GoGoD and AI

Higher level discussions, analysis of professional games, etc., go here.
Post Reply
John Fairbairn
Oza
Posts: 3724
Joined: Wed Apr 21, 2010 3:09 am
Has thanked: 20 times
Been thanked: 4672 times

GoGoD and AI

Post by John Fairbairn »

The GoGoD database, already used in connection with AI by researchers at Windows and Facebook, amongst others, has a new contribution to record.

See the following Scientific American article:

https://www.scientificamerican.com/arti ... e-playing/

Shame we no longer have Martin Gardner there to give us his take on it.
pajaro
Lives in gote
Posts: 558
Joined: Tue Mar 23, 2021 6:13 am
GD Posts: 0
KGS: pajaro
IGS: pajaro
Has thanked: 28 times
Been thanked: 142 times

Re: GoGoD and AI

Post by pajaro »

When I was reading the article, I had the feeling that it was written short after the Lee Sedol games. Because it's still talking about how pros are going to include the IA lessons, new moves, and so on. But the change has already happened. IA play is everywhere, and the number of news moves is increasing, which takes me to the interesting point.

What I found interesting in the article is the discussion about new moves. Not many new moves before the IA era, a lot after it. It felt like pros thought "oh, well, this is good enough, don't mess with what works". A bit conservative. Not too different with what happened in the time of Shin Fuseki.

All (most?) players now have the new ideas in mind, as we know, and the human stregth has boosted. Some players are though to be as strong as the original Alpha Go. And younger and future players, well, to them this was of playing is natural, so the level is raised.
Elom0
Lives in sente
Posts: 732
Joined: Sun Feb 20, 2022 9:03 pm
Rank: BGA 3 kyu
GD Posts: 0
KGS: Elom, Windnwater
OGS: Elom, Elom0
Online playing schedule: The OGS data looks pretty so I'll pause for now before I change it.
Has thanked: 1028 times
Been thanked: 32 times

Re: GoGoD and AI

Post by Elom0 »

pajaro wrote:When I was reading the article, I had the feeling that it was written short after the Lee Sedol games. Because it's still talking about how pros are going to include the IA lessons, new moves, and so on. But the change has already happened. IA play is everywhere, and the number of news moves is increasing, which takes me to the interesting point.

What I found interesting in the article is the discussion about new moves. Not many new moves before the IA era, a lot after it. It felt like pros thought "oh, well, this is good enough, don't mess with what works". A bit conservative. Not too different with what happened in the time of Shin Fuseki.

All (most?) players now have the new ideas in mind, as we know, and the human stregth has boosted. Some players are though to be as strong as the original Alpha Go. And younger and future players, well, to them this was of playing is natural, so the level is raised.
Yes, I feel like AlphaGo came nearly at the perfect time, which probably would have been around march 2018, since apparently some pros were just meditating in front of the board for training since they didn't know what else to improve.
Post Reply