From: Eliezer S. Yudkowsky (firstname.lastname@example.org)
Date: Sat Mar 15 2003 - 06:34:05 MST
Damien Broderick wrote:
> At 06:14 AM 3/15/03 -0500, Eliezer wrote:
>>Steven Pinker gets everything right. Impressive.
> Weelllll-- Like this?
> "Now, why are technological predictions so often wrong? First, there's a
> habit of assuming that technological progress can be linearly extrapolated.
> If there's a little bit of progress now, there will be proportional
> progress as we multiply the number of years out.
> "Engineers sometimes refer to this as the fallacy of thinking that we can
> get to the moon by climbing trees. A little bit of progress now can be
> extended indefinitely. "
> Sensible caution before such an audience, no doubt, but unlikely to be
> correct except accidently (since the extrapolation toward Singularity
> *won't* be linear in that boring sense).
Yes, but Damien, he's *right*. Let technological progress be a vector
which produces vector changes in culture and society. How do naive
futurists (Pinker may have never encountered a non-naive futurist)
extrapolate the future? If we have delta time of ten years from, say,
1990 to 2000, with corresponding delta tech producing delta culture and
delta society, then the naive futurists project that in 2010 we will see
2000 plus the same old delta tech, delta culture and delta society.
This is, indeed, analogous to thinking you can get to the moon by climbing
trees. Now you *can* get to the moon, much faster than climbing a very
tall tree would allow for, but not by an extension of the climbing principle.
The Singularity has nothing to do with supposed or real acceleration in
technological progress, delta tech or d-squared tech, except insofar as
the absolute level of tech starts creating delta smart.
-- Eliezer S. Yudkowsky http://intelligence.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT