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

How much did AlphaGo Zero cost? (2018)

yuzeh.com

Post reply on HN
Print view View original

11–20 of 179 posts

Re: How much did AlphaGo Zero cost? (2018)

#11

Post by zucker42 »

I wonder if the code and network weights will ever see the light of day. I wonder what the eventual value proposition of working on this sort of stuff is. I suppose they are just going to try to apply the algorithms to better things.

I've been interested in the application of AlphaZero to chess. It's sad that this many resources were devoted to something which we can't even use to play chess as of now. Leela (the open source reengineer) is really strong, but the crushing results presented in the AlphaZero paper never materialized. And this article just shows how hard they are to replicate.

Re: How much did AlphaGo Zero cost? (2018)

#12

Post by NVHacker »

bsaul wrote: post #6
ipsum2 wrote: post #2

Alpha Go Zero*, which was trained from scratch, without human games. I've also heard rumors that AlphaStar ( https://deepmind.com/blog/article/alphastar-mastering-real-t... ) was essentially put on hold because it was too expensive to improve/train. The bot wasn't able to beat StarCraft champions and _only_ got to a grandmaster level.

Interesting ! i didn't understand why they stopped the alphastar project. They pretended they reached their goal, but clearly haven't.

The goal was never to beat all pros. https://deepmind.com/blog/article/AlphaStar-Grandmaster-leve...

Re: How much did AlphaGo Zero cost? (2018)

#13

Post by 29athrowaway »

irjustin wrote: post #4

Alpha Go Zero inspired the development of an open source version, Leela Go Zero which Leela Chess Zero is forked from by the same guy who made Stock Fish. Lots of people contribute what I imagine are amounts of CPU Power/money to the Leela Chess Zero project[1]. Would love to see Alpha Chess vs Leela Chess. [1] https://training.lczero.org/ [edit] I've caused terrible confusion by melding Leela Go and Leela Chess when…

I think KataGo is stronger than Leela Zero.

https://github.com/lightvector/KataGo

Re: How much did AlphaGo Zero cost? (2018)

#15

Post by FartyMcFarter »

hjnilsson wrote: post #7

Another way of thinking about how efficient the brain is: By the article’s numbers, about 5.5 million TPU hours were required to train the machine to play as well as a Go champion. A Go champion might have trained for 8 hours a day, for 15 years (age 5 to 20). That is about 40 000 hours. In other words, machines required 137 times longer to learn the game, and at twice the power consumption! There is still a lot of r…

Go champions don't learn from zero. They learn from teachers, books, and playing against each other. This knowledge is built over hundreds, or thousands of years.

Re: How much did AlphaGo Zero cost? (2018)

#16

Post by phreeza »

hjnilsson wrote: post #7

Another way of thinking about how efficient the brain is: By the article’s numbers, about 5.5 million TPU hours were required to train the machine to play as well as a Go champion. A Go champion might have trained for 8 hours a day, for 15 years (age 5 to 20). That is about 40 000 hours. In other words, machines required 137 times longer to learn the game, and at twice the power consumption! There is still a lot of r…

This comparison is not entirely fair because the human brain also benefits from priors baked in over the entire course of evolution.

Re: How much did AlphaGo Zero cost? (2018)

#18

Post by vbezhenar »

hjnilsson wrote: post #7

Another way of thinking about how efficient the brain is: By the article’s numbers, about 5.5 million TPU hours were required to train the machine to play as well as a Go champion. A Go champion might have trained for 8 hours a day, for 15 years (age 5 to 20). That is about 40 000 hours. In other words, machines required 137 times longer to learn the game, and at twice the power consumption! There is still a lot of r…

Human does not learn Go from the scratch on himself. He's using teachers, books which present compressed knowledge which was crystallized from many millions of human hours.

If you would ask someone to learn Go, but only present him rules of the game, he'll likely be weak player (although probably with some original strategies).

Re: How much did AlphaGo Zero cost? (2018)

#19

Post by Drakim »

phreeza wrote: post #16
hjnilsson wrote: post #7

Another way of thinking about how efficient the brain is: By the article’s numbers, about 5.5 million TPU hours were required to train the machine to play as well as a Go champion. A Go champion might have trained for 8 hours a day, for 15 years (age 5 to 20). That is about 40 000 hours. In other words, machines required 137 times longer to learn the game, and at twice the power consumption! There is still a lot of r…

This comparison is not entirely fair because the human brain also benefits from priors baked in over the entire course of evolution.

So did the machine, albeit indirectly.

Re: How much did AlphaGo Zero cost? (2018)

#20

Post by ben_w »

phreeza wrote: post #16
hjnilsson wrote: post #7

Another way of thinking about how efficient the brain is: By the article’s numbers, about 5.5 million TPU hours were required to train the machine to play as well as a Go champion. A Go champion might have trained for 8 hours a day, for 15 years (age 5 to 20). That is about 40 000 hours. In other words, machines required 137 times longer to learn the game, and at twice the power consumption! There is still a lot of r…

This comparison is not entirely fair because the human brain also benefits from priors baked in over the entire course of evolution.

OTOH, I expect that avoiding human evolutionary priors is necessary for superhuman performance.
Post reply on HN