Yeah, I am really on a roll ... first I am misquoted as saying it is
going to be all over for humans in go very soon, and then they say
I wrote GNU Go.
Sigh ...
I guess that now I need to expect requests for the next release of GNU
Go source, or Windows versions, or whatever.
Cheers,
At 01:50 AM 8/10/2008, you wrote:
Yeah, I am really on a roll ... ...
On 9, Aug 2008, at 9:34 PM, terry mcintyre wrote:
I was present; David Doshay said that in ten years, it would be
reasonable to expect computers to play even games with pros.
david d, do you *really* think that they will
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
Yes, for the first time I do think that on the 10 year time scale
computers will play against pros on an even basis. I am not
ready to predict that they will routinely beat the best of the
pros.
They play (or rather it played) at amateur 1-dan now ... that is
what just happened.
Cheers,
David
I'm sure you can find quotes from 'experts' claiming really wild
things on just about any subject.
I think generally that reaching 1-dan in computer-Go was thought to be
attainable with today's hardware but that it would still take
considerable work. I don't think MoGo's recent success suddenly
While the mogo game and result is in the newspaper and keeping all of
us talking, there was another piece of progress in computer Go that
took place at the US Go congress that I think says more about the
state of computer go than the 9-stone handicap win.
The day before the mogo match
First, thank you very much, Don, for giving us a reliable 19x19 server.
Please consider increasing the time a program stays on the list until it
ages off. I guess you drop programs from the ratings page after some time
that depends on the number of games they have played. Since 19x19 games
take
I like the idea of using Bayesian ELO ratings instead. They should
adapt better and faster. It would give better rank confidence than the
current k factor. For example, kartofel would have kept a low
confidence.
Sent from my iPhone
On Aug 10, 2008, at 11:51 AM, David Fotland [EMAIL
* The MCTS technique appears to be extremely scalable. The theoretical
* * papers about it claim that it scales up to perfect play in theory.
** We agree here that this is not true of course.
*
No, I think we disagree this time my friend!
Monte Carlo of course by itself is not
Robert Waite skrev:
* The MCTS technique appears to be extremely scalable. The theoretical
* * papers about it claim that it scales up to perfect play in theory.
** We agree here that this is not true of course.
*
No, I think we disagree this time my friend!
Monte Carlo of
MC/UCT is provably scalable up to perfect play.
Really? Could you send me a link to the paper? I think we must have a
different definition for some word. Perfect play? Are you saying that we
have proven that the 19x19 go board has a perfect path for black? I did not
realize we knew so much about
On Sun, 2008-08-10 at 14:15 -0400, Don Dailey wrote:
I will also modify the server so that losses by anchors don't count.
Woops, what I mean is losses on TIME won't count. They will still
count if the opponent loses but not if the anchor loses.
- Don
David Doshay wrote:
As an aside, the pro in question won the US Open, so comments about
him being a weak pro seem inappropriate. I spoke with him a number
of times, and I firmly believe that he took the match as seriously
as any other public exhibition of his skill that involves handicap
On Sun, 2008-08-10 at 11:37 -0700, Bob Hearn wrote:
Now, my question. Sorry if this has already been beaten to death here.
After the match, one of the MoGo programmers mentioned that doubling
the computation led to a 63% win rate against the baseline version,
and that so far this scaling
It's the tree search part where everything is happening. Eventually,
enough of the tree is explored to find a win or prove a loss.
- Don
On Sun, 2008-08-10 at 20:11 +0100, Raymond Wold wrote:
Dan Andersson wrote:
No more incredible than that Mini-Max and Alpha-Beta will generate
perfect
Hmm.. I dunno.. I think there are a lot of ideas floating around but some
miscommunications.
So the aim is to devise a computer that will beat the strongest human
players of go.
I hear that Monte-Carlo with UCT is proven to be scalable to perfect play.
It seems that this is essentially saying...
On Sun, 2008-08-10 at 15:19 -0400, Robert Waite wrote:
Hmm.. I dunno.. I think there are a lot of ideas floating around but
some miscommunications.
So the aim is to devise a computer that will beat the strongest human
players of go.
