248 – Could Internet fail “catastrophically”?

Posted on August 20, 2013 by

0


Internet has become a vital part of modern society. 25 years ago Internet barely existed. Today, the globalized world in its current form could not survive without it. Internet has been one of mankind’s great enablers. I would place it in the same league ad plant and animal domestication.

I was asked the other day whether Internet could fail “catastrophically” and whether one should make contingency plans for it. I felt challenged by the question – and not in the politically correct sense of the term.

248

On the technical side the question puzzled me. Internet had been designed as a network precisely in order to avoid catastrophic failure. One cannot “take out” a network without first cause and mover. one Internet has no head one can destroy, disabling subordinate structures in turn. This kind of design has consequences. It works, but we can no longer say why. Where there is no linear chain of causality, we are left with the puzzle of “emergent” properties of the system.

But what about the “rule side” of Internet?

Gasping, I grasped for analogy. Among the greatest network structures we can point to, are the markets. Markets too are a complex network. There is no center as well as subordinate periphery. The markets are simply a set of neutral rules determining how transactions are carried out. This a set of rules covers property rights, contracts, liability, whatever. They reflect a grown accumulation of experience. The rules are neutral, in the sense that they apply to everyone and should not arbitrarily favor a person or a group.

As we deepen our understanding of markets, or as the context changes, we fix the rules to improve their functioning. This is an adaptive process, based on trial and error. Of course, over time the rules may become encumbered with exceptions, as favors to interest groups accumulate like barnacles (just think of the tax code with its innumerable details favoring this or that private interest). Major reforms may become necessary to restore faith in the neutrality and fairness of system. True, adaptations should be slow but relentless, rather than operate by fits and starts. This is hardly ever the case.

Markets function because participants all abide voluntarily by the rules – we call it trust. Trust is an unspoken social compact that makes social life possible. Mistrust is a terrible waste of time. Of course, just in case, the state is there to enforce the rules when needed. The state can only enforce so much, however. Without self-discipline the system would soon collapse.

A “catastrophic” failure of the markets would occur if the state fails. The law of the jungle replaces the rule of law. Civil war and revolution e.g. challenge the very existence of the state. Experience tells us that such events badly weaken the effectiveness of markets. Nothing need be intrinsically wrong with the market rules – but the surrounding circumstances hinder their working. In terms of complexity economics a “catastrophic” failure is a situation where the rules no longer apply. When the state fails, we have a “catastrophic” event.

Repairing a “catastrophic” failure points in the direction of strengthening the enforcement capacity of the state; not necessarily to a reform of the rules (though one may use the opportunity to weed out privileges).

Civil war is an acute form of “catastrophe”. Corruption, on the other hand, is akin to a chronic degeneration. The rules remain in place, but everyone bends them as she sees fit; the state is party to the corruption, or is impotent to do anything about it, because corruption becomes so diffuse. Corruption is civil war in slow motion. Also: corruption and incompetence are the twins that bring down the state. Usually one observes corruption as an archipelago in a sea of incompetence. They are a subtle poison, for they involve not so much the state as an institution, but all of its agents. The German state was soon reborn after 1945. Freeing a state of corruption and incompetence may take a generation or more. The twins replicate themselves readily.

What does this analogy tell us about the Internet? Forces external to Internet are more likely to trigger “catastrophic” failures than intrinsic problems. When its rules are no longer applied, Internet will wither too. The causes of the failure are not intrinsic to the Internet, but external to it.

Planning for a “catastrophe” only addresses intrinsic causes. In a complex system intrinsic deficiencies usually have precursors. The system bends before its breaks. Tremors shatter complacency. Vigilance about the functioning of the system should suffice to yield early warning. Failure of the context, however, cannot be foreseen.

A “catastrophic” failure of the Internet: What, me worry?

Posted in: Uncategorized