October 6, 2009

The promises of the Information Age should not blind us. The more we heap data into the modern equivalent of haystacks, the more truth resemble the proverbial needle concealed in the hay. Faulty concepts however can be more damaging than the sheer proliferation of doctored data.

Take for example Randy Cohen's recent pronouncement on teacher rating sites (*). To a professor who asks "may I suggest to my more-satisfied students that they post a rating?", the ethicist in residence at the New York Times Magazine answers by the negative. We encountered a similar issue when writing about physician ratings. That such Internet sites are unfair in practice is not under dispute. The rub is with the remedy.

My own proposal is to compel each rating to carry the identity of the rater and to enable rating by both parties. As their users would rather shrink from the burden of responsibility, no doubt these measures would threaten the commercial success of rating sites. No one expects their owners to endanger their lucrative use of anonymized delations nor the potential benefit of offering advertisers the personal information they do collect from the raters. Meanwhile the powerlessness of professors precludes their mimicking Medical Justice, which would forbid their students to post a rating.

To stay within the bounds of practicality then, why not coax favorable reviews? Toyota has shown what efficiency can be achieved in this domain. The professor, it must be noted, does not talk about rating herself, either directly or via her friends and family. So why not endorse her good natured remedy? It can only be because it would threaten the integrity of the rating, "cooking the books" as goes the quote of a tenured professor.

This dangerous error of judgment confuses appearances with reality. True, the publishing of anonymous ratings look like a democratic ballot and tempering with democracy, however small in the beginning, is ruinous in the end. What professor would want to be known as a campus Khamenei? But Internet ratings are not democratic. To shower this tool with the reverence we ought to feel for democracy may actually undermine democracy.

Popularity is indeed one source of truth and Internet is congenial to its popularization. Unfortunately what allows "the wisdom of crowds" to emerge can equally channel "the madness of mobs". Andrew W. Lo had financial markets in mind when contrasting these two outcomes (**), but his remark is valid wherever popularity is in play. YouTube has already been used to hunt teachers out of a job. Now "someone [has created] a poll on Facebook asking respondents, 'Should Obama be killed?' The choices were: No, Maybe, Yes, and Yes if he cuts my health care". Thomas L. Friedman pointedly reminds us how real killers are inspired and feel justified by such virtual madness (***).

Speaking of financial markets, they strive to create liquidity but do so in a dynamic environment. Adding the potential for earthquakes to oceans is of course the formula for unforeseeable tsunamis. Still one can try to better protect coastal areas, better estimate the likelihood of earthquakes, better measure the propagation of tsunamis once under way. Reviewing "new models for markets" (****), John Authers credits Andrew Lo, professor of financial engineering, with "arguably the best known idea, the adaptive markets hypothesis, by borrowing from evolutionary theory".

I leave the analysis of adaptive market theory to another time. Without casting any aspersion on it, one may remark the present popularity of evolution theory can be a source of some severe misconceptions of its own. John Kay for instance shrewdly uncovers intellectual traps beneath "Herbert Spencer['s...] striking phrase 'survival of the fittest'" (*****). Being here today implies neither the past superiority nor the future relevance of all one's current attributes. Barring Providence, a particular evolutionary path is "[neither] inevitable [nor] desirable". "One roll of the evolutionary dice", it is also likely to be irreversible according to Dr. Thornton, a biology professor, quoted by Carl Zimmer (******).

Having forced some welcome humility on us all by coining a phrase of his own, "evolution is smarter than you", John Kay has a second warning to impart. "The specific mechanisms of organizational evolution differ from those of biological evolution". He stops short however from clarifying what is for me the crucial difference. While Darwin's theory has displaced Lamarck's as the better explanation of the biological record, there is no reason today for extending Darwin's victory to the science of organizations. Invoking "the" theory of evolution in the instance risks building castles in the air.

Take Microsoft as led by Bill Gates. For John Kay, it would be an error to think of the former as "the expression of the will" of the latter and he is right up to a point. Yet, when Bill Gates famously decided Microsoft had to adapt to the Internet and incorporate a browser in its operating system, wasn't it a case of Lamarck's giraffe stretching its neck by a couple of inches to better graze the trees, thus depriving Netscape from vital fodder?

In a world subject to irreversibility, the deliberate will of individuals is, for socio-economic constructs, as determinant as the accidents in replication. To a scientific mind, there is evolution all right but still no consistent theory behind it. For a field study, forget General Electric and Microsoft. Tackle an organization with a slightly more involved historical record like GDF Suez (1).

For poets, the investment bank founded by William I of the Netherlands (2), the canal dug by Ferdinand de Lesseps (3), the gaz and electricity distributed by Gérard Mestrallet (4), all sing the single minded focus of turning fluids into cashflows. For writers, the unlikely history of this tenacious surviver would make a gripping novel, what with a President of the United States, Republican no less, tacitly approving the nationalization of its main asset by publicly backing an Arab firebrand against the Israeli government of the day. For the financial engineer, it remains an unfolding mystery.

To think misconceptions do not contain an element of truth is of course another misconception. Google does not automatically rank the truth on top of the lies, but, if you have enough time, it will help you find it. And if you value your privacy, bid your time for evolution to work out a solution.

Philippe Coueignoux

• (*) ........... Grading the Ratings, by Randy Cohen (New York Times Magazine) - September 27, 2009
• (**) ......... When the wisdom of crowds becomes the madness of mobs, by Andrew W. Lo (Financial Times) - September 25, 2009
• (***) ....... Where Did 'We' Go?, by Thomas L. Friedman (New York Times) - September 30, 2009
• (****) ..... Wanted: new models for markets, by John Authers (Financial Times) - September 29, 2009
• (*****) ... Evolution is the real hidden hand in business, by John Kay (Financial Times) - September 30, 2009
• (******) . Can Evolution Run in Reverse? A Study Says It's a One-Way Street, by Carl Zimmer (New York Times) - September 29, 2009
• (1) see GDF Suez in the Wikipedia.
• (2) see the Company de Suez in the Wikipedia.
• (3) see the Suez Canal in the Wikipedia.
• (4) the Wikipedia is wholefully lacking on the laborious birth of GDF Suez. But how ensure a truthful expansion of Gérard Mestrallet's stub?