More on Chess

It’s no accident that the first big effort at demonstrating Artificial Intelligence was played out in a game of chess.

In the late eighties IBM’s Deep Blue super computer was pitted against world chess champions.

Finally in 1996 the machine defeated Garry Kasparov in game one of a six-game match.

No one truly believes that this effort conformed to Turing’s test for Artificial Intelligence, namely that Deep Blue was able to fool a human into thinking that Deep Blue was a human.

All it showed was that Chess is a very constrained mathematical problem that can be solved quite happily by code. The constraint up until 1996 was simply processing power.

In any case, the Turing test is retarded. A much better test for AI is when a machine has a bank account.

This is a smart-arse way of saying that we humans live in the past, present and future (to varying degrees) whereas machines live in a timeless state, blissfully unaware of the past, the present or the now.

The desire to own money and to have a bank account indicates some sense of living in the future rather than the now, and a memory of the past.

When a machine desires money and a bank account we will know that a machine has some degree of consciousness and also some fear for it’s future safety. These are very human traits.

Back to chess; my assertion would be that a human chess expert is practising Artificial Unintelligence (AU) by forcing their brains into a constrained mathematical domain.

My version of the anti-Turing test is this: Artificial Unintelligence will be achieved when a first human can convince a second human that the first human is actually a machine.

This has already been achieved. I’ve met a few and I don’t envy them.