Joe Lonsdale

I think there's a lot of different levels you can define of what...

like, what constitutes the computers being way, way ahead of us.

And I think there's, like, artificial systemic intelligence.

There's general intelligence is lower.

And, like, one of the questions is, like, at what point, like, do people not even help the AI improve itself, which is kind of a scary question, right?

So if you want to compare, if chess is the right analogy, if you go back to chess, it used to be that, like, people were better than the machines, but the machine could help you with tactics.

So there's two types of chess.

There's the tactics, which is like the next three, five, 10 moves.

The computers are really good at that, even from 1960s, because you can see the next 10 moves.

But people will be much better at knowing what's happening way farther ahead, right?

Position, place.

So there's things the human mind's better at, and there's things the computer is better at.

And as the computer got better at chess, it got to the point where the best computer was better than the best person.

But for a long time, if you combined a computer and a person, it still made it a lot better, because there were still some things that people were better at.

So the people were so much better at position play that there would be a chimera.

Combined, it was better.

And then it got to a point where the computer is so freaking good, it doesn't even get any better if you add a person.

It doesn't need to help anymore.

It's just so far ahead.

Damn.