this post was submitted on 26 Jan 2024
911 points (98.6% liked)

Comic Strips

12529 readers
2742 users here now

Comic Strips is a community for those who love comic stories.

The rules are simple:

Web of links

founded 1 year ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] Maven@lemmy.world 25 points 9 months ago (3 children)

It wasn't until 2015 that the top Go player lost to an Ai while chess lost in 1997. It's wild how big that gap is when you think about how much tech had to improve to make it possible.

[–] AllonzeeLV@lemmy.world 8 points 9 months ago

It's an order of magnitude more complex than Chess, which I am just ok at, so kudos!

[–] Lev_Astov@lemmy.world 3 points 9 months ago (2 children)

And didn't people still find holes in the Go AI's algorithm and proceed to dunk on it afterward?

[–] Pringles@lemm.ee 3 points 9 months ago

Iirc someone figured out that if you didn't make it obvious that you were encircling the AI, it wouldn't take any preventative measures.

[–] sukhmel@programming.dev 1 points 9 months ago

It's a bit complicated to understand what an "algorithm" is in case of a neural network. Besides, I haven't heard of recent human wins over an AI in Go, can you point me to read about it?

[–] idunnololz@lemmy.world 2 points 9 months ago* (last edited 9 months ago)

This is actually more impressive about AI. People used to think Go AI wouldn't be able to beat a human player until like 2050. I certainly thought that when I learned it in like 2010. Back then the strongest AI was like 1 Dan (amateur) at most. (9 Dan is the highest rank and professional 9 Dan which you need to play professional games to get to are much stronger than an amateur 9 Dan which is like 9 Dan from an online website. Also the games rankings go from 30 kyu which is the lowest rank to 1 kyu which is the highest "amateur ranking". After 1 kyu is the Dan ranks ranging from 1 to 9)