increase even more. So if you are a software engineer, web developer, frontend developer, or backend developer, you have to create APIs again and again or make some button or some element inside the front end again and again. So a complete function or a snippet that might have taken you ten to fifteen minutes to write ChatGPT can be generated very quickly. Then you copy and paste it, and if there are some bugs inside it, the work of fixing them will be left.

Ask Git Commands

The next thing for which we can use ChatGPT is to “Ask for git commands.” That is, if we use Git-Git repeatedly, there are many commands, so now that we forget them, we can use ChatGPT to find which Git to use in which situation. I can ask it, “Git command to combine multiple gits commits into one.”

What did it do? It told me which commands I had to use and through which process I had to complete that thing. If I want, I can also ask for fundamental things inside Git, like “Git command to commit changes.” So there is a basic command it will generate and give me.

So it has generated the whole process of commitment and push for us.

Write Documentation

The next thing we, as coders, can use ChatGPT is to “write documentation.” If I talk about my old code, this was my C# code with Smtp. What can I do for its documentation? I can write “write documentation for my code.” 

What did it do? It has generated proper documentation for the entire trim piece of code. And its significant advantage can be in open source because many times we create many git repos, and if we have to generate a readme file, then what can we do instead of writing it ourselves? We can use ChatGPT to generate it. Or for college-level projects – if we want to develop our documentation, we can do that. If you are working inside a company and paste the big codes here and generate their documentation, it will generate more complex and incorrect output. But for simple things, we can use this well for things up to the project level.

Look for Roadmaps

And one last way we can use it is if we have to learn any new technology, such as programming or coding, or go into a new field, then we can also use it to generate a roadmap. 

Although we will not give much advice about this method, sometimes the roadmap that comes with it has multiple things in it, out of which we have to learn only one thing, but there needs to be more clarification about it. 

We can say, “Generate a roadmap to become a frontend engineer. Which technologies to learn.”

Now here, the process described is precisely the same, but if we ask the same question slightly differently, like “I want to become a frontend engineer, which technologies to learn?”

Like, “I want to be a software engineer. What should I learn?

Now it has generated a good roadmap for all those in software engineering. Still, as a fresher, you should learn programming languages and data structure algorithms and have knowledge of databases and any web development. If you do not do web and app development, that will also work. But that thing is not specified here. So only a human or a mentor or a senior or teacher in your college can tell that apart from this thing, you can do this thing. 

Apart from this, there should be knowledge of version control, which is good. If you learn at the college level, it would be good to know about testing and debugging, but you should also know about cloud computing. As a fresher, extensive knowledge of cloud computing is optional. If you are interviewing for companies like Google, Microsoft, or Amazon, it is seen as a plus if you have basic knowledge of agile development. Only a little is expected of you regarding agile development in a fresher role.

Soft skills are essential, but you will be judged fairly in tech interviews. So maximum things have been told correctly, but if you want to see it from a guidance perspective or from the perspective of learning new things, a human touch becomes necessary there. If you ask your seniors, you will understand better and get an experienced view.

But definitely, ChatGPT is a perfect tool. Suppose an experienced person already wants to use it to optimize his day-to-day work, which is the work of coding. In that case, this tool will keep improving with time, and we never know if we will get better AI tools available in the coming years to make the lives of coders easier.

Moonpreneur is on a mission to disrupt traditional education and future-proof the next generation with holistic learning solutions. Its Innovator Program is building tomorrow’s workforce by training students in AI/ML, Robotics, Coding, IoT, and Apps, enabling entrepreneurship through experiential learning.