Ali Eslami

what its potential is.

And so, yeah, towards the end of my PhD, it was pretty clear that this is going to be big.

And so I decided to stay.

I did a postdoc and then joined school.

Yeah, totally.

I mean, the topic of my PhD was computer vision specifically.

So it's how do we get computers to see the meaning in images, right?

That's one way of defining it.

So when I look at the scene, I can understand that I'm in a circular room and there's people.

I understand the context.

Even 10 years ago, if you were to show this photo to a computer,

the most I could say is that this pixel is red, this pixel is black, but otherwise it was basically a mystery to the computer itself.

And yeah, as I mentioned to you earlier, as an expert in the field, I genuinely didn't think that we would make meaningful progress on this problem in my lifetime.

It felt so out of reach because we had no clue how to proceed.

And

Even to me, if you were to show me what we can do today, 10 years ago, 99% would not have believed you.

I would have said this is impossible.

Right.

Well, one thing that's really changed is 10 years ago when we were doing research, it really felt like it was separate from the real world.

So when friends or family would ask me, what are you working on?