Menu
Sign In Search Podcasts Libraries Charts People & Topics Add Podcast API Blog Pricing

Andrej Karpathy

πŸ‘€ Speaker
3433 total appearances

Appearances Over Time

Podcast Appearances

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

You might not even be able to pick us up in that simulation.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

And so how do you prove that you exist, that you're intelligent and that you're part of the universe?

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Yeah, the puzzle is basically alerting the creator that we exist.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Or maybe the puzzle is just to break out of the system and just stick it to the creator in some way.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Basically, if you're playing a video game, you can somehow find an exploit and find a way to execute on the host machine in the arbitrary code.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

For example, I believe someone got a game of Mario to play Pong just by exploiting it and then...

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

basically writing code and being able to execute arbitrary code in the game.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

And so maybe that's the puzzle, is that we should find a way to exploit it.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

So I think some of these synthetic AIs will eventually find the universe to be some kind of a puzzle and then solve it in some way.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

And that's kind of like the end game somehow.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Yes.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Yeah, I think so.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Is that what physics is essentially?

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

I think it's possible that physics has exploits and we should be trying to find them.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

Arranging some kind of a crazy quantum mechanical system that somehow gives you buffer overflow, somehow gives you a rounding error in the floating point.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

We'll find some way to extract infinite energy.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

For example, when you train reinforcement learning agents in physical simulations and you ask them to, say, run

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

quickly on the flat ground, they'll end up doing all kinds of weird things in part of that optimization.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

They'll get on their back leg and they'll slide across the floor.

Lex Fridman Podcast
#333 – Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI

And it's because the optimization, the enforcement learning optimization on that agent has figured out a way to extract infinite energy from the friction forces and basically their poor implementation.