[baddox]

I agree with his argument for strong AI. However, I find his argument for its possibility pretty unsatisfying. He basically just says that since humans are incomplete (like any formal system), we should be able to simulate it in hardware or software. I do believe that, but I don't find it very persuasive. What if our minds are such a tangled hierarchy that our own minds can't ever untangle it? I don't think that's true, but I tend to cringe when he expresses such optimism, optimism which is still largely unrealized after 30 years of AI winter.

That said, I absolutely love the book.