The University of California has been working on programs that teach computers how to play poker.
While I am very much in favor of taking all of our most social activities and engineering them so that they can be enjoyed without interacting with any other humans, I think that this is a terrible idea. Why? Robot doom, of course. Do we really want our computers learning how to tell when humans are bluffing? ("There's no one else hiding in this house. I swear, you don't need to hunt down my wife and children, you already killed them!") Worse yet, do we want them learning how they can bluff? ("I LIKE YOU HUMAN, I PROMISE I'LL NEVER HARM YOU.")
In the end, it will probably just be the computers distracting us with poker while they send the hunter/killer squads after us.
Monday, August 18, 2008
The Computer Always Wins
Labels:
Robot Doom,
Science
Subscribe to:
Post Comments (Atom)
The header image is adapted from a photo taken by Bill McChesney and used under a creative commons license.
No comments:
Post a Comment