[unrev-II] "S" curve limitations related to bootstrapping and evolution

From: Paul Fernhout (pdfernhout@kurtz-fernhout.com)
Date: Thu Jan 20 2000 - 08:32:56 PST

From: Paul Fernhout <pdfernhout@kurtz-fernhout.com>


People often are reluctant to get involved with anything that seems to
promise big returns for a small investment. It just sounds "too good to
be true". Bootstrapping also has some scary aspects, both in who uses it
for what, and in what the future will bring, especially if the future
comes at us even faster as a result of bootstrapping.

In a sense, bootstrapping as described now is like a nuclear explosion
-- a few neutrons are set off going through nuclear material to
make more neutrons and so on until there is an explosion. This is
positive feedback. Yet, even in explosives there is some limit to the
size of the explosion -- this is due to limited resources based on the
amount of nuclear material placed in one area.

Another issue is explicit negative feedback. Negative feedback is like
the way water level in a toilet tank causes a float to rise which shuts
off a valve, stopping more water from coming into the tank. Positive
and negative feedback together make up living systems.
The early cyberneticists found that without negative feedback all their
control devices oscillated into uselessness (or self-destruction).

In human designed systems, positive and negative feedback together often
reflect a sense of "purpose" built into the mechanism. For example,
when a thermostat detects cold below a range, it turns on a heater. When
the thermostat detects warmth beyond a range, it turns off the heater.
Positive and negative feedback together reflect the "purpose" the
inventor of the thermostat had in mind -- creating a device that keeps
temperature in a narrow range.

Positive feedback and negative feedback (through limited resources or
active control) also lead to the common "S" curve seen throughout
Sometimes one "S" curve can build on another, leading to multiple
plateaus. For example the population of a new species of frogs on an
island might be limited by available fresh water, and form an "S" curve
to "carrying capacity", the most frogs the environment can support.
Eventually, there are so many frogs living over such a long time that a
few survive being carried way by birds-of-prey and are accidentally
dropped (alive) onto new islands which they then populate. So the total
frog population across all islands summed together might appear to be a
series of "S" curves, as new populations form based on almost chaotic

"Bootstrapping" as a concept needs to incorporate the idea of an "S"
curve and multiple plateaus if it is to seem more realistic -- given
that no exponential growth curve in nature lasts for very long. There
are always diminishing returns as some resource is exhausted -- until of
course there is another breakthrough (like a frog dropping on a new

For more details on "S" curves, see:

A diagram of what I mean by multiple plateaus:


The evolutionary theory related to "Punctuated vs. Gradual Evolution"
inquires about the issue of multiple plateaus in depth.

As a simplified summary based on (still unpublished?) thinking by my
advisor when I was in the SUNY Stony Brook Ecology and Evolution
department, Lev Ginzburg http://life.bio.sunysb.edu/ee/ginzburg.html ,
the issue may depend on the time scale you look at and your sampling
rate. When you look very close, things are punctuated (mutations), when
you look far away, they are gradual (sets of mutations in some
direction), even further, punctuated (whole new areas of resources open
up occasionally, after long periods of stasis, but those mutations and
situations are rarer), even further, gradual again (trends across
resource use over long period of time), and so on.

Unless "Bootstrapping" incorporates this idea of multiple "S" curves,
users may give up hope once the hit the first plateau defined by limited
resources or negative feedback. I would suggest the 1968 demo was made
at a plateau, at the top of a bootstrapping "S" curve related to
1950s-1960s NLS/Augment work, (from which we have since slid). I'm not
sure what the limiting resource was in 1968 -- perhaps it was the
imagination of the people who saw the demo, or perhaps it was the
mismatch between funders' world views and how far Doug and the NLS team
had progressed. I'm sure the people who participated could say more
exactly what the limiting resource was in the 1970s.

That NLS/Augment work needs to be repeated as an open source project,
but, it will likely still plateau (at a new level perhaps).
However, since the environment has changed (the internet, cheaper
computing), from that plateau, NLS/Augment's evolution will likely
then surge off again in some new directions, heading to new plateaus, as
the global "carrying capacity" for innovative IT projects has changed.

To conclude, I'm not saying the idea of positive-feedback bootstrapping
is wrong. I just think bootstrapping needs to acknowledge its
limitations and build them into the bootstrapping process and the
presentation of the idea.

-Paul Fernhout
Kurtz-Fernhout Software
Developers of custom software and educational simulations
Creators of the open source Garden with Insight(TM) garden simulator

--------------------------- ONElist Sponsor ----------------------------

Independent contractors: Find your next project gig through JobSwarm!
        You can even make money by referring friends.
<a href=" http://clickme.onelist.com/ad/jobswarm2 ">Click Here</a>


Community email addresses:
  Post message: unrev-II@onelist.com
  Subscribe: unrev-II-subscribe@onelist.com
  Unsubscribe: unrev-II-unsubscribe@onelist.com
  List owner: unrev-II-owner@onelist.com

Shortcut URL to this page:

This archive was generated by hypermail 2.0.0 : Tue Aug 21 2001 - 18:56:36 PDT