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

Stuart Russell

๐Ÿ‘ค Speaker
See mentions of this person in podcasts
1598 total appearances
Voice ID

Voice Profile Active

This person's voice can be automatically recognized across podcast episodes using AI voice matching.

Voice samples: 1
Confidence: Medium

Appearances Over Time

Podcast Appearances

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And then he dies of misery and starvation.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And this is, you know, it's a warning.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

It's a failure mode that pretty much every culture in history has had some story along the same lines.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

You know, there's the genie that gives you three wishes.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And, you know, third wish is always, you know, please undo the first two wishes because I messed up.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And, you know, when Arthur Samuel wrote his checker playing program, which learned to play checkers considerably better than Arthur Samuel could play and actually reached a pretty decent standard.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

Norbert Wiener, who was one of the major mathematicians of the 20th century, sort of the father of modern automation control systems,

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

He saw this and he basically extrapolated, as Turing did, and said, okay, this is how we could lose control.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And specifically that we have to be certain that the purpose we put into the machine is the purpose which we really desire.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And the problem is, we can't do that.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

The line is gray between the two.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

So theoretically it's possible, but in practice it's extremely unlikely that we could specify correctly in advance the full range of concerns of humanity.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

Well, we learn, yeah, I mean, as we grow up, we learn about the values that matter, how things should go, what is reasonable to pursue and what isn't reasonable to pursue.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

Yeah, so I think that what we need to do is to get away from this idea that you build an optimizing machine and then you put the objective into it.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

Because if it's possible that you might put in a wrong objective, and we already know this is possible because it's happened lots of times, that means that the machine should never take an objective that's given as gospel truth.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

Because once it takes the objective as gospel truth, then it believes that whatever actions it's taking in pursuit of that objective are the correct things to do.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

So you could be jumping up and down and saying, no, no, no, no, you're going to destroy the world.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

But the machine knows what the true objective is and is pursuing it.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And tough luck to you.

Lex Fridman Podcast
Stuart Russell: Long-Term Future of AI

And this is not restricted to AI, right?