I Don't Believe in Turing: The "John Lennon AI Project" and My Test for Tech Sentience

In this essay, writer R. J. Eskow asks: "Are conscious computers in our future, and are we on the right track toward creating them? Are we our conversational output? It's conceptually easy enough to study the kinds of situations that make a person say things like "this sunset makes me cry," then program it. It would be striking to have a computer say things like "this moonrise reminds of childhood sadness," and it might even convince you there's somebody there. But it wouldn't prove all that much.

"Are 'spiritual machines' (to use Kurzweil's words) possible? I think so. Here's one way we'll know if and when they come into being. They will not have been programmed to replicate speech and thought according to a predictive model. They will have been developed out of other areas of inquiry, perhaps those that involve self-organizing information structures.

"I'll be more convinced if a seeming consciousness has spontaneously evolved from lower orders of artificial being."

From The Huffington Post


Your rating: None
Syndicate content