Re: How hard a Singularity?

From: James Higgins (
Date: Sat Jun 22 2002 - 19:25:21 MDT

At 08:13 PM 6/22/2002 -0500, you wrote:
>James, when we say human equivalent we don't mean performance, we mean
>algorithms. Doing the thinking using the same algorithms that a human
>uses. We assume we can figure out these algorithms because we have an
>example to study. These human equivalent algorithms will run faster and
>more effectively on a silicon substrate than a biological one.


What proof do you have that these algorithms will be more effective in
silicon? It is very plausible that they will, at first, be much less
efficient in silicon since they have been refined for many, many years to
be effective on our biological substrate. The first AIs may also be short
on raw computing power in comparison to the human brain. Once the first
reasonable AI is up and running we'll know if these algorithms are more or
less efficient in silicon, not before.

James Higgins

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:39 MDT