Jonathan Li

who am i?

That's a hard question.

As you may have noticed, I have a fairly unconventional background for someone who calls themselves an ai researcher.

I started coding eight years ago when I was captivated by the magic behind a neural network. I used to work at Cohere on synthetic data, reasoning, and RL infrastructure. Prior to that I published NLP research in an academic capacity at Queen's University. Now I'm building a company (with my brother again).

I’ve always been interested in the existential questions—which is likely why my interests bounce around so often, from AI to physics to building companies—though I’m honestly unlikely to succeed, at least on my own.

I’ve recently started thinking a lot about my nonconventional path. This introspection has given me a lot of conviction on how we should build the next generation of ai.

If you want to learn more about how I think, take a look at some of my writing.