On Mon, Sep 03, 2001 at 07:48:20PM -0700, Tim Maroney wrote:
>
> http://www.kurzweilai.net/meme/frame.html?main=/articles/art0134.html?m%3D1
>
> But the input data is not cited, and the Y-axis is anything but clear to me
> for the more sweeping graphs.
Shame on Ray for not giving us academic people citations to bite into.
:-(
> Mass use of inventions, paradigm shift rate --
> these are useful things to look at, but where are the inputs? There are no
> references.
> There hasn't been a paradigm shift in physics, for instance,
> since quantum physics was developed eighty years ago.
How do you measure paradigm shifts? What about quantum field theory, the
standard model, or string/brane/whatever theory?
An objective way of measuring change in physics would be to look at when
new fundamental forces are introduced, and when they are unified.
So we have gravity and magnetism (ancient), electricity (1700's or so),
Maxwell unifying electromagnetism 1873, Chadwick & Bieler introduce the
strong nuclear force 1921, Enrico Fermi introduces the weak force in
1933, Weinberg and Salam does electroweak unification in 1967 and
Fritzsch and Gell-Man QCD in 1973.
Taking differences we get the series ~2000, ~123, 48, 14, 34, 6 years.
It does seem to speed up quite a bit, but of course my resolution is
extremely crude, we need a more fine-grained measure of physics
advances.
Now, if physics is a limited field we should see things behave like a
sigmoid or something similar. Maybe the delay in grand unification is a
sign of that - QCD is after all 28 years old by now.
> Economic graphs like education and
> manufacturing seem rather beside the point
Actually, they might be good indicators of what is actually done. The
number of literate people worldwide (per cpaita) is after all an
objective thing that tells us quite a bit about how we are doing and
does imply social revolutions just about everywhere.
> > I think one can look at aggregate values, like energy production per
> > capita, or the efficiency of various motors/generators to get an
> > estimate of overall changes.
>
> I also don't think (as I mentioned with respect to Moore's Law and
> population growth) that isolated exponential graphs demonstrate much. Has
> anyone done graphs of miles traveled per capita per year in the first few
> decades after the train was deployed, for instance, or casualties per
> bowman-hour in the century after the invention of the longbow? I wonder if
> they are any less exponential. Without some such control data, it seems
> difficult to support the hypothesis that there is something special about
> today's technological advances compared to those of previous centuries.
Most growth curves are sigmoids, that is why aggregates are interesting.
For example, the number of miles travelled per capita per year
regardless of what vehicle is used is a very interesting statistic it
might be possible to plot over historical times. Another measure might
be the amount of watts of energy that could be bought by an average
worker-day's vages.
> There would need to be some metric of social significance, and again, I'm
> just not seeing that. All I see on that point is hand-waving. The number of
> patents issued is going up, therefore we will all be supermen by 2030. Huh?
You are of course right in that all these graphs do not prove that the
Singularity is coming. But how would a metric of social significance
help us in that respect either? If the average index of social relevance
was doubling every 6 years, what would that really tell us? Relevance is
a qualitative issue.
So, maybe we should not look for exponential curves, but rather at
feedback loops driving faster and more pervasive changes. Is technology
diffusion in (and between) societies increasing? Is the average
scientist more efficient at solving problems now than in the past? Are
there any economic feedback loops introducing logics of acceleration?
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:40:25 MDT