Comment by conistonwater | Hacker News Reader

Comment by conistonwater | Hacker News Reader

Parent Full thread

> In its ability to identify and share new insights about one of the most contemplated games of all time, AlphaGo offers a promising sign of the value AI may one day provide, and we're looking forward to playing more games in 2017.

What I find fascinating is how different AlphaGo's impact was from the impact of early chess engines. Once chess engines became decent (not even good—before even Deep Blue), they identified tons of inaccuracies in published chess literature. These were missed tactics, hard-to-see moves, requiring only relatively shallow calculation but a computer's precision. These inaccuracies were found even in classical annotations of well-known chess games, as well as standard books about openings. I believe John Nunn was known for this kind of work.

AlphaGo hasn't achieved nearly the same impact. Have they even tried to identify the same types of inaccuracies in classical Go books? Can you imagine how absolutely cool it would be for a go engine to find errors in Invincible? Maybe they tried, but didn't find any inaccuracies, so now this negative result is sitting in one of their file drawers? I really wish they were more active with this sort of thing.

Once chess engines became decent (not even good—before even Deep Blue), they identified tons of inaccuracies in published chess literature.

Within one year of their invention?

Well, AlphaGo beat Lee Sedol, so I think it wouldn't be fair to compare timelines so directly. With chess engines, the compute power just wasn't there in the beginning, but it's all already there for go. Also, speaking as an optimist, yeah! why not? To be honest, my suspicion is that they aren't trying.

I suspect this is a visibility problem. Go is a more popular game in the East I bet there are a few grad student level problems doing just what you described at Tokyo U and other prestigious Schools far from the West.

Good point, I forgot about DeepZenGo. But it's very low-hanging fruit, though, which is the main reason why I'm surprised they haven't already done this.

Are there any open source or widely available Go engines that are even remotely as strong as AlphaGo (which I think we can safely assume plays beyond the 9p level, i.e., at superhuman strength)? If the only seriously strong engine is completely tied down behind the closed doors of a company that has largely already moved on to other challenges, I think that's the issue right there -- or at least a large part of it.

I think the strongest commercial engines are Zen and CrazyStone. Zen has certainly become pro-strength on fairly modest hardware (it had an exciting BO3 against a Japanese 9-dan, which it lost 1-2), but the strong version isn't available to the public yet, I think.

Interestingly, for the past 3 or 4 days there's been a mystery Go Bot playing on 2 servers calling itself "Master" which is currently about 50-0 (!) against professional opponents, many at the very highest level. Nobody knows its identity yet. It's stronger than the commercial engines, and its presence must be some sort of stunt or test by someone.

http://lifein19x19.com/forum/viewtopic.php?f=10&t=13913

Now confirmed to be AlphaGo.

https://twitter.com/demishassabis/status/816660463282954240

It's 58-0 now. They've been playing 10 games per day since December 29.

Diagrams with moves at https://tieba.baidu.com/p/4922688212?pn=150&

Google translate helps with the comments but they don't add much to the diagrams.

They are playing fast games with 3 time periods of 30 seconds per move. It means that one can exceed the 30 seconds limit and move on to the next period. When the limit is exceeded in the last period the game is over.

Fast games usually increase the chances of the stronger player if the difference is large.

Edit: 59-0 just after I posted this.

I think the next strongest engine is Zen ( http://senseis.xmp.net/?ZenGoProgram ) which plays at amateur 9 dan level. CrazyStone is strong too.

Strongest open source engine is Pachi, which has achieved an amateur 4 dan ranking on the KGS go server. Fuego is almost as good.

Go is different than chess. There is no huge blunders to be found in well-known games. What Alphago can find are creative moves that happen to be very slightly better than "normal" ones.

Since a week, I hear news about a bot named "Master" that may be an evolution (or a concurrent clone) of Alphago.

You can study the games: http://www.dgob.de/yabbse/index.php?topic=6381.msg208264#msg...

There are speculations that the next game between Alphago and best players would be an handicap game.

IMHO Alphago is starting a revolution in the fuseki theory that is bigger than fixing a couple of broken openings.

Deepmind is more focused on general AI (AGI) than the game of Go. The AGI community has always considered beating a Go champion to be a significant milestone towards AGI, and it was expected to be years in the future. Now that Deepmind won that battle, there's little PR benefit - or benefit to AGI progress - to focusing on go.

I don't mean to slight the game of Go, but it is a perfect information game which could make it non-central to AGI research.

Reply on news.ycombinator.com