3 ms·
Yes, machines can out-perform humans; that's why we build them. A car with a rock on the peddle will out-run an olympic runner; but it isn't competing in a rac
by mjburgess 4y ago
Yes, machines can out-perform humans; that's why we build them.
A car with a rock on the peddle will out-run an olympic runner; but it isn't competing in a race.
To play, to compete, to do any of these things is to intend. One has to have the relevant mental states to "play" at anything.
A computer which scans a move tree and emits moves does not intend to do anything; it's just a car with a rock on the peddle. Of course it "wins", in as much as a river might, or a gust of wind.
We are interested in games because we play them; not because we "arrive at a win condition". Winning is only significant because it's play, not because of the board state.
- calf 4y agoFurthermore anyone who studied branch and bound in algorithms class ought to understand that a chess engine wins games but does not understand the game and cannot explain the game. That's partly what Chomsky meant by explainability; the fact that AlphaGo/MuZero can beat humans at chess says nothing about rational intelligence any more than a conventional chess engine can beat humans says anything about its capacities. It says even less because a neural network obscures its own learned algorithm, by design.