Former chess world champion Garry Kasparov likes what he sees of computer that could be used to find cures for diseases
G
arry Kasparov is not only humanity’s greatest ever chess player but its highest-profile victim of artificial intelligence. His loss to IBM’s super computer Deep Blue in 1997 made global headlines and left him feeling bitter and, well, blue. Yet there is a warm glint in his eye when he talks about AlphaZero, the game-changing chess program that took just four hours to teach itself to become the strongest in history.
“For me, as a very sharp and attacking player, it is a pleasure watching AlphaZero play,” he said after playing in a charity tournament for Chess in Schools before the London Chess Classic, which runs until Monday. “We all expect machines to play very solid and slow games but AlphaZero just does the opposite. It is surprising to see a machine playing so aggressively, and it also shows a lot of creativity. It is a real breakthrough – and I believe it could be extremely helpful for many other studies in the field of computer science.”
Aggressive. Creative. Helpful. These are words you might not normally associate with artificial intelligence. Indeed, they sound rather human. But for AlphaZero’s creator, Demis Hassabis, the CEO of Deep Mind, this is just the start of what it might be able to do. As he points out, the next step is to use its capabilities to solve real-world problems – such as protein folding, which is responsible for diseases including Alzheimer’s, Parkinson’s and cystic fibrosis. But he also expects AlphaZero to be able to develop stronger and lighter materials, better medicines and eventually become flexible enough to adapt to new situations.
“Deep Blue could play chess well,” explains Hassabis. “But that is all it could do. It couldn’t play noughts and crosses or Connect 4, or any such simple games. In other words, it couldn’t demonstrate two components which are core to what defines human intelligence – our flexible intelligence and our learning capacity.”
AlphaZero is different. In a long-awaited paper published in the journal Science last week, the authors explain how it learned to conquer Chess, Go and Shogi by playing millions of games against itself via a process of trial and error called reinforcement learning. In more than 1,000 games against Stockfish, a regular winner of the computer chess world championship, it won 155 games with only six defeats, with the rest drawn.
Garry Kasparov and Terry Chapman play chess against the inventor of the Alpha Zero, Deepmind’s Demis Hassabis, and Matthew Sadler at the Google HQ London. Photograph: Lennart Ootes/AlphaZeroSource
No comments:
Post a Comment