Einzelnen Beitrag anzeigen
  #11  
Alt 15.12.17, 15:44
Hawkwind Hawkwind ist offline
Singularität
 
Registriert seit: 22.07.2010
Ort: Rabenstein, Niederösterreich
Beitr?ge: 3.063
Standard AW: Alpha Zero schlägt Stockfish

BTW, this is the opinion of the computer chess pioneer Ed Schröder (author of "Rebel") about the Deep Zero - Stockfish match:

Zitat:
Zitat von Ed Schröder
I am not a conspiracy theorist at all but this Alpha Zero thing needs to be questioned for various reasons and I will serve for some limited time as devil's advocate to warm up the discussion.

#1. We are offered games without depths and scores.
#2. I don't believe it's possible 4 hours of self play can do that, not even 4000 hours, it's insane.
#3. I haven't seen an official statement from Google.
#4. Games can be played by a correspondence player with the help of Houdini tactical (or so) and cherry-picked the best ones, ignoring all the lost games also making use of massive take-backs.

CONSEQUENCES
If this type of learning is possible then (IMO) mankind is in deep trouble, imagine what in theory would be possible. Examples.

#1. Predict the stock market and get rich.

#2. Suppose politicians ask the question:

Dear comp, how do we solve global warming?

Alpha 666

Reduce the population from 7 billion to 500 million and the Earth will restore itself.

------

Shoot !
quoted from the American computer chess forum

http://www.talkchess.com/forum/viewt...=743214#743214
---

hmm, warum schreib ich eigentlich auf Englisch?

Ge?ndert von Hawkwind (15.12.17 um 16:02 Uhr)
Mit Zitat antworten