From American Go Association newsletter,
COMPUTER BEATS PRO AT U.S. GO CONGRESS: In a historic achievement, the MoGo computer program defeated Myungwan Kim 8P (l) Thursday afternoon by 1.5 points in a 9-stone game billed as “Humanity’s Last Stand?” “It played really well,” said Kim, who estimated MoGo’s current strength at “two or maybe three dan,” though he noted that the program – which used 800 processors, at 4.7 Ghz, 15 Teraflops on a borrowed European supercomputer – “made some 5-dan moves,” …
Complete writing is here.
The game of chess has long been conquered and mastered by machines, and now they are eyeing on even greater problem domain, the game of Go ! But dont worry, the computers will not conquer the earth and enslave us in the near future just yet. That’s because of the nature of AI discipline, which focuses more on narrowly defined problem space, with limited rules. At least I never heard of any successful general-purpose AI machine, or even existence of any such project.
By the way, it was a 9-stone handicap game, meaning that the computer was allowed to move 9 times before the human made his first. The positions of these handicap stones are usually predetermined to spread evenly over the board. So yea, humans are still that much better than machines in Go.
Kim estimated that MoGo’s strength is about 2 or 3 dan (amateur) which I further estimate to be about the level of average amateur go players who play and study the game consistently for 1-3 years. And this same performance is achieved by hardware of 15 Teraflops, or 15,000,000,000,000 floating point operation per second. I am very curious as to what techniques of AI are involved in a computer go program. In fact I’ve already downloaded open source gnugo project, although being slightly demotivated by its size. Haizzz.. so huge, where to find time to study :( .
By the way, from Artificial Intelligence : A Guide to Intelligent System by Michael Negnevitsky, I read that
- Human brain memory capacity is 10^18 bits.
- Human brain processing power is 10^15 bits per second.
Well, maybe that explain subconscious memory, dreams, and so on ? After all, we only remember 0.1% of what we know on every second, with others somewhere in the background. (”Remember”, for me, means “being processed”).
Anyway, that above book on Artificial Intelligence is superb, really excellent with good review from many other readers. I have only read the first chapter, and did not fall asleep despite having no reason whatsoever to learn AI. OK, maybe I am not qualified to review that book since I’ve only read one chapter out of nine. But a technical book that keeps you wanting to read even after 24 pages, isn’t it good ?



Yes – I never found any technical book that keeps me awake…so its good for you!
Wow, thanks for the interesting article. I was put off by Chess due to it being conquered by machines, especially after Kasparov lost to Deep Blue.
Good to know that humans aren’t obsolete, YET!
similar to me. machines being so good discouraged me from playing chess also.
but any successes by machines are actually humans achievement also. it shows that we are getting better at understanding “thinking process” and maybe its nature.
wonder if go will stay interesting to me, even after computers beat us.
Well, with the Nehalem Intel chips coming out, probably only a matter of time before AI overtakes us in Go and other areas, though hopefully not in my lifetime…
A bit scary, if one thinks about it.
I’m sure Go will remain interesting since it’s still a long way to go before AI reaches such an advanced level…
when will AI ever be self-aware?
im not sure if it will be long before they take over the throne on Go board, avatar.
history have shown never to underestimate the growth of technology.
self-aware?
the aim of AI is to pass Turing test, “can the computers fool human that they are human too, by acting, thinking, lying , pretending like them .. ?”
the ‘nature’ of AI whether, ‘it truly thinks’, ‘it truly feels’ or so on is no longer the concern, except to philosophers maybe.
Dear Waq,
I certainly hope not… because that’s one of the last bastion of intellectual games… not yet conquered by AI.
Well, AI can either turn out to be a good thing or otherwise…it’s either an Asimov-ian future or a Matrix/Terminator one.
I hope it’s the former but looking at the trend, it just might be the latter
[...] Machine Invasion. Human Outsmarted? « Waq’s Words and World [...]