I recently started 3rd year of my CS degree, and I realised something, I am one of the first generation of college students who got access to a whole different paradigm of learning with AI. Upto this point everything I heard about it was theoretical , the most significant anecdote in my memory about AI stems from my 6th or 7th grade English book, an Isaac Asimov story. [1]
Coming back to the point, I wanted to talk about how it has changed my daily life , to be clear I am not claiming this has revolutionized my daily life, but I noticed some subtle change in my habits , some for better some for worse.
Two of my obvious and most significant use case as a college student with AI is writing/learning & coding, there are others too but these use-cases had significant impact on me as a person.
When I started using LLMs, I was surprised not by its ability to write but its ability to adapt different writing styles. This opened lots of doors, some good like I no longer have to spend too much time writing applications/emails for some mundane task and some bad like now I could easily slack through my creative classes at college, I no longer found writing technical essays or homework's that required hours of research a tedious tasks ,since AI would generate for me within seconds Although sometimes it did get me into trouble with cited sources but I soon learnt the trick with prompting could help with that but the million dollar question is
Well, When I was in high school, we were often asked to write answers to literature questions spontaneously, and this exercise was done so often that I really feel it improved my ability to articulate things, certainly I might not be a great writer but I am able to atleast articulate things well. ( I hope this might be obvious at this point but English is not my first language, but I have been learning it from past 15 years, so yeah I should be good by now). So, I really can’t comment if it has degraded my capability to write well but it certainly has made it better to express it better ( which is certainly cannot be said for the second use case. More in next section).
AI has certainly made me lazier, when I recieve an email or I write something, I try to write a rough/raw reply to it notepad and ask it to improve it later and after fair amount of prompting I get satisfactory results. For example, let’s take this blog I am pretty sure before publishing I would pass it through AI to improve the tone and grammatical errors. Wait! Now when I think about it , I will also attach my manually written draft, so if you wish to compare my writing style and published one, you can do so , it’s certainly not a good metric but nonetheless its a metric !
At the same time it has various advantages for me, now when I apply for various internship I no longer have to write my SOP’s several times from scratch, I can just prompt-bomb [3] it with all the info about my the internship , my goals , my resume and it whips up a great first draft and from there with a bit of prompting things become a lot easier, what would have taken me days to write , could be completed in matter of hours. So that’s definitely going into a plus column.
Now coming to the second and most controversial use-case ( sometimes it keeps me up at night !!)
I feel if you are reading this blog, there is a great chance, you have used it for coding, and before we dive into this , a caution I used to big believer in AI boosting your learnings, I still am, but for coding purposes I am not sure anymore. Before we go any further we need to clarify some things including “What does being good at coding mean?” . For the purpose of this blog ( and for me personally) it means to be converting your thought/ideas into a practical software application with a reasonable understanding of its working. Of course, you can disagree with me , I am hoping you do :)
Now, that we have defined it, let’s talk about how AI has shaped me , when I entered college I was given the same advice that every years millions of freshers recieve, learn the theory, practice and implement , and if you get stuck “google it, use stack overflow, refer to documentation etc. “ but as soon as AI arrived ( chatgpt in November 2022) , things changed a lot, unlike math's, AI was good at coding ( here I mean it was able to generate code for whatever I wanted it do) and suddenly most of my tasks were being outsourced to you. Mind you, it’s not that I was not doing/trying my homework but when I got stuck or maybe I didn’t know how to start, I gave or framed the problem to AI, and voila a working answer is in right in front of me.
While at a surface, it might not seem a huge problem, but let’s take an example , my OOPs course, I read and implemented all the examples from my course book but when somebody told me to write an example problem from scratch I was blank , and here the things change from legacy learning, when you didn’t have AI, your option was to google it and hope to find the exact answer for the problem you are looking for or go through the text again to have a better understanding, but with AI, you don’t need to do all this, just give it the problem or your thoughts and it customizes the answer to your needs, and I was able to understand it , so it felt like I satisfied all the criteria , I was able to convert my thoughts to code, I have a reasonable understanding of it and it works !!!! So, I thought I was getting good at coding, at first glance this might seem very obvious , and you might be saying this to the screen “You’re moron , you didn’t write the code thus you’re not good at coding” but hear me out its not that obvious , It’s not like I didn’t understand the concept I knew when I had to use inheritance, its just I wasn’t able to translate into code, I used a tool to do so, I told what I wanted to do, it did for me , I read it and understood it and implemented it. While this gave me some sense of confidence and understanding , it took something very important from me, which I realised very late, my knowledge of what not to do, In legacy learning you go through lots of sites, lots of trial and error, and in order you learn what not to do….. and this I missed on with AI.
It’s not all gloomy and sad , I mean at this point it’s fairly obvious I should stop using AI for coding , it would solve all my problem , but hear what benefits it brought for me. To give context, I am a guy who easily get’s intimated with things especially in tech, like if I read something once or twice and didn’t understand I would think I don’t have a pre-req understanding for it ( which is only somewhat true in CS, that’s a topic for another day). It would have stopped me learning a lot of things, but with AI , the possibility was limitless , I wanted to learn about AI in my first semester but by seeing all the maths terms, it would have stopped me but with chatgpt I was able to perform MNIST classification in python , with only a high level understanding of things, sure this might not seem a big deal but think with a perspective of a 18-year old kid who recently learnt loops in python, this was a big deal, but if you notice I missed a step for being good at coding, “ I didn’t have a reasonable understanding” and to overcome this I soon found a loop-hole I started learning about libs like sklearn, PyTorch, I started looking about APIs being used, and soon enough I started to understand what is happening in the code, so in a way I started doing reverse learning, I first converted my idea into code and then learnt how and what was happening in my code, this does satisfy the criteria of “reasonable understanding”, and soon enough I was able to do all kinds of things with it , I was able to run before I could walk well, but I didn’t realize I was using a walker[4] to run for a long time… I call it the “Curse of AI” I can understand code but I can’t write it.