Re: Dispersing AIs throughout the universe

From: Aaron McBride (themcbrides@mindspring.com)
Date: Thu Jan 08 2004 - 19:40:06 MST


Ben Goertzel wrote:

>Eliezer,
>
>It might be significantly easier to engineer an AI with a 20% or 1% (say)
>chance of being Friendly, than to engineer one with a 99.99% chance of being
>Friendly. If this is the case, then the broad-physical-dispersal approach
>that I suggested makes sense.
>
>
>>
>>
It still doesn't make sense from a Friendliness point of view. You're
dooming the inhabitance of 4 out of 5 ship-worlds to live with an
unFriendly AI.

Also, it seems to me that either you'll know how to architect the Seed
AIs so that they all have a 99.99% chance of being Friendly, or you
won't, and they'll have a 0.00000....03% chance of being Friendly. I
admit that I have no evidence for this either way (who does?), but
doesn't it seem reasonable to you that Friendliness doesn't ever come
with 20% odds?

That all aside... it does sound like a very interesting dream. :)

-Aaron



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:43 MDT