In his famous article in 1950 Alan Turing proclaimed: "I believe that by the end of the century ... one will be able to speak of machines thinkingwithout expecting to be contradicted." Clearly the famous Turing got it wrong (Tom Melzer, G2, 18 June). And even if the Turing test had been passed, he would still have been wrong. The Turing test is the crudest of behavourism. In my lifetime Turing has gone from obscurity to awe-inspring icon, but this seems to be exaggerated. He certainly did good work on the notion of computability, though that was based on something much more fundamental called Godel's theorem. It doesn't seem to be true that he was the inventor of the computer, though again he did good work on early versions of programming computers. I can't comment on his contribution to breaking the German codes, since none of the accounts I've seen explain exactly how they did it. But I understand that it took them a long time, and that when they'd done it, Doenitz became suspicious and changed all the codes. It is perhaps time for a more balanced view of Turing.
Roger Schafir
London