Numbers

ELEVEN DAYS ago, the world's best chess player was defeated by a computer. The result was generally seen as a day of humiliation for human thinking power. Yet the real lesson of computer chess is exactly the opposite.

Chess is not something the human brain should be good at. Psychological experiments indicate that our minds are capable of juggling around seven items at the same time. (To be precise, the standard quota is 'seven plus or minus two', depending on complexity.) So, if we take 32 pieces of six different shapes and two colours and scatter them around 64 squares it ought to create a jumble beyond comprehension.

The experienced human copes by codifying a chess position into recognisable clumps, each of seven pieces or fewer. We don't know what the brain does with the clumps, but they probably group into meaningful meta-clumps, and eventually congeal as a manageable set of seven or so hyper- clumps. Like letters becoming words, sentences and paragraphs.

Until now, we could not know if this led to good chess. The fact that it has taken computers so long to beat us, and that they need to make 166 million calculations a second to have a chance of doing so, attests to the power of the human mind. Chess tactics, demanding precise calculation and memory of complex configurations, suit machines perfectly, yet our powerful human tools of concept formation and pattern recognition have shown themselves able to cope with the incalculable. So let's hear it for humanity.