Re: AI over the Internet (was Re: making microsingularities)

From: Eliezer S. Yudkowsky (
Date: Sun May 27 2001 - 13:26:41 MDT

James Rogers wrote:
> This does not follow, and it baffles me that so many relatively
> knowledgeable people have simply accepted it. Distributed computing,
> *particularly* over slow high-latency links, is totally unsuitable for a
> very large percentage of the algoriİhm space (and essentially *every*
> plausible AI architecture).

This is a seed AI we're talking about. Suppose that humans had evolved to
write code. Can you imagine them talking about the difficulty of daily
life in a 3D-world? "Oh, yeah, right, like you're going to filter through
trillions of incoming photons in all directions to build up a model of
multidimensional objects that rotate and accelerate and deform.

If it's even theoretically possible for a human being to write distributed
code, then you have to figure that a reasonably mature seed AI that
started out on a tight Beowulf network will be able to adapt verself
pretty easily for the Internet. Or FPGAs. Or nanocomputers. Or a
galactic abacus if that's what's available.

I wouldn't want to develop an AI running on the Internet *initially*.
With this I agree. But I do not agree that network latencies - or
firewalls, for that matter - would present a significant obstacle to a
mature, computer-native general intelligence.

-- -- -- -- --
Eliezer S. Yudkowsky
Research Fellow, Singularity Institute for Artificial Intelligence

This archive was generated by hypermail 2b30 : Mon May 28 2001 - 10:00:08 MDT