Research Philosophy
Q: How do you choose what to work on?
A: I seek out things that seem more complex than they really are.
I treat research as an extension of learning. I just do what I did in high school to learn math, but now I'm learning things nobody knows about yet. The hardest part is distinguishing between what's well known, what's known but not well understood, and what's an open problem. When learning, you can assume everything is well known, but research is a lot more nuanced (e.g. what is entropy?). There are bad explanations everywhere, and most of the time it's lazy presentation, but sometimes it hints at something deeper.
Q: What are the applications of your work?
A: I don't know, and if I knew, I'd probably be doing something else. I'm not motivated by applications. My purpose is to invent better explanations for phenomena we deem inherently confusing and complex.
That being said, I can't imagine a world where understanding life and the physics of information does not have applications. If we were able to invent computers and AI before understanding computation or intelligence, imagine what we could do after!
Q: If you're trying to understand life, why aren't you researching biology or chemistry?
A: Imagine you're a young grad student in the 90s hoping to contribute to our understanding of intelligence. Would you choose to study neuroscience given what we know today? It seems obvious that you would have learned a lot more about intelligence by learning to code and studying AI!
That's why I believe that the best way to answer What is Life? is to learn to code prompt AI and simulate
digital creatures all day. Programming lets you run tens or hundreds of experiments a day, while in a wet lab
it takes on the order of days or even weeks to run a single experiment. Plus, I already learned how to code
during my time researching AI and building a startup.
Q: Are you going to grad school?
A: Probably not. Academia is too risk averse and slow for what I'd like to do. Instead, I'm assembling a group of motivated independent researchers to work on hard problems in a hacker house environment @ calculus.house.
Q: How did you learn all this?
When I saw ALife simulations for the first time, I didn't understand what I was looking at, so I read papers and ran some experiments. The experiments left me confused, so I kept playing around until I could explain what was going on. That's how I wrote my first paper while on hiatus from my startup. After that, I still couldn't explain everything, so I wrote some equations and did some more experiments. That's how I made my second major discovery, etc...
I'm also an AI poweruser. This lets me skip past the esoteric details and jargon that make papers unnecessarily confusing.