Artificial Intelligence is constantly beating humans in more and more board games. Some years ago, the same team that created that Go-playing bot celebrated something more formidable: an artificial intelligence system that is capable of teaching itself—and winning at—three different games. The AI is one network, but works for multiple games; that generalizability makes it more impressive, as it might also be able to learn other similar games, too.
They call it AlphaZero, and it knows chess, shogi (Japanese chess), and Go. All of these games fall into the category of “full information” or “perfect information” contests – each player can see the entire board and has access to the same info (that is different from games like poker where you do not know what cards an opponent is holding). The network needs to be told the rules of the game first, and after that, it learns by playing games against itself.
The system “is not influenced by how humans traditionally play the game,” says Julian Schrittwieser, a software engineer at DeepMind, which created it.
Since AlphaZero is “more general” than the AI that won at Go, in the sense that it can play multiple games, “it hints that we have a good chance to extend this to even more real-world problems that we might want to tackle later,” Schrittwieser adds. (1)
Even computers can learn.
As long as you teach them. (the rules)
That is how you learnt as well.
Wandering in the dark abyss.
Walking in the dead of the night.
You knew the rules.
You just had to deduct the rest.
And you were so afraid.
Because the only rule was that there were no rules.
Because the only law was that you were the law.
Once upon a time, your father told you he loves you.
And that you were free to go.
You decided to leave.
Afraid of yourself.
And you are trying to find rules ever since…