From: Christian Rovner (crovner@lacasilla.com.ar)
Date: Wed Mar 12 2003 - 14:02:14 MST
Sébastien Chikara wrote:
>
>Here's the latest flippant argument I had against the singularity
>hypothesis:
>"One of the most common fallacies people fall into is to assume any given
>trend will retain its current path forever. It ain't so."
If your hypothesis is that the Singularity can be predicted, analyzed or
understood by extrapolating current trends, I would have to disagree. All
those graphs, whether correct or not, are mostly irrelevant to the Singularity
nowadays. The creation of smarter-than-human intelligence does not depend on
Moore's Law or any other statistical trend. Better hardware would make some
things easier, of course, but it's not a *necessary* condition anymore. The
only hard problem that separates us from a Singularity is the complete,
correct and detailed theory of intelligence. (And, in case we want to tell the
story, we better have an equally good theory of Friendliness too.)
So I didn't really deal with your friend's argument but with yours. However, I
hope you agree.
-- Christian Rovner
Singularity Institute for Artificial Intelligence, http://intelligence.org/
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT