RE: SIAI's flawed friendliness analysis

From: Ben Goertzel (ben@goertzel.org)
Date: Sat May 24 2003 - 22:02:34 MDT


Philip,

There's a decent argument that wisdom is NECESSARY for Friendliness to be
properly executed....

I suspect this is the SIAI position...

However, it may be useful to think about engineering/teaching/growing AI
wisdom ... as part of achieving Friendliness and as a valid goal on its
own...

ben g
  -----Original Message-----
  From: owner-sl4@sl4.org [mailto:owner-sl4@sl4.org]On Behalf Of Philip
Sutton
  Sent: Saturday, May 24, 2003 8:43 PM
  To: sl4@sl4.org
  Subject: RE: SIAI's flawed friendliness analysis

  I said:
> But I think there are aspects of the AGI morality issue that the
> Institute itself hasn't even flagged.

  Ben replied:
> Such as?

  (From memory) They don't consider the need for AGIs to be friendly to
anything else other than humans. And they see the goal to be achieving
friendliness whereas I think we need to have friendliness plaus wisdom.
(Trying to specify wisdom and implement wisdom capability is going to be
tricky for sure, but I suspect it's necessary too.)

  Cheers, Philip



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT