Re: Existential Risk and Fermi's Paradox

From: Eliezer S. Yudkowsky (
Date: Wed Apr 18 2007 - 11:17:00 MDT

Larry wrote:
> It now seems to be standard dogma to say "evolution doesn't advance", a
> reflexive response to the old dogma showing mankind as the end goal of
> evolution. Neither is accurate. Evolution is like a random walk, and
> random walks do expand over time.

A common misconception, which I believe was first spread by the
scientifically dishonest Stephen J. Gould, misrepresenter extraordinaire
of modern evolutionary theory. (For more on Gould see

Evolution has a complexity limit, the amount of genetic information that
it can support against degenerative mutation, given a copying fidelity
and a maximum selection pressure. This amount would not exceed 10^8
bits for mammals. It would have been reached long ago. Thus, the
random walk has an upper limit and does not expand. Since then we have
simply been substituting new complexity for old complexity as adaptive
challenges change.

Much of this discussion on this thread shows misunderstanding of basic
evolutionary biology combined with no realization that more advanced
understandings exist. I recommend "Adaptation and Natural Selection" by
George Williams for starters (this is the book often said to have
started the "Williams Revolution" of the late 1960s in biology).

Eliezer S. Yudkowsky                
Research Fellow, Singularity Institute for Artificial Intelligence

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:57 MDT