Wow, congrats to the AlphaGo team! Would love to see a more detailed
analysis later today.
On Mar 9, 2016 9:55 AM, "Marc Landgraf" wrote:
> It was pointed out by Lee Sedol after the game and Kim Myungwan during
> the game, that Q5 should have been better at R4. I would say this was
> the final st
Everybody here probably knows it, but just in case -- there's a
commented broadcast uploaded here:
http://www.youtube.com/watch?v=vFr3K2DORc8
I don't play well enough to understand how good the commentary is, though.
With best regards,
Sergey Nikolenko.
On Wed, Mar 9, 2016 at 10:55 AM, Marc Land
The amount of points AlphaGo was ahead is also meaningless because it
started playing slow moves when it got ahead. After the big fights were
finished it already knew it was going to win easily.
On 2016-03-08 23:55, Marc Landgraf wrote:
It was pointed out by Lee Sedol after the game and Kim My
Many Faces thought alpha go was ahead most of the game. It looked to me like
the turning point was when Alphago cut in the center then gave up the two
cutting stones for gains on both sides (but not so strong…).
Congratulations Aja!
I watched it at Google in Mountain View with about 10
Btw, is there any information on what hardware AlphaGo is running. And
how does it compare to the version used against Fan Hui?
2016-03-09 9:31 GMT+01:00 David Fotland :
> Many Faces thought alpha go was ahead most of the game. It looked to me
> like the turning point was when Alphago cut in the
As Demis said during the press conference, it's roughly the same amount of
hardware.
On Mar 9, 2016 17:56, "Marc Landgraf" wrote:
> Btw, is there any information on what hardware AlphaGo is running. And
> how does it compare to the version used against Fan Hui?
>
> 2016-03-09 9:31 GMT+01:00 David
Congrats to the AlphaGo team also from me!
Von: "David Fotland"
> Many Faces thought alpha go was ahead most of the game.
Similar with CrazyStone. After move 26 CS gave 56 % for AlphaGo
and never went below this value. Soon later it were 60+ %, and
never went lower, too.
Ingo.
_
Wow - didn't expect that. Congratulations to the AlphaGo team!
Ingo wrote:
> Similar with CrazyStone. After move 26 CS gave 56 % for AlphaGo
> and never went below this value. Soon later it were 60+ %, and
> never went lower, too.
Did it show jumps at some of the key moves the human experts thoug
Congratulations to AlphaGo people!
Are there strong humans who have an opinion, on whether this advantage at
move 26
is real ?
In pro games, does it happen often that there is a clear advantage at move
26 ?
--
=
"I will never sign a doc
Will geforce 940m do cuDNN?
From: Hiroshi Yamashita
To: computer-go@computer-go.org
Sent: Wednesday, February 3, 2016 2:55 AM
Subject: Re: [Computer-go] What hardware to use to train the DNN
Hi David,
I use GTS 450 and GTX 980. I use Caffe on ubuntu 14.04.
Caffe's install is diffi
Congrats to Aja and alphago team
2016-03-09 8:43 GMT+01:00 Hiroshi Yamashita :
> AlphaGo won 1st game against Lee Sedol!
>
> Hiroshi Yamashita
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/com
This comment should be very good, it was done by a 9 dan pro, the top rank
in go.
2016-03-09 9:27 GMT+01:00 Sergey Nikolenko :
> Everybody here probably knows it, but just in case -- there's a
> commented broadcast uploaded here:
> http://www.youtube.com/watch?v=vFr3K2DORc8
> I don't play well
Aya also thought AlphaGo was ahead 55% - 65% most of game.
But it is because Aya thought Black N-18 group is 65% alive at move 75.
This Black must live 100%. So I think MCTS tend to think White is ahead.
I'm curious whether AlphaGo understands this group is over 80% alive.
Thanks,
Hiroshi Yamashi
I found Michael Redmond's commentary very good. Helped a weak player
like me understand what was going on, but occasionally went over my
head, which I'm sure others appreciated. He came across as a class act.
His partner, on the other hand, (I forget his name) was more of a
hindrance to the pr
You may also want to check out AGA's commentary by Andrew Jackson and
Myungwan Kim. They don't run out of magnetic stones.
https://www.youtube.com/watch?v=YZPKR7HzM_s
On 09/03/2016 16:13, Richard Lorentz wrote:
> I found Michael Redmond's commentary very good. Helped a weak player
> like me under
Hello,
I had the honour to watch round 1 of the match together with FJ Dickhut
(German 6-dan; player in the codecentric Challenges 2014 and 2015) and
Andreas Fecke (3-dan; Go cartoonist: "Stones").
Here are some photos with comments:
http://althofer.de/lee-sedol-alphago-round-1.html
Ingo.
Hi!
On Wed, Mar 09, 2016 at 04:43:23PM +0900, Hiroshi Yamashita wrote:
> AlphaGo won 1st game against Lee Sedol!
Well, I have to eat my past words - of course, there are still four
games to go, but the first round does not look like a lucky win at all!
Huge congratulations to the AlphaGo t
To me the most surprising fact in this success is that Reinforce can be
applied on such a huge neural net.
That's a huge number of parameters for a reinforcement learning algorithm -
ok it's with a great computational power,
still I would not have guessed that reinforce could be applied on somethin
Hi!
I think the technique of hashing move pairs from the search tree and
reuse them in
the playouts if the context matches, could plausibly be the major
improvement of Alphago that
we witnessed today.
Another thing I noticed is that Alphago does not use any statistics from
the playouts other
In my opinion, the thing that programs do worst is ko.
Lee did not play any kos, except one minor irrelevant
one in the lower left. This game was so simple that
the program could accurately model the whole board.
If Lee wants to win, he needs to start 2 or 3
simultaneous kos.
Michael Wing
Congr
dear Go researchers,
>> > Found a 582 move 3x3 game...
>> Can you give us sgf?
>
> I took the effort of trying to format the 582 game in a more insightful way.
> I ended up with lines of positions that mostly add stones, only starting
> a new line when a capture of more than 1 stone left at most 4
Congratulations, Aja and David! Very remarkable win!
Pasky,
It's too early to conclude any, I think, because no records
of losing games have been published, ie., no weakpoints of
AlphaGo are open. I believe that the essential problems which
come from current MCTS (bottom-up) framework, such
I predicted Sedol would be shocked. I'm still routing for Sedol. From
Scientific American interview...
Schaeffer and Fotland still predict Sedol will win the match. “I think the pro
will win,” Fotland says, “But I think the pro will be shocked at how strong the
program is.”
>
> P.S. Lee Sed
Yes, I think the programs will have similar biases. In this game Sedol had
some groups that were alive, but needed correct responses to stay alive. Even
though the pro's stones won’t die, the playouts sometimes manage to kill them.
This makes the program think it is more ahead than it actual
24 matches
Mail list logo
