hal@finney.org wrote:
>
> Scerir forwards:
>
> > Only when the
> > researchers introduced a "rogue" minibrain with more intermediate
> > neurons to analyze the past rounds did it attain more than a 50 percent
> > success rate. Their simulations suggest that intelligence often hinges
> > on how much one can make use of the data in its physical
> > environment. (Wakeling and Bak, Physical Review E, November
> > 2001)
>
> Gee, let's see, they made a network with more layer-two neurons and it did
> better than the ones with fewer. Making a bigger BRAIN made it do better.
> Does that surprise anyone? Seem like an earthshaking result?
Those of you who, like me, have been thinking "gee, what a bunch of
blithering idiots" should remember that the blithering idiots may be the
folks who wrote the press release, not the scientists themselves.
Possibly someone asked the researchers "Well, can you explain this
research to the public?", understood maybe one-twentieth of the resulting
information stream, and then used that to write the press release. Say, a
researcher said something about the embodiment problem as background
material and the writer thought that was the research result. Anyone
remember when the IBM marketroids announced the achievement of
teleportation in a Scientific American ad?
-- -- -- -- --
Eliezer S. Yudkowsky http://singinst.org/
Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2b30 : Sat May 11 2002 - 17:44:17 MDT