Richard Clark wrote:
. . .
All-in-all, the introduction of a new technology is frequently
confused as a better version of an old technology - something like
saying facsimile would replace the newspaper - or that the utilities
would pay us to use nuclear power. All the "forecasts" mentioned in
this article rank right up there with these world class pipe dreams.

On the other hand, sometimes the assumption that the new technology will
simply replace the old falls staggeringly short. The total market for
transistors was initially seen as being the same as for tubes -- replace
each tube with a transistor, and that's it. Hardly worth developing the
technology to overcome the gnarly manufacturing problems (e.g., extreme
purity requirement of the base material). No one foresaw the integrated
circuit, making it practical to put the equivalent of hundreds of
transistors in a pocket calculator, wris****ch, or even an electric
iron. The transistor made possible a whole new technology with
applications which were altogether impossible and therefore unimaginable
with tubes. But the best that the soothsayers can ever seemingly do is
to extrapolate from what we've got right now. Maybe the nanotubes won't
end up being simply a replacement for wires, but the basis for a whole
new technology we can't now conceive.

And maybe they won't. Every entrepreneur does his best to convince
investors that his invention will be the next integrated circuit or his
garage company the next Microsoft. But the odds are sure against it.

Roy Lewallen, W7EL