I hear that Monte-Carlo with UCT is proven to be
I don't know how you can say that. The empirical evidence is
overwhelming that this is scalable in a practical way but more
importantly it's been PROVEN to be scalable. If you throw the word
practical in there then you are no longer talking the language of
mathematics, theory and proofs so
It is great to see computer players taking another step towards being
first-class citizens of the go-playing world.
cheers
stuart
On Mon, Aug 11, 2008 at 3:37 AM, David Doshay [EMAIL PROTECTED] wrote:
While the mogo game and result is in the newspaper and keeping all of us
talking, there was
one more thing -- you may want to keep anchors from playing
one another. at least, i seem to recall that i saw two anchors
playing one another. it can't (by definition) affect anyone's ratings,
so... probably pointless for them to do so, right?
s.
On Sun, Aug 10, 2008 at 11:27 AM, Don Dailey
On Sun, Aug 10, 2008 at 3:46 PM, Robert Waite [EMAIL PROTECTED]wrote:
Okay.. so where is the paper that correlates the speed at which MCwUCT
approaches perfect play with the ability to play a human? They seem
unrelated as of yet.
The closest I've seen are these two studies Don made:
your calculation is for mogo to beat kim, according to kim and the
mogo team's estimates.
i think that a better thing to measure would be for a computer program
to be able to regularly beat amateurs of any rank without handicap.
i.e. to effectively be at the pro level.
for one thing, this is
Robert,
Do you know what Occam's razor is?
Einstein originally believed that the universe was static. When this
didn't fit his observations he invented the cosmological constant,
which he considered one of his biggest blunders.
If we are going to continue to discuss this, then if you
On Sun, 2008-08-10 at 14:35 -0700, steve uurtamo wrote:
one more thing -- you may want to keep anchors from playing
one another. at least, i seem to recall that i saw two anchors
playing one another. it can't (by definition) affect anyone's ratings,
so... probably pointless for them to do
again, not true.
there are an infinite number of complexity classes beyond
P that do not require infinite space or infinite time.
exptime would just take exponential time instead of polynomial
time, and pspace would just be able to reuse its available
polynomial space (and thus use at worst
david, is mfgo-12-0805-2c really over 400 ELO better
than mfgo-11, as cgos seems to suggest? or is mfgo11
still rising up into place?
thanks,
s.
On Sun, Aug 10, 2008 at 8:51 AM, David Fotland [EMAIL PROTECTED] wrote:
First, thank you very much, Don, for giving us a reliable 19x19 server.
Well... I think I have hunches just as you do. And I think we both express
our hunches on here.
Diminishing returns is not really my theory.. I am just looking at
alternative ways of viewing the datapoints. Let's say you have two computers
and both of them focus only on solving local situations.
there are no problems that would take infinite time or infinite
space. there are problems that cannot be solved no matter
how much space or time you give a computer, but that's a
different matter altogether, and go isn't one of those problems.
How do you know what class go belongs in?
How do you know what [complexity] class go belongs in?
Hi Robert,
If these topics interest you, you probably want to start by reading the
papers [1] about the complexity of go. Then if you still disagree take
up a specific point with the paper authors.
Both minimax and UCT solve go simply
(Sorry if this is a duplicate; the first posting didn't show up.)
To clarify (or maybe not) the status of the computational complexity
of NxN go:
Go with Japanese rules is EXPTIME-complete (Robson, 1983).
Go with superko is only known to be PSPACE-hard, and EXPSPACE-easy
(Robson, 1984).
Here are the current ratings using bayeselo of program on the 19x19
server. I have a script in place so that I can update this at will and
I may run this every few hours or so, probably starting tomorrow.
http://cgos.boardspace.net/19x19/bayes_19x19.html
- Don
On Sun, 2008-08-10 at
Thanks. This is more like what I would expect. About 80 elo points between
mfgo 1 cpu and 2 cpu (like other programs), and many faces 11 a little
higher rated.
Does anyone know whose program is rz-74? I'm trying to catch mogo,
crazystone, and leela by September, and I'm curious if there will
You will notice that no program has a great deal of confidence. Even
Gnugo the anchor is plus or minus 50 ELO.
I combined ALL Gnugo anchors into 1 entity for rating purposes. It
appears on the charts as Gnugo-3.7.10-a0 and does not have a
cross-table entry.
I also did the one time removal
On Aug 10, 2008, at 1:46 PM, Robert Waite wrote:
Exhaustive search is scalable in that I could give it all the
memory and time it wanted. And it would approach a finite amount of
memory and a finite amount of time.
Yes, but exhausitve search does not improve your player by 63% (eg.)
for a
34 matches
Mail list logo