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

Google reveals secret test of AI bot to beat top Go players

nature.com

Post reply on HN
Print view View original

151–160 of 214 posts

Re: Google reveals secret test of AI bot to beat top Go players

#151

Post by thomasahle »

conanbatt wrote: post #85
bluetwo wrote: post #21

Earlier quoted context omitted.

I don't know much about Go but I thought a comment about AlphaGo was interesting that it played in a way to marginally beat the player, which was different that most masters played, which is to clearly beat the opponent by as wide a margin as possible. Is this accurate? Does MasterP also use this style? Are there humans that can play this way? (I'm asking you because you seem to know what you are talking about here.)

Thats not true. Go is a game where the safety of the lead is more important than the amount at all times, and its taken into account strategically all the time. What alphago showed however, is that once in the lead it made mistakes: small mistakes that didnt jeopardize the game, but lost the lead. If all paths lead to rome, it doesnt matter which is shorter. Humans however, always think of the best after safety. What…

Modern chess engines take a lot of effort to hide this problem. For example, an engine might know from a table that it can win a king+rook vs king+pawn endgame, and then throws away its second rook to reach such a position.

Humans using the engine don't like this however, and so the authors build in heuristics to make it play more "humanly". Things like using scores for positions, rather than "chance of winning" is also largely for the sake of the users.

Re: Google reveals secret test of AI bot to beat top Go players

#152

Post by thomasahle »

conanbatt wrote: post #127
iopq wrote: post #122

Earlier quoted context omitted.

The highest level players play special variations against each other hoping that the other player doesn't know it. They CLEARLY play against what they think the other player knows, not what they know (since they have studied this variation on purpose to prepare)

As a semi-pro player I can assure you thats not how you play Go.

Interesting. From chess it seems that, for important matches, players will deeply study each other's games and try to get the other player into positions that they may be less used to playing and less comfortable with.

Is this not done in go?

Re: Google reveals secret test of AI bot to beat top Go players

#153

Post by thomasahle »

zeven7 wrote: post #145
dfan wrote: post #7

One thing that isn't made clear in this writeup is that Master plays in a very nonhuman style, as opposed to the version of AlphaGo that beat Lee Sedol, which mostly played like a strong human except for a few surprising moves. My first guess when I saw Master's games was that it was a program like AlphaGo that had its policy network trained from scratch rather than being bootstrapped by being given the goal of imita…

> had its policy network trained from scratch rather than being bootstrapped by being given the goal of imitating the moves of strong humans I just watched a couple of games, and I can't agree. Master's fuseki looks a lot like human fuseki. It plays some original, unusual, confusing, non-human looking moves, yes. It also plays some moves at what would normally be considered the "wrong time" conventionally. But the fu…

Some people were discussion that it seemed like Master just didn't care that much about the opening, since "it knows it'll win anyway".

On chess computers certainly seems to have supported that basically any opening move is playable with tight enough play.

Re: Google reveals secret test of AI bot to beat top Go players

#154

Post by Radim »

ogrisel wrote: post #150
zeven7 wrote: post #145

Earlier quoted context omitted.

> had its policy network trained from scratch rather than being bootstrapped by being given the goal of imitating the moves of strong humans I just watched a couple of games, and I can't agree. Master's fuseki looks a lot like human fuseki. It plays some original, unusual, confusing, non-human looking moves, yes. It also plays some moves at what would normally be considered the "wrong time" conventionally. But the fu…

Maybe it does not use the same opening style when it plays against a version of itself. It would be very interesting to have pro-players comment on published records of alphago self-play. Maybe alphago has discovered a new balance between black and white (that is a new optimal value of the komi) but when playing with the human defined of the value of the komi its optimal style is also different than what it would be…

You mean, like these three AlphaGo self-play games from September? ;)

https://deepmind.com/research/alphago/alphago-games-english/

(analysis by Gu Li and Zhou Ruiyang, two top pros; standard komi)

Re: Google reveals secret test of AI bot to beat top Go players

#155

Post by aaron695 »

conanbatt wrote: post #127
iopq wrote: post #122

Earlier quoted context omitted.

The highest level players play special variations against each other hoping that the other player doesn't know it. They CLEARLY play against what they think the other player knows, not what they know (since they have studied this variation on purpose to prepare)

As a semi-pro player I can assure you thats not how you play Go.

