From: Mike Linksvayer (ml@gondwanaland.com)
Date: Tue Apr 09 2002 - 21:44:54 MDT
On Mon, 2002-04-08 at 03:18, Samantha Atkins wrote:
> SIGH. If it things 1000 times slower than humans then it is not
> a very credible threat. It can't move fast enough to escape
> detection and extermination. Vinge did not exactly deal with
> this possibility. Also, if it runs too slowly it is
> questionable whether certain forms of intelligence will gel at all.
We have zero experience detecting or exterminating rouge AIs, nor is
anyone watching, or even theorizing about what should be watched for
AFAIK. Speed is only one means of avoiding detection and extermination.
A very slow distributed AI could also have a goal to get faster --
insinuate itself into low-latency/high-cpu networks, move work requiring
such there. A few datacenters stealthily used, the thing starts running
closer to human timescales and can interact more effectively with "our"
world.
Apart from its relevance to the potential for an [ab]useful distributed
AI, I think the roles speed and intelligence play in terms of getting
"intelligent work" done is an interesting question. An uploaded rat
running at many times the speed of a physical rat presumably won't get
any more intelligent work done. How much work could a IQ300-equivalent
AI that runs at 1/1000 human speed get done? Perhaps not much, but it
might be extremely interesting work.
Mike Linksvayer
http://gondwanaland.com/ml/
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:13:21 MST