August 12, 2008

After "The End of Oil", comes "The End of Food", reviewed by Harry Eyres (*). Will Paul Roberts end his "End of" series with "The End of Data"? While this would fit neatly in our taking food, energy and data as the engines behind the three ages of History, the reader may think I am ready to go to no end to find a good beginning. Doesn't data differs from food and energy in that it can be used without being consumed?

What if the comparison had more depth than apparent at first glance? War and waste weigh heavily against humanity meeting its needs in terms of food and energy, apt as we are to manufacture want in the midst of overabundance. Second we are far from overtapping the physical limits of our current planet. It's just that most of the solar energy it receives and the biomass it sustains is not so easily converted to human use.

In looking at financial disclosure documents, Aline van Duyn (**) similarly argues that, while "prospectuses have a lot of information", much "information is not always of value" to the investor. What characterizes data relative to food is not so much a question of greater availability as of human greater ability to generate the wrong kind to feed its appetite for the right kind.

Take the FBI inquiry on the anthrax letters which killed five people seven years ago. Understandably the public wants to know who is the murderer. Scott Shane and Eric Lichtblau (***) (****) are left to relay circumstancial evidence against a scientist who committed suicide last month. If the first suspect had escaped exposure in the same way, who doubts the case would have been closed years ago? But is justice served when lack of mental strength makes you guilty as accused? In former times, judicial torture had also for effect to favor those best able to bear physical pain.

There lies a plot for a detective novel. Deadly, traceable germs are stolen from a lab and used to make enough victims to raise the ire of the public. One by one starting with the weakest, the ten lab employees who had access to the germs are hounded to their grave by the police. The real culprit? Not even one of the suspects! Little did we know, the evil genius had passed herself as the cleaning lady to wreak vengeance on successful rivals.

Even if the information were all of the right kind, Aline van Duyn warns us "too much [of it] can result in people making the wrong decisions". I would add that the time taken to make a choice, right or wrong, tends to increase exponentially with the number of underlying criteria. The abundance of bad data and the difficulty to reach rational conclusions are why "mechanisms" were worth a Nobel prize in economics. To quote Tim Harford's obituary of the late Leonid Hurwicz (*****), "a mechanism [is] a kind of "message center", where market participants would submit information, true of false, and a pre-specified set of rules would adjudicate the result".

Mechanisms need not destroy individual freedom. In Tim Harford's example, an eBay auction, every bidder is free to act as he or she deems best. But, while some US Supreme Court observers are wont to model its decisions with a simple set of rules, it is best to avoid applying mechanisms to reach legal decisions. Even market mechanisms need revisiting in the light of the power derived by the entity controlling the "message center".

In a classical mechanism, its administrator can profit handsomely from violating the participants' privacy. A successful mechanism creates huge amounts of data and there's gold in them thar hills. Read Miguel Helft on how Google let marketers ever deeper into its archive of searches (******). May Hurwicz lead us beyond communism and capitalism. But let us face reality and prepare for The Resistible Rise of Real Big Brother.

If we do not require our economic mechanisms to be confidential by design, if we do not limit the roles of the rulers, expect the beginning of the end

Philippe Coueignoux

• (*) ............. Hungry Planet, by Harry Eyres (Financial Times) - August 2, 2008
• (**) ........... Information is not always of value - ask the usual suspects, by Aline van Duyn (Financial Times) - July 31, 2008
• (***) ......... Scientist's Suicide Is Linked to Anthrax Inquiry, by Scott Shane and Eric Lichtblau (New-York Times) - August 2, 2008
• (****) ....... F.B.I. Presents Anthrax Case, Saying Scientist Acted Alone, by Scott Shane and Eric Lichtblau (New-York Times) - August 7, 2008
• (*****) .... Nobel-winning economist who put a premium on truth, by Tim Harford (Financial Times) - July 19, 2008
• (******) .. Google's New Tool Is Meant for Marketers, by Miguel Helft (New-York Times) - August 6, 2008