The long tail of a complex system
Most attributes of complex systems studied by the exact
sciences may be approximated by the Gaussian (normal) distribution. Asymmetric
or skewed distributions are transformed so as to make them normal, e.g.,
the log-normal distribution. On the other hand life generates only skewed
distributions. Hitherto those distributions were also transformed so as
to make them Gaussian. The short asymmetric tails were regarded as noise
which ought to be reduced by an appropriate transformation.
The Gaussian distribution (model) is generated from random phenomena, and
since nothing in life is random the Gaussian
model fails to capture some essential attributes of life. Life is
manifested by the allometric law with
asymmetric tails.
After a century of ignoring these tails as random noise, developers encounter
this phenomenon in the Internet, and since it is bad for business, it is
unpleasant. Particularly since the area under the long tail is
larger than the body that wags it. Hitherto you believed that the
distribution-body harbors most of your customers only to discover that they
hide in the long tail. The phrase “Long Tail” was coined by Chris Anderson,
and highlighted by an influential essay by Clay
Shirky, "Power Laws, Weblogs and Inequality" (February
2003) http://en.wikipedia.org/wiki/The_Long_Tail.
The long tail is a manifestation of the allometric law (a
power law) which is an expression of life, and since the Internet is a
living system it behaves allometrically. Thus, in order to improve your
business, you ought to abandon the outdated Gaussian distribution, and
turn your attention to life. You cannot generate a long tail from random
elements, since phenomena generating long tails are creative. My CA
models are such creative constructs which generate long tails.
The long tail carries with it a sad message which led some to conclude that
Web 2.0 is dead. On one hand the web is flooded with information,
yet most of it is redundant, irrelevant, and misleading. Suppose that
you fell ill with cancer. You turn to the experts and get treated. Nevertheless
you worry. They talk about probabilities of cure, while you want a definite
answer. What about alternative cancer treatments? Google spits out 10,700,000
links within 0.09 seconds. You start reading and reading only to encounter
the non-relevant. The distribution of these links forms a body with a long
tail. The first n pages consist of the body, while the creative links
are distributed somewhere in the long tail.
More on cancer
The optimists hope to remedy this with new search engines of Web 3.0 . Yet
the future lies in Web-Bionics, whose basic ideas were
introduced in this thread.
Back to complexity index