You never study your up coming opponents previous games?

Re: Google reveals secret test of AI bot to beat top Go players

#156

Post by codehotter »

inimino wrote: post #148
dfan wrote: post #7

One thing that isn't made clear in this writeup is that Master plays in a very nonhuman style, as opposed to the version of AlphaGo that beat Lee Sedol, which mostly played like a strong human except for a few surprising moves. My first guess when I saw Master's games was that it was a program like AlphaGo that had its policy network trained from scratch rather than being bootstrapped by being given the goal of imita…

Professional go player Otake Hideo supposedly said he would ask for three stones if playing against God. That was before AlphaGo. My guess has always been that the real gap is much higher, and a perfect player could give 6 or even 9 stones to the world's top players. In perfect play there would be no such thing as joseki. I don't think it would be a game we would even recognize.

As play gets closer to optimal it gets more and more difficult to play more efficiently than your opponent. To play so much more efficiently than your opponent to overcome a handicap of 6 stones strikes me as extremely unlikely at pro level.

At amateur level, say a 2d vs a 2k player would perhaps have a 99.9% winrate. To make it an even game, would take 4 handicap stones. But at pro level, I think it's possible for player A to have a 99.9% winrate against player B, but on two handicap stones, for player B to be the favorite.

Even an engine that is enormously successful against top human players would struggle at high handicaps vs them.

Re: Google reveals secret test of AI bot to beat top Go players

#157

Post by tasuki »

Retric wrote: post #134
jkkramer wrote: post #26

Earlier quoted context omitted.

I would call its style unorthodox rather than nonhuman. It still plays common josekis (standard opening sequences) but often chooses uncommon variations. Its mid game is full of startling moves backed by VERY good reading. There's definitely still discernible strategy that us mortals can learn from. If I recall correctly, the version that beat Lee Sedol was trained on amateur games plus self-play. My guess would be t…

Perfect play is likely inhumanly aggressive on blacks part part and white making zero moves. Compared to that this is very human style of gameplay simply based on a different strategy culture as it where.

I don't see why perfect black play should be any more aggressive than perfect white play. Care to elaborate?

Re: Google reveals secret test of AI bot to beat top Go players

#158

Post by on_and_off »

halflings wrote: post #89
derefr wrote: post #75

Earlier quoted context omitted.

I've heard a similar approach described in military strategy at all levels: rather than looking for a single dominant tactic, you try with each "move" to create so many potentially-viable future positionings at once that your opponent cannot predict you in order to effectively concentrate their effort. It'd be very scary to watch a "sibling" to AlphaGo play a 4X game.

Directly thought of StarCraft while reading this comment: never revealing your strategy and constantly attacking while being defensing and investing in economy to keep your opponent on the defensive. Like another commenter said, StarCraft 2 is DeepMind's next bet.

I am not too familiar with StarCraft but from what I can gather, it relies deeply on micromanagement. It seems to me that any half competent AI strategy wise would destroy any human player just with raw clicks per seconds during fights.

Re: Google reveals secret test of AI bot to beat top Go players

#159

Post by iopq »

Locke1689 wrote: post #133
iopq wrote: post #128

Earlier quoted context omitted.

That's just appeal to authority. Lee Sedol's move against AlphaGo in game 4 could be considered a trap because it actually didn't work, but it was complicated enough to trick AlphaGo.

Appeal to authority is only deductively invalid, not inductively invalid.

But it's not even a good authority, since strong amateurs (semi-pros) have many opponents (amateur tournaments are played in the same day or weekend and have many matches), while top players prepare specifically for one opponent in tournament finals (each match played on a different day).

Re: Google reveals secret test of AI bot to beat top Go players

#160

Post by eutectic »

Diederich wrote: post #55
Florin_Andrei wrote: post #20

Earlier quoted context omitted.

When playing against a much stronger player, it's always very hard to figure out why they play tenuki (make a move somewhere else on the board, apparently uncorrelated to the current fight). When AI gets strong enough (and it seems like it has already), it will just tenuki everyone all the time, while winning. Sounds like exactly what's happening already. It's past the event horizon for human understanding.

> past the event horizon for human understanding I think this phrase is going to pop up more and more frequently.

This thread is literally the only Google search result for this phrase (for me)...
Post reply on HN