Ali Eslami

And artificial intelligence felt really interesting because it has computer science, but it also had philosophy.

It had theory of mind.

It had a bit of neuroscience.

And, um,

I thought that might be a fun thing to spend a couple of years thinking about.

But the analogy I make is, back then, it felt like studying teleportation feels to us today.

It's like, cool idea in theory, but we've never seen anything like teleportation.

Probably it's not going to happen.

But yeah, sure, if you want to go and study that, go ahead.

So that's what AI felt like when I started.

And even when I did my PhD, I genuinely didn't think that I would have a career in this field.

It felt like an intellectual curiosity.

Well, actually during my PhD, there was this kind of sequence of scientific breakthroughs that people at the time called deep learning.

We don't use that word so much anymore, but deep learning, which is basically a statistical technique, came to the fore and people started realizing that this thing has legs.

And very gradually, but at the same time, very rapidly, people who were in the know kind of,

for the first time in, you know, 50, 60 years started thinking, oh, hang on a second.

It feels like we've cracked how we might be able to make progress in this field.

And that was when I kind of switched from, oh, this is kind of a fun thing to do, but realistically, I'm going to go and do startups or go and work in a bank or something to, oh, hang on.

I happen to be within the group of like 300, 400 people in the world who know what this thing is about.

And