distributed AI was Re: computronium prime-oxide

Timothy Bates (tbates@bunyip.bhs.mq.edu.au)
Thu, 19 Nov 1998 16:09:54 +1100

Hi all,

Mike Linksvayer said
>Ever since reading Vernor Vinge's "True Names" I've thought a distributed
>artificial intelligence would be one of the neatest things one could do
>with distributed computing (the "Mailman" from the aforementioned story
>was a distributed AI that got its name because it answered questions
>slowly, as if by post).

Great book. The mail man reminds me of a computer implemented out of people passing notes to each other instead of transistors passing voltage. This in turn reminded me of what is called "Turing Functionalism" This is the position that a Turing machine (however implemented) would be conscious if it ran the consciousness program. One envisages the paper tape program ticking over for millenia and then someone cuts it: was the machine half conscious, never conscious?

I asked Daniel Dennett if he was a Turing functionalist: he said yes and no ;-)

I think he said yes, but that interaction is critical, he views intelligence as being embedded in the machines environment as well as in its own hardware. For this reason he doubted that a paper tape Turing machine could be conscious as the environmental context on which consciousness is scaffolded (in his, but not my) view, would be invisible to it: a kind of Nyquist limit for implementation speeds.

best,
tim



"I program because I was programmed to program."

Dept Psychology
Macquarie University
Sydney NSW 2109
Australia
Ph 61 (2) 9850 8623
FAX 61 (2) 9850 8062