Are you getting dumber with AI? (Programmer POV)

Are we becoming dumb with AI?

This question is worth wondering upon. Are we getting dumb with AI?

Is using AI making all of us less capable of producing or creating something unique?

Before I continue further, this is hand written article and my english isn’t that good(in terms of writing). So pardon if there are some spelling and grammatical mistakes.

Reason of this thought came from when I noticed how learning has shifted over the years. I remember before AI — we would approach problem in a structural way.

• Open book, docs or whatever — Go through basic concepts, theories and how why something works the way it works.
• Then jumps directly to implementation part — which forces you to face errors , which frustates you.
• Then you re-inforce yourself to understand things at deeper level, and then you master it.

But now comes the problem, AI took away that friction and this is one of the reason why I feel like It’s making all of us average if not dumb.

Okay hear me out — As I am programmer so I will explain from that perspective.

I am not talking about AI writing faster code than me, I love that part — but something else happens too. It took my friction away.

For example:

• I never cared about AI wrote this line of code instead of me.
• But when it spits 1000 lines of code at each iteration and all you can do is stare at it.

Either you spend time understanding each line, or keep ignoring those and vibe code/blindly accept everything.

Now understanding part is solved for those who are already experienced in their domain. But it wouldn’t turn out well for fresher, and by fresher here I mean you as well — regardless of how many years of experience you have in your domain.

Just because you and me have learned over the years concepts, and did our journey in mistakes to learn things, doesn’t mean that we will do same in future. In future, we also might do exactly how majority of population is doing.

Now there is no problem in this as your are doing things at fire speed. but so do everyone — and that’s one of the reason you might see all vibe coded apps and websites look same, those Neon layout + gradiant font, which is good but its average. And those vibe coders or people in general don’t have much of freedom because their understanding haven’t reached enough so that they can write custom instruction detailed for AI to follow. And as long as AI taking decision on their behalf, more or less it will follow same pattern.

This thing creating bunch of average people instead of creating innovators. And if we categorize then we are left with these category of people:

• Section A : Who hates AI and like we don’t give d*mn. I mean they will sooner or later, but they are still doing it in traidtional way.
• Section B : Who doesn’t know fundamental and don’t have strong foundational knowledge, yet as AI doing things — they are relying on it, but the problem with this one is that, they will all be producing same crap, if they don’t understand concepts enough to create a reciepe.
• Section C : These are lucky ones who got their foundational knowledge from experience but still in danger if they rely on AI blindly when they are learning new things. because then they will become Section B
• Section D? : <I don’t know this one, we gotta define this one, This section will have to build a framework through which they can accelerate foundational knwoledge while building stuff> — Mix of Section B * Section C.

Listen, I am not saying AI ain’t helping — I for real would love to skip part where I am forced to remember how this specific code is written, but having an understanding will make a lot more difference.

• Person 1: Imagine once person prompted: Build me beautiful web app for business with good UI and animation , millions of other will prompt the same, and they all get same.
• Person 2: Instead of jumping straight, he learned what color theory is, what tools and techs are being used, how GSAP, Framer Motion, Lenis, 3D, webGL, etc works. what happens in each of them, why not everyone uses it, what are drawbacks, when to use and when avoid. How to draw concept and pick color pallete, etc… (Again its shallow level understanding, but even with only this much of cociousness if one would build that app, it will look far better.)

Again I am not saying you to focus on those repetitive useless question that have 0 value, but we also have to make sure in this process that we at least grab foundational level understanding before we jump, though a funny story — I had been giving interviews right now, while I can explain my projects, and how I have built features and things, I am still failing at sily things like: typeof null , or write this query which I had written in far optimized way, but again I can’t recall or even write it in syntax wise( I need to do Top 50 interview question and I will clear those). But If I have code I can explain it regardless. So right now, things are still traditional for while I guess in job market.

But this article was designed more in a way to discuss how a learning would and should look like in upcoming era, where humans should pause and dive deep and when he should go with flow… And who will enforce that habit in them, because we don’t then all we would produce average.