Re: Computers that improve themselves

From: Jimmy Wales (jwales@aristotle.bomis.com)
Date: Sun Apr 22 2001 - 14:31:55 MDT


Brian Atkins wrote:
> Yes evolving stuff in general can do some cool stuff, but it's a real
> pain to debug. The scientist in this article still hasn't figured out
> how his new chip layout even works... I would not want to be flying in
> a spaceship or some other kind of critical system that relied on such
> opaque chip designs.

While I certainly appreciate the sentiment, and tend to agree, it
occurs to me that we actually do this all the time. Like, every
time you get in a cab. You're trusting your life to a computational
system (human cab driver) that operates in a pretty opaque manner,
and which is known to be highly error prone. :-)

The only point I'm making is that if there was a silicon based cab
driving unit, with a self-evolved design that no human actually
understood, we might still end up trusting it *even if* we knew it not
to be perfect. It just needs to be better (and/or sufficiently
cheaper) at that particular job, on average, than the human.

> Using a raw evolution-derived piece of technology without
> completely understanding how it works is a recipe for disaster IMO.

Absolutely. Like, for instance, having humans in charge of military
missions or the "war on drugs".

-- 
*************************************************
*            http://www.nupedia.com/            *
*      The Ever Expanding Free Encyclopedia     *
*************************************************


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:36 MDT