🚀 go-pugleaf

In article <Bwxqxw.sq.1@cs.cmu.edu> amon@elegabalus.cs.qub.ac.uk
writes some stuff about extrapolating exponential progress. I agree
with some of it and disagree with some of the rest, and I think the
briefest way of getting my point across is by replying to this
snippet:

			My own feelings are that 2020 is about the limit  
   at which things start getting very strange: the exponential will be  
   essentially vertical with respect to human perceptions of time.

Technological development is an aspect of human activity. This will
remain true until such time as we have strong AI (where this can be
defined as computers which can `do' technology on their own). We don't
have that now, we're not particularly close, and I don't believe we
will have it before 2020. Until that time, there is a maximum slope to
the technological graph, determined by things like how fast are
results disseminated, how fast can experiments be set up, and (most
limitingly?) how quickly any given researcher can think of a new idea.
Thus until we have strong AI, the exponential _cannot_ become
`essentially vertical'. Sure, it'll get steeper, but not indefinitely.

A separate point is that a lot of people don't like the
faster-and-faster aspect of modern life. This portion of the
population will grow (IMO) as the slope steepens, and may well be able
to apply brakes at some point. Speaking for myself, I don't want to
have a newer, better, faster car every day (or even every year). Every
five or ten years is fine by me.

Nick Haines nickh@cmu.edu