Mainstream Go Sightings

[dmf]

In "The Weakest Link" UK show that aired on 18 november, one of the final questions was: "The board game thought to have originated in China and played by two people on a ninteen by nineteen grid has what two letter name?"

[MoxElliot]

The show "Arrested Development" surprisingly had a go scene at the end of an episode. Thanks to netflix having them all available I got a screen shot.

[Jedo]

Recently saw Summer Wars and there's a few scenes of the grandmother playing go.

Awesome movie btw, get it in blu ray if possible.

[kex]

Kroeker, K.L., A New Benchmark for Artificial Intelligence, Communications of the ACM, pp. 13-15. No.8, Vol. 54, Aug 2011.

[cyclops]

Kroeker, K.L., A New Benchmark for Artificial Intelligence, Communications of the ACM, pp. 13-15. No.8, Vol. 54, Aug 2011.

An abstract might help to interest me.

[msgreg]

Kroeker, K.L., A New Benchmark for Artificial Intelligence, Communications of the ACM, pp. 13-15. No.8, Vol. 54, Aug 2011.

An abstract might help to interest me.

"Computers are unable to defeat the world's best Go players, but that may change with the application of a new strategy that promises to revolutionize artificial intelligence."

[Mike Novack]

a)This posting belongs in a different section of the forum (computer go)

b) I think you might have meant a more informative abstract?
I haven't bothered to keep up my membership so can't read the paper. But I rather suspect it isn't really about go playing specifically but rather whether our concept of AI needs to change. Previously there was just one (other) sort of AI that "just worked" as opposed to depending on human knowledge (a "neural net" gets "trained" to be able to perform a task -- the human trainers have to tell it "right" or "wrong" but otherwise it evolves a solution as it "learns" to perform the task). In this case the MCTS algorithm solves the problem of playing a reasonably strong game of go by a different "just works" method.

Note that this may really be more about the "level" at which we look at intelligence. A natural intelligence (like our own) at some level is below consciousness with it all just being connections or lack of connections netween neurons (modification of connections). In other words, at some low level, it too "just works".

c) The isn't going to be about some (new) breakthrough in go playing programs. You perhaps hadn't gone far enough searching out bits about this paper that you can see without being a member/subscriber. But feugo isn't one of the stronger programs at the current state of the art (just the strongest "free" one).

[RBerenguel]

a)This posting belongs in a different section of the forum (computer go)

Hi Mike, I think this depends on what "mainstream" means to the submitter... (but I agree that this suits a post in Computer Go.) If I saw something Go related in the ArXiV for dynamical systems, I may think this as "mainstream Go sighting" (well, I may refrain from posting it here, but would think it is close enough to think about it)

[balistic]

I actually remember when you could play go on steam (valves game library client) against people on your friends list. That was back in 2003 from memory, you could also play chess and checkers.

All of my friends tried it at least once, it reached a mainstream of gamers. Anyone else remember that? There was also chess and checkers and stuff.

[msgreg]

"Bitter Seeds and the Game of Go"

Mainstream "Gweek" podcast 032 Smargo Kifu iPad app review and basic description of the game begins at 40:43 and lasts for about 3 minutes.

http://boingboing.net/2011/12/26/gweek- ... nd-th.html

[Phelan]

http://xkcd.com/1002/

[BaghwanB]

http://xkcd.com/1002/

Good to see go put in context with some other pastimes on this scale...

Bruce "Play-a" Young

[kivi]

That paper (Kroeker) is not a proper scientific paper, rather a popular science magazine style talking about Monte Carlo search applied to Go. Just 3 pages. Nothing much worth mentioning, though probably good to get attention of people that didn't hear about Go and/or Monte-Carlo.

References can be more useful if you are interested:

Arneson, B., Hayward, B., and Henderson, P.
Monte Carlo tree search in hex, IEEE
Transactions on Computational Intelligence
and AI in Games 2, 4, Dec. 2010.

Chaslot, G.M., Winands, M.H., and Herik, H.J.
Parallel Monte-Carlo tree search,
Proceedings of the Sixth International
Conference on Computers and Games,
Beijing, China, Sept. 29–Oct. 1, 2008.

Enzenberger, M., Müller, M.,
Arneson, B., and Segal, R.
Fuego: An open-source framework for
board games and Go engine based on
Monte-Carlo tree search, IEEE Transactions
on Computational Intelligence and AI in
Games 2, 4, Dec. 2010.

Rimmel, A., Teytaud, O., Lee, C.S.,
Yen, S.J., Wang, M.H., and Tsai, S.R.
Current frontiers in computer Go, IEEE
Transactions on Computational Intelligence
and AI in Games 2, 4, Dec. 2010.

Winands, H.M., Bjornsson, Y., and Saito, J.
Monte Carlo tree search in lines of action,
IEEE Transactions on Computational
Intelligence and AI in Games 2, 4, Dec.
2010.

[kex]

That paper (Kroeker) is not a proper scientific paper, rather a popular science magazine style talking about Monte Carlo search applied to Go. Just 3 pages. Nothing much worth mentioning, though probably good to get attention of people that didn't hear about Go and/or Monte-Carlo.

Indeed. It was a mainstream paper: not a scientific paper but a popular science article, and not in a go magazine but on a non-go magazine. I thought this thread was about telling others about mainstream sightings, not about assessing their educational value.

[rubin427]

http://xkcd.com/1002/

Good to see go put in context with some other pastimes on this scale...

Bruce "Play-a" Young

This comic just made front page of USGO.org's news feed.
I don't know about you guys, but I am personally interpreting that as an official endorsement of the game "seven minutes in heaven" by the AGA board.

I'll be sure to check next month's meeting minutes for the official announcement.