This article appeared in 1983 in the ORSA Applied Probability SIG under the pseudonym “Anonymous”
In another country far away, there were power plants that generated cheap electricity by unclear means. Since operation of the plants involved some risk, the Unclear Regulatory Commission was established to license power plants for operations. The URC decreed that any plant could have a license if its probability of an unclear accident was smaller than 0.000001 per year. The plant operators said they were uncertain.
There was in this far away country a consultant who made his living by propagating uncertainty. When he heard about the licensing decree, he asked the URC if they were certain that 0.000001 per plant year was small enough. The URC said it was uncertain. He asked the plant operators if they were certain their probabilities of unclear accidents were less than 0.000001 per year. The plant operators said they were uncertain.
The URC and the plant operators decided they needed to propagate uncertainty. The consultant agreed to show them how to do it–for a fee, of course. He said, let’s pretend probabilities are random variables, and let’s pretend uncertainty is characterized by the standard deviation of the random probabilities. He then propagated uncertainty by a magical procedure. (We call it numerical integration.)
When they were all finished, they were aghast at the effects of uncertainties on the 0.000001 limit and the power plant accident probabilities. The URC was uncertain whether 0.000001 was small enough, and the plant operators were uncertain about their estimates of unclear accident probabilities. For instance power plant A was in an earthquake zone, and its operator concluded the uncertainty in his unclear accident probability was larger than the probability.
In this far away country, it was believed that you can solve any problem if you throw enough money at it. Thus, there were economists in this far away country too. The consultant was enough of an economist to know marginal analysis. He showed the URC how to optimally allocate its budget among sources of uncertainties-for a fee, of course. He said, spend the money where you get the biggest bang per buck; that is where the marginal reduction in uncertainty is largest per dollar.
The consultant advised the plant operators to propose research projects to the URC to reduce uncertainties. He advised the operator of power plant A to propose to reduce uncertainty by 10% per year at a cost of $100,000 per year-for a fee, of course. The power plant operator asked the consultant how to reduce uncertainty. The consultant said, “Eliminate persiflage.” This sounded good to the plant operator so they stuck that into the proposal.
The URC received many proposals and subjected them to marginal analysis. Fortunately for this story, power plant A’s bang for buck score was right on top, and the URC certainly didn’t want persiflage in the power plant, although they were unclear on where it was located. So power plant A got the contract.
Power plant A asked the jubilant consultant how to reduce uncertainty 10% per year. The consultant agreed to show them-for a fee, of course-equal to the present worth of 50% of $100,000 per year, payable immediately.
Here is what the consultant showed them:
[% reduction in year i] = 0.10*[uncertainty in year i]*100%, i = 1, 2,…
He then moved to a faraway country and settled in Los Angeles.
The moral of this story is that there is a lot of money to be made in uncertainty, but if you want to reduce persiflage, you better get paid for it up front and plan to move far away.
“Persiflage (Rhet.); ‘whistle talk’. Irresponsible talk of which the hearer is to make what he can without the right to suppose that the speaker means what he seems to say; the treating of serious things as trifle & of trifles as serious. “Talking with one’s tongue in one’s cheek’ may serve as a parallel. Hanna More, quoted in the OED, describes French, as the ‘cold compound of irony, irreligion, selfishness & sneer’: irony, paradox & levity, are perhaps rather the ingredients of the compound as now conceived.” H. Fowler, Modern English Usage.
This article originated in the Artificial Stupidity Research Institute. I am a mere “Communicant” of the ASRI.
In the late 1960s my ex-wife and I had several 4-hour dinners with Carlo Cipolla and his mistress. My ex-wife worked in the same office as his mistress, and both were very good cooks. Carlo taught part time at UC Berkeley and stayed in the Durant Hotel. Among other topics, we discussed human stupidity. I recommend Carlo’s book, “Basic Laws of Human Stupidity”, and I give credit to Carlo for inspiring the Artificial Stupidity Research Institute.