Å®ÉúСÊÓƵ

Out of Control by Kevin Kelly, Fourth Estate, pp 521, £16.99

The title of Kevin Kelly’s weighty book derives from a phrase currently
used by robot builders to define their hoped-for product: ‘Fast, cheap
and out of control’. Building robots with even a modest repertoire of simple
physical actions is far more difficult than intuitively expected, and one
capable of carrying a cup of tea without spilling it, for instance, eludes
the practitioners. Building an intelligent robot is an even more distant
target.

Recently, a coterie of roboticists has turned away from the traditional
top-down approach (that is, a powerful central computer programmed to generate
intelligent behaviour). In its place, the researchers have adopted a bottom-up
concept in which they assemble many simple cognitive functions, out of which
(they hope) intelligent behaviour will emerge. The absence of central programming
leaves the system out of control.

The new philosophy is strongly influenced by recently evolving ideas
about biological systems, in which complex structures or behaviours are
seen to emerge from the interaction of many small parts. This importation
of biological concepts into the way we construct complex systems, from computer
software to information networks, will become more pervasive and more powerful
and will transform society, argues Kelly.

‘Our future is technological,’ he concedes, ‘but it will not be a world
of grey steel. Rather our technological future is headed toward a neo-biological
civilisation.’ Humans have always exploited nature, first as hunter-gatherers,
then as farmers, then as forgers of new materials dug out of the Earth as
the Industrial Revolution transformed the planet. ‘Now,’ suggests Kelly,
‘Bios is yielding up her mind – we are taking her logic.’

Kelly’s thesis is that the logic of mechanics is limited in the complexity
that can be coaxed out of it. ‘Truly complex systems such as a cell, a meadow,
an economy, or a brain (natural or artificial) require a rigorous nontechnical
logic. We now see that no logic except bio-logic can assemble a thinking
device, or even a workable system of any magnitude.’

Kelly sees himself as a champion of an imminent revolution, and in such
role coins new words and laws. For instance, he suggests the word ‘vivisystem’
for ‘the common soul between the organic communities we know as organisms
and ecologies, and their manufactured counterparts of robots, corporations,
economies, and computer circuits’. All such systems follow unifying principles,
which Kelly has enshrined as nine ‘laws of god’ – the fundamental forces
that are ‘shared by all self-sustaining, self-improving systems’.

By now it should be apparent that at the core of Out of Control is
complexity theory, which is concerned with the emergence of order – usually
unpredictable – out of dis-order. Most researchers studying complexity are
working on physical systems, such as computer models. Only a few are biologists
or working with natural systems, although their number is growing. By
concentrating on biological systems and their properties, Kelly has chosen
the more interesting and complex of complex dynamical systems.

He is right to see biological systems as the richest source of concepts
for future exploitation, if they can be discerned and understood. The power
of such systems – in their steady evolution and improvement – raises the
real possibility, however, that artificial systems based on them might quickly
come to elude human understanding, given their self-generated complexity.

In his work for the book, Kelly interviewed most of the prominent complexity
theorists, and readers familiar with the subject will meet old friends here.
Stuart Kauffman’s story of the discovery of ‘order for free’ is retold,
as is Chris Langton’s role as midwife of the new field of artificial life.
Tom Ray’s evolving ecosystem in a computer is there, too, and so are Stuart
Pimm’s insights into the stability of real ecosystems. And so on.

Kelly’s odyssey took him from the Media Lab at the Massachusetts Institute
of Technology to Disney Studios, from Biosphere 2 to a prairie in the Midwest.
It is perhaps not surprising that Kelly, formerly publisher and editor of
the Whole Earth Catalogue, should be awed by his visit to the ill-starred
eco-bubble in the desert that is Biosphere 2. ‘I felt I was standing in
a cathedral,’ he said of his experience in the nearly completed structure,
in the spring of 1991.

There is no doubting Kelly’s conviction that a new age is dawning. And
he correctly identifies novel frontiers of thinking about biological systems
that will change the way the natural world is perceived. These same insights
can and have been applied to artificial systems, such as financial markets
and the rise and fall of civilisations. It is also true that the principles
of the theory have been applied directly, such as the improvement of computer
programs through their own evolution.

Despite this, Kelly’s excursions through the different fields of the
theory, as related in Out of Control, surprisingly lack the coherence or
cohesion that would be expected – if the intellectual movement is as cogent
as he argues. And the prospect of a neo-biological civilisation seems distant
in these pages, despite the repeated statement of its putative imminence.
The message may well be correct, but the messenger has striven too hard
to deliver it, thus causing it to be obscured.