From: Michael Vassar (michaelvassar@hotmail.com)
Date: Fri Aug 19 2005 - 11:53:04 MDT
Actually, I find it odd to hear Phil Goetz arguing about how NASA does
things and about how things are done in science. I agree that the single
genius, or the small band of innovators, are not "how things are done" in
the 21st century, but I think that it's worth noting that per-capita, and
moreso per dollar, not a lot does get done in the 21st century. Certainly
not compared to the late 19th century. Specialization in science appears to
me to have been less an innovation than a ghastly mistake. Sort of like the
abandonment of place notation around the 8th century BC or the replacement
of mouth-to-mouth resuscitation with the use of bellows in the late 19th
century. These things happen. Progress and metaprogress are not a streight
road. There may even be anthropic selection in place, preventing UFAI so
far by retarding progress.
Anyway, if we need mental superheroes to get AI, and we don't need mental
superheroes to get IA for making mental superheroes, it would seem that a
two step transition process is in order. Eliezer thinks that the necessary
IA is acquired expertise on failings of human cognition. I agree, but think
that biotech or superior recruiting for more rapid acquisition of said
expertise might be a good intermediate step.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:52 MDT