The Death of Cool in AI Research

The Death of Cool in AI Research

Arshia
"Science is what we understand well enough to explain to a computer. Art is everything else we do." — Donald Knuth

Something I’ve noticed a lot—honestly, felt it myself too—is this crazy FOMO that everyone in AI/ML seems to have. There’s this weird rush to jump into projects or, as we love to say, “get into research.” It’s like everyone’s running to do something just to feel like they’re not falling behind.

At first, I thought it was just an Iran thing—like, people want to buff up their résumés for applying abroad. So yeah, maybe spending time on foundational stuff doesn’t feel as urgent because the focus is on cranking out results ASAP. But when I started checking out what’s going on internationally (thanks, Twitter), I realized this rush isn’t just about résumés. The field is growing so fast, and the hype is so intense, that it feels like everyone’s scrambling to catch up—bingeing ML/DL/LLM courses and signing up for research before they even know what’s going on.

It’s not just students either. Even course designs reflect this. I'm telling this since I was looking at Sharif’s ML syllabus this semester, and it’s super DL-heavy. There’s barely any room to slow down and dive into the details or core concepts anymore. It’s all about moving fast, and honestly, that’s not just an ML thing—it’s the vibe everywhere right now.

The field itself is becoming a little bit boring cause back in the day, doing ML meant designing models, playing with architectures, and solving complex, puzzle-like problems. It was creative. But now? It’s like the field has turned into Lego for grown-ups. Everyone’s just stacking pre-trained models, calling APIs, or tuning prompts. Sure, it’s efficient, but where’s the creativity? I'm always getting this LLMs or AI in this “assembly-line” vibe is everywhere.

I’m still at the start of my journey, and yeah, I’m guilty of following the herd sometimes. But I try to keep some perspective about where I’m heading. Even if my “vision” for the path ahead isn’t perfect and has a ton of bumps, it’s better than running blind. That little clarity goes a long way, man.


Report Page