From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Wed Dec 19 2001 - 14:22:46 MST
Jeff Bone wrote:
>
> "Eliezer S. Yudkowsky" wrote:
>
> > The CFAI architecture doesn't preclude an infrahuman AI being used for
> > commercial purposes.
>
> Would it preclude a posthuman AI from being used for such purposes, and if so,
> how and why?
The technical answer is that yes, it would, because an posthuman AI makes
independent plans in pursuit of Friendliness, which means that a
commercial agency has no hold over the consumer; that is, no ability to
deny the benefits of the posthuman AI to the consumer. The only way you
can make money off the creation of a posthuman Friendly AI is by
collecting voluntary tips from a grateful populace, just as with other
nonrival, nonexcludable goods.
The nontechnical answer is that a posthuman Friendly AI makes the existing
economy irrelevant. After some thought, I have finally come up with a
compact way to express my intuitions on this subject: Let's suppose the
current value of the planetary economy is a quadrillion dollars flat. Now
let's suppose that a quintillion dollars worth of wealth is dumped into an
egalitarian Friendly AI. Who cares who used to be rich or poor? Who's
even going to notice? The wealth differentials we've been fighting over
are ripples in the surface of the pond, while the tidal wave gathers over
the horizon.
-- -- -- -- --
Eliezer S. Yudkowsky http://intelligence.org/
Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT