Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Richard Sutton and Andrew Barto Win 2024 Turing Award

awards.acm.org

Post reply on HN
Print view View original

21–30 of 119 posts

Re: Richard Sutton and Andrew Barto Win 2024 Turing Award

#21

Post by amelius »

khaledh wrote: post #4
ofirpress wrote: post #3

Good time to re-read The Bitter Lesson: https://www.cs.utexas.edu/~eunsol/courses/data/bitter_lesson...

Indeed a bitter lesson. I once enjoyed encoding human knowledge into a computer because it gives me understanding of what's going on. Now everything is becoming a big black box that is hard to reason about. /sigh/ Also, Moore's law has become a self-fulfilling prophecy. Now more than ever, AI is putting a lot of demand on computational power, to the point which drives chip makers to create specialized hardware for it…

Well, take compiler optimization for example. You can allow your AI to use correctness-preserving transformations only. This will give you correct output no matter how weird the AI behaves.

The downside is that you will sometimes not get the optimizations that you want. But, this is sort of already the case, even with human made optimization algorithms.

Re: Richard Sutton and Andrew Barto Win 2024 Turing Award

#22

Post by dfan »

DavidPiper wrote: post #19
ofirpress wrote: post #3

Good time to re-read The Bitter Lesson: https://www.cs.utexas.edu/~eunsol/courses/data/bitter_lesson...

This describes Go AIs as a brute force strategy with no heuristics, which is false as far as I know. Go AIs don't search the entire sample space, they search based on their training data of previous human games.

The paragraph on Go AI looked accurate to me. Go AI research spent decades trying to incorporate human-written rules about tactics and strategy. None of that is used any more, although human knowledge is leveraged a bit in the strongest programs when choosing useful features to feed into the neural nets. (Strong) Go AIs are not trained on human games anymore. Indeed they don't search the entire sample space when they perform MCTS, but I don't see Sutton claiming that they do.

Re: Richard Sutton and Andrew Barto Win 2024 Turing Award

#23

Post by zackkatz »

Very cool to see this! It turns out my wife and I bought Andy Barto’s (and his wife’s) house.

During the process, there was a bidding war. They said “make your prime offer” so, knowing he was a mathematician, we made an offer that was a prime number :-)

So neat to see him be recognized for his work.

Re: Richard Sutton and Andrew Barto Win 2024 Turing Award

#26

Post by signa11 »

DavidPiper wrote: post #19
ofirpress wrote: post #3

Good time to re-read The Bitter Lesson: https://www.cs.utexas.edu/~eunsol/courses/data/bitter_lesson...

This describes Go AIs as a brute force strategy with no heuristics, which is false as far as I know. Go AIs don't search the entire sample space, they search based on their training data of previous human games.

> ... This describes Go AIs as a brute force strategy with no heuristics ...

no, not really, from the paper

>> Also important was the use of learning by self play to learn a value function (as it was in many other games and even in chess, although learning did not play a big role in the 1997 program that first beat a world champion). Learning by self play, and learning in general, is like search in that it enables massive computation to be brought to bear.

important notion here is, imho "learning by self play". required heuristics emerge out of that. they are not programmed in.

Re: Richard Sutton and Andrew Barto Win 2024 Turing Award

#28

Post by optimalsolver »

anonzzzies wrote: post #5
khaledh wrote: post #4

Earlier quoted context omitted.

Indeed a bitter lesson. I once enjoyed encoding human knowledge into a computer because it gives me understanding of what's going on. Now everything is becoming a big black box that is hard to reason about. /sigh/ Also, Moore's law has become a self-fulfilling prophecy. Now more than ever, AI is putting a lot of demand on computational power, to the point which drives chip makers to create specialized hardware for it…

I am still hoping AI progress will get to the point where the AI can eventually create AI's that are built up out of robust and provable logic which can be read and audited. Until that time, I wouldn't trust it for risky stuff. Unfortunately, it's not my choice and within a scarily short timespan, black boxes will make painfully wrong decisions about vital things that will ruin lives.

>AI can eventually create AI's that are built up out of robust and provable logic

That's the approach behind Max Tegmark and Steven Omohundro's "Provably Safe AGI":

https://arxiv.org/abs/2309.01933

https://www.youtube.com/watch?v=YhMwkk6uOK8

However, there are issues. How do you even begin to formalize concepts like human well-being?

Re: Richard Sutton and Andrew Barto Win 2024 Turing Award

#29

Post by HarHarVeryFunny »

DavidPiper wrote: post #19
ofirpress wrote: post #3

Good time to re-read The Bitter Lesson: https://www.cs.utexas.edu/~eunsol/courses/data/bitter_lesson...

This describes Go AIs as a brute force strategy with no heuristics, which is false as far as I know. Go AIs don't search the entire sample space, they search based on their training data of previous human games.

First there was AlphaGo, which had learnt from human games, then further improved from self-play, then there was AlphaGo Zero which taught itself from scratch just by self-play, not using any human data at all.

Game programs like AlphaGo and AlphaZero (chess) are all brute force at core - using MCTS (Monte Carlo Tree Search) to project all potential branching game continuations many moves ahead. Where the intelligence/heuristics comes to play is in pruning away unpromising branches from this expanding tree to keep the search space under control; this is done by using a board evaluation function to assess the strength of a given considered board position and assess if it is worth continuing to evaluate that potential line of play.

In DeepBlue (old IBM "chess computer" that beat Kasparov) the board evalation function was hand written using human chess expertise. In modern neural-net based engines such as AlphaGo and AlphaZero, the board evaluation function is learnt - either from human games and/or from self-play, learning what positions lead to winning outcomes.

So, not just brute force, but that (MCTS) is still the core of the algorithm.

Post reply on HN