Former World Chess Champion Garry Kasparov recently talked to neuroscientist and philosopher Sam Harris, on topics ranging from Russian-American politics, to the future of artificial intelligence:
As the domain expertise of the participants definitely lies in AI rather than politics, I’ll confine my comments to that aspect of their dialog! (You can find this part of the conversation at around 1:16 in the podcast.)
Kasparov emphasizes that one of the major advantages of chess-playing computers is simply their relentless accuracy. He implies that an expert chess player paired with a low-powered computer ‘assistant’ may be more than a match for top-class computer-only performance: the assistant could eliminate low-grade human inaccuracies. This is a very different picture from the ‘superhuman intelligence’ version of AI. Note, however, that in other contexts, AI-play has been praised for its innovative character:
I wrote in a bit more depth about intelligence and AI a few months ago:
As that article discusses, I think there is a kind of creativity and intelligence that cannot, even in principle, be reproduced by conventional computers. This view complements Kasparov’s: the marriage of human intelligence and computer cognition may be the way of the future.
Don’t have an account? Sign up