Out of Control

by Kevin Kelly

New York: Addison-Wesley Publishing Company. 1995, 521 pp. (paper).

Long a major figure in cyber-culture, the executive editor of Wired, turns in this book to futurology. The wisdom of contemporary oracles, like that of their predecessors, emerges from dark sayings and plays on words. Although Kelly admits the future is unforeseeable, he argues that we can domesticate the things that make it too complex to grasp if we replace our logic by “bio-logic.” His book is an effort to explain what that might mean.

Its theme is the contemporary convergence of technology and biology, or what the author views as such, that consists in the replacement of external control by self-organization and self-direction. An automobile does what we make it do, while a horse has a mind of its own. The technology and social arrangements of the future, the author says, will be so complex and adaptable that they will be more like a horse than an automobile, and only through the equivalent of taming, training and selective breeding will we be able to harness them.

Kelly’s concern is less science than how science and technology are shaping the future. His treatment, which often leaves technical specifics rather out of focus, reflects the techno-libertarian view that technology and networking are creating a world of unprecedented individual choice and creativity. He surveys a variety of topics in vogue among popular science writers, from chaos theory to artificial life, and draws from them what he calls “laws of god”—unifying principles for self-sustaining systems that become autonomous and creative, and therefore “out of control,” by merging biological principles and technology. Since self-organization can apply to situations of any size and complexity, even to the world as a whole, he is able to supplement his libertarianism with techno-mysticism. Just as the order of the free market arises from the acts of myriad independent agents, and the human mind arises (so it is said) from the actions of simpler non-conscious bodily systems, a world mind could arise and perhaps already has arisen through the self-organization of the living and non-living systems that compose the world as a whole.

These speculations are thought to imply the replacement of the authorities and institutions that structure daily life, such as families, governments and employers, by free individual choice, infinitely-varying contractual arrangements, and a dimly-seen emergent cosmic mind, all arising out of a universal bio-electronic network. A Marxist might notice that these implications are favorable to Silicon Valley capitalism; others might be worried by the implied lack of concreteness and stability in social relations. The issue, however, is whether they truly foreshadow the future. The author believes that the scientific theories he discusses suggest they do.

A difficulty is that the theories tell us less than one might expect. Kelly says that Out of Control could be summarized as an update on the current state of cybernetics, the study of automatic management and control. As he recounts it, though, cybernetics and the other studies under consideration have had a spotty history. For all the excitement and froth, the understanding of the workings and self-organization of whole systems has advanced very little. Basic concepts such as “complexity” and “emergence” remain murky and hard to clarify. Chaos theory has yielded few useful results, and the artificial intelligence movement has met unyielding difficulties. Even the theories that are most firmly accepted explain less than one would expect. After more than a century and a quarter of research, it remains obscure how Darwinian evolution could give rise to the limitless creativity attributed to it, and its usefulness for understanding the course of future events is severely limited by its apparent lack of direction.

The striking thing about the topics discussed, in fact, is how little is known about them, and what is known suggests that further progress will be difficult. So why the excitement over theories that tell us so little? The answer may lie in a basic impulse that has shaped our modern world. At the heart of Out of Control is faith in the human power to predict and control, and the will to maintain that faith in adverse circumstances.

The book deals with issues relating to the reduction of systems as complex as the weather or human society to the kind of order we can control through knowledge of its mechanism. Attempts to do so routinely fail, and chaos theory suggests failure is inevitable because of the world’s irreducible complexity. It follows that if we limit ourselves to natural science, there are fundamental aspects of the world and our relationship to it that we can neither understand nor control.

It is difficult for a modern American to accept failure and ultimate ignorance in important matters. Those who reject traditional religious and metaphysical views in favor of inspirations drawn from modern natural science are disturbed by the prospect that there are demonstrable limits on what science can do for us, and if that prospect seems unavoidable they look for a way of softening its implications. Saying, as the author does, that our logic, which can deal with only very limited complexity, will be replaced by “bio-logic,” which will do better, or that we will transcend the limits on our ability to control events by creating systems that will be “out of control” but designed to serve our ends, comes at least very close to verbal exorcism of the world’s intractability.

