[Computer-go] Notes from the Asilomar Conference on Beneficial AI
sligocki at gmail.com
Thu Feb 9 18:28:37 PST 2017
I remember hearing widespread surprise at many of AlphaGo's moves vs. Lee
* Game 1 - Moves 10&12 - Attach and pull back after pincer. Which is widely
taught as bad play.
* Game 2 - Move 37 - 5th line shoulder hit. Likewise students are taught to
almost never play 5th line shoulder hit.
and to a lesser extent:
* Game 2 - Move 15 - Peeping so early.
Those are the examples I remember. Don't remember any references to style
pros didn't like (except maybe that AlphaGo likes the middle more than most
On Thu, Feb 9, 2017 at 7:32 PM, Richard J Lorentz <lorentz at csun.edu> wrote:
> Thanks for the interesting link. Indeed, some good reading there.
> One quote that I've seen various versions of a number of times now: " More
> interesting for the rest of us, AlphaGo is playing moves and styles that
> all human masters had dismissed as stupid centuries ago."
> Can any one point me to concrete examples of such moves, styles, and/or
> games? With some commentary?!
> On 02/09/2017 02:42 PM, Freeman Ng wrote:
> My favorite blogger's account of this conference that the Deep Mind team
> also participated in.
> Note: he's neither a Go player nor a computer programmer, which limits his
> AlphaGo related reporting, but I thought this might be interesting to the
> list nonetheless.
> Computer-go mailing list
> Computer-go at computer-go.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Computer-go