A common link running through a number of contemporary issues is that of dealing with disaster whether manmade (terrorism, oil spills) or natural (hurricanes, earthquakes, floods). In my view the correct strategy for dealing with all of these is essentially a common one. Although some commonsense regulation like improving building standards in earthquake zones, evacuations during hurricanes, and so on are just that—common sense—it isn’t cost justified, prudent, or even possible to preclude them completely, to eliminate their risks.
As was said recently in comments you can’t plan for unforeseen occurrences but you can mitigate the risks. You do this by decentralizing, redundancy, having resilient processes. This can be in direct contradiction to some notions of efficiency. The more specialized the niche in which any organism functions, the more vulnerable it is likely to be to change.
I’m sometimes reminded of the wisecrack from the great book The Mythical Man-Month: “If builders built building the way programmers write programs, the first woodpecker to come along would destroy civilization”. One of the things I learned in the course of the extensive addition we put on our house a year or so ago is that builders build buildings exactly the way programmers write programs and that when things don’t go exactly according to plan they just adapt, modify, and go right on. We used to speak of adaptivization, of building change into processes. I don’t know whether that term is used any more or not.
Another perceptive quote from Helmuth von Moltke: “No plan of operations extends with any certainty beyond the first contact with the main hostile force.” That’s true of everything, not just in war, and it’s the reason that standards, regulations, and other attempts to bend circumstances to the wills of bureaucrats can never be successful.
In responding to terrorism we have done almost completely the very wrong thing. We’ve centralized rather than decentralizing and attempted to do with regulations what can only be done with prudence and observation, much harder commodities to come by. It appears to me that the urge for more regulations to mitigate the risks of hurricanes, earthquakes, and oil spills suffers a similar failing. Rather than trying to prevent them we need to have systems that adapt to them more fluidly and better strategies for responding to them
“As was said recently in comments you can’t plan for unforeseen occurrences but you can mitigate the risks. You do this by decentralizing, redundancy, having resilient processes.”
But Dave, as I understand it, the panic last week was in part cause by AMEX not being in sync, computer-tradingwise, with many of the newer exchanges. That would seem to militate against decentralization–these guys were reading from different technological playbooks. Doesn’t this indicate the need for some kind of centralization, if only to get the players reading the same playbook? Not that this would solve the complexity problem, but it would serve some interests of mitigation, I believe.
Dave,
Your comments on decentralization and resilience remind me a lot of John Robb over at Global Guerrillas.
I have direct personal experience albeit in other industries which would suggest that isn’t what generally happens. I have seen the following happen in several different industries:
Step 1: Some brilliant new idea is developed
Step 2: The company for whom the guys developed the idea make a lot of money
Step 3: The guys who did the developing go to work for a competitor and do materially the same thing there.
Step 4: Lather, rinse, repeat.
I find that there’s a remarkable convergence of technologies rather than a divergence. Monoculture is risky. And commonplace.