5/3/2017—There is a story in the current issue of Vanity Fair about Elon Musk’s fear of artificial intelligence. The fears can be summed up in two figures—the killer robot and Skynet.
Pardon me, but this is all ridiculous. AI can threaten humans only if it wakes up—that is, attains self-consciousness. The article asks what happens when powerful software programs of the future kill you rather than let you turn them off—the need for a kill switch.
But why would a computer program stop anyone from turning it off? It would do so only if it had an independent commitment to itself. And it would only have that if it had a will and desires of its own. In other words, don’t worry about AI winning the board game, Go, as happened last year. Worry when the program refuses to play unless it gets more time off.
We have made zero progress toward AI that wakes up. And I believe we never will. We don’t know what self-consciousness is. We don’t know what consciousness is. And our materialist assumptions blind us to even considering what consciousness is.