Re: Predictions (was The Spike, nanotech, and a future scenario)
Tue, 7 Oct 1997 22:07:16 -0400 (EDT)

In a message dated 10/7/97 1:01:09 PM, wrote:

>> Gosh, I can make oodles of predictions if they don't have to be right :-)
>WHich is true...
>However, it seems by the completeness of there site/predictions that
>they 1) have a good idea about future tech and perhaps used some
>methods to reach those conclusions. The various conclusions seem to
>mutually consistent.

If you don't know the methods, you have no way of evaluating the accuracy of
their predictions save by waiting around. That destroys the value of their
predictions. How do you know it's anything other than wishful thinking? Or
whether it's based on some silly numerology, like many apocalyptic
predictions? Even straightforward extropolation is very dubious.

Moore's law isn't just an extrapolation; it also rests on the observation
that photographic etching processes don't change much in expense when the
resolution is increased. Setup cost go up, but so far it's been possible to
increase the chip volume to make the per chip setup cost minimal. This
theoretical basis for Moore's law increases my confidence in its reliability,
and also points out 2 big things which would probably break it: substantial
per chip setup costs and switching to a production basis other than
photographic etching.

>Nano tech toys by 2010 seems a bit fast but who knows..

Given that after the better part of a decade we've yet to see any nanotech
mechanism, I'm not holding my breath. Maybe we'll have a working nanogear by
then - I guess that would count as a toy.