Retraction about intelligence and computing power

From: Philip Goetz (philgoetz@gmail.com)
Date: Mon Jul 31 2006 - 19:51:06 MDT


I forgot to make a retraction.

Some time ago, I made some posts to the effect that algorithms for
intelligence might necessarily scale logarithmically with computing
power - so that some numeric measure of intelligence would scale as
the logarithm of the computing power.

I believe this is true for most objective numeric measures of
"intelligence" that I can think of, such as factoring primes. I think
that we shouldn't plan on having intelligences that can hold 1000
items in their short-term memory. However, it also seems true that
small increments of intelligence would produce great effects and
disruptions. One might guess that the impact of intelligence grows
exponentially with intelligence. So that we arrive with impact that
increases linearly with computing power. :)



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT