Andrej Karpathy

So you're teaching some kind of a special language by which it can save arbitrary information and retrieve it at a later time.

And you're telling it about these special tokens and how to arrange them to use these interfaces.

It's like, hey, you can use a calculator.

Here's how you use it.

Just do 53 plus 41 equals.

And when equals is there, a calculator will actually read out the answer and you don't have to calculate it yourself.

And you just like tell it in English.

This might actually work.

Yeah, I think so.

So Gatto is very much a kitchen sink approach to reinforcement learning in lots of different environments with a single fixed transformer model, right?

I think it's a very sort of early result in that realm.

But I think, yeah, it's along the lines of what I think things will eventually look like.

Yeah, I'm not a super huge fan of, I think, all these interfaces that look very different.

I would want everything to be normalized into the same API.

So, for example, screen pixels versus the same API.

Instead of having different world environments that have very different physics and joint configurations and appearances and whatever, and you're having some kind of special tokens for different games that you can plug, I'd rather just normalize everything to a single interface.

So it looks the same to the neural net, if that makes sense.

So it's all going to be pixel-based pong in the end.

I think so.

I'm not a morning person.