People often ask me, “Why are you into NLP? Why LLMs? Why not computer vision, reinforcement learning, or some other cool field in data science?”

And honestly? The answer has always been… process of elimination. I tried computer vision, and it just didn’t click. But I’ve always known that’s kind of a weak excuse.

If I really think about it - why am I drawn to language instead of images? After all, there’s that saying: A picture speaks a thousand words. Shouldn’t an image be more efficient than a wall of text?

Maybe there’s something deeper that I haven’t fully unpacked.

First guess? Language has always been my weakest subject. Back in school, my English grades were always the lowest among my friends. I wanted to improve, just to keep up. Over time, I did - but I was never great at it. Maybe, on some subconscious level, working with NLP is my way of continuously improving my relationship with language.

Second guess? When I got into data science, I dabbled in different areas - GANs, structured data models. A lot of those fields felt… established. Well-researched, predictable, a bit too solved. But language? Language is messy. Text is unstructured, ambiguous, full of nuance. Yeah, a picture might tell a thousand words in an instant, but no one’s out there processing a thousand words for you in a split second. And that’s what excites me - making sense of all that chaotic, free-form text.

From a personal perspective, NLP just feels more useful to me. As an individual, I’d rather have something that helps me process vast amounts of text quickly than something that identifies objects in an image. That’s where I see the most value - not just for businesses, but for myself.

So I guess that’s why NLP and LLMs pull me in more than CV, RL, or anything else. It’s not the most logical reason, but hey - sometimes, you just gotta go with what sparks your curiosity.