The true implication of the studies Kelly recounts is that the world is not reason’s oyster. What lies behind his speculations is less the development of new scientific theories regarding “out of control” systems that make them somehow manageable than progressive abandonment, made more palatable by the rhetoric of scientific advance, of the illusion that control of the world as a whole through science is possible. The rhetoric is radical but the practical lesson is conservative to the core—in the author’s words, “don’t mess with what works.”

The conservative implications extend to fundamental issues regarding social life. The difficulty of restoring vanished ecologies suggests that it may be impossible to restore what social experimentation destroys. Intentional interventions in computer simulations of evolution suggest that in complex systems even small-scale reform that seems clearly beneficial is of doubtful value; even though intervention appears in the short run to increase the fitness of simulated organisms, in the longer term the organisms subjected to it tend to die out.

Radicals have long dreamed of an endlessly flexible and responsive social order that would facilitate the expression of free-flowing creativity. In its most up-to-date form the dream is that networked electronics will usher in an age of freedom greater than any previously imagined; universal connectivity will transform the world by liberating the wonderful creativity of decentralized self-organizing systems.

As the author observes, however, complex networks behave in ways one would not have expected. In fact, the studies detailed in Out of Control make it clear that the dream of limitless flexibility and freedom through self-organizing networks is no more real than the socialist dream of liberation through central bureaucratic control.

When he descends from vision to specifics, Kelly recognizes that neither complexity nor interconnectivity lead to kaleidoscopic freedom. He discusses at length Kauffman’s Law, that after a point that is soon reached interconnectivity freezes a system, and Braess’s Paradox, that adding routes to a congested network only slows things down. More generally, a striking feature of self-organizing systems is how stable they are or soon become. For any such system there are never many configurations that work long-term. An ecology such as a salt-water aquarium is unsettled for a while and then “pops” into stability, after which further changes fall within far narrower limits. The tendency toward stability is a necessary correlate of self-organization; since there is no external force to govern things, there must be internal dynamics that drive the various parts willy-nilly toward a particular common order.

The networked future will therefore be less free than many expect. With or without electronics, an assemblage of human beings forms a human society, and can best be understood as such. Men have well-defined social behavior, and they will continue to re-enact that behavior when connected by electronic networks. It therefore appears that what emerges from the net will bear a striking resemblance to the familiar machinery of daily life, such as families, employers and governments, and that the author’s “global mind” will look very much like what we now call cultural institutions and trends.

Kelly presents the future as a “Network Era” guided by new principles, but we have always lived in uncontrolled networks and have always known it. The ideal of absolute rational control has appeared occasionally in times of crisis, for example in China during the Warring States period and in Europe during the Wars of Religion and again in our own times, but in each case it has been decisively criticized by subtler and more comprehensive thinkers. Many such thinkers, favoring an organic social vision, have even made use of what the author would call “bio-logic.”

The dream that man can take charge of his world and destiny is intoxicating, but there is always a morning after. The dream of a rational and orderly society has repeatedly turned destructive, and in an age with too much faith in technology it is salutary to be reminded, by Kelly or anyone else, that centralized design and control are of limited use, that absolute coherence and stability mean death, and that in dealing with complex systems consequences cannot be determined by adding up results in particular cases. Reminding us of our situation, however, puts us not in a new world but back where we have always been.

As Kelly observes, the only way to see what a complex system does is to run it. It follows that the way to find out what an interactive network of human beings does is to study what has happened in actual societies. Our journey therefore ends with history, which is, now as ever, the queen of the sciences of man. If the world is too complex to control, as the author suggests, it must be understood through the sciences of systems we can not dominate or fully understand. These include above all the classic sciences of human things. From Plato to Pascal and Burke, and down to our own time, thinkers have analyzed the difficulties of applying mechanistic reason to human affairs and assessed the possibilities actually available to us. Those uncertain how to proceed in confused times can make no better start than to reread and ponder the classic analyses

The foregoing review has bounced around among publications that requested or expressed interest in it, but has never seen print and seems likely to remain an orphan, so here it is on the web.