Government by Google

Chris Anderson, the editor of Wired, has a short piece there called, ambitiously, “The End of Theory.” (Hat tip: On the Media.) It argues that the traditional way of doing science – construct a hypothesis, collect data to see if the hypothesis is falsified, repeat as necessary until the hypothesis becomes a generally accepted “theory” – is about to be made obsolete by the Google way:

But faced with massive data, this approach to science — hypothesize, model, test — is becoming obsolete. Consider physics: Newtonian models were crude approximations of the truth (wrong at the atomic level, but still useful). A hundred years ago, statistically based quantum mechanics offered a better picture — but quantum mechanics is yet another model, and as such it, too, is flawed, no doubt a caricature of a more complex underlying reality. The reason physics has drifted into theoretical speculation about n-dimensional grand unified models over the past few decades (the "beautiful story" phase of a discipline starved of data) is that we don't know how to run the experiments that would falsify the hypotheses — the energies are too high, the accelerators too expensive, and so on.

Now biology is heading in the same direction. The models we were taught in school about "dominant" and "recessive" genes steering a strictly Mendelian process have turned out to be an even greater simplification of reality than Newton's laws. The discovery of gene-protein interactions and other aspects of epigenetics has challenged the view of DNA as destiny and even introduced evidence that environment can influence inheritable traits, something once considered a genetic impossibility.

In short, the more we learn about biology, the further we find ourselves from a model that can explain it.

Just as Google ranks web pages’ importance in a given search not by their content but by the degree to which other web pages refer to them, science may, instead of starting with a story and then seeing whether it plays well out amongst the data, be shifting to a model where scientists look at the data first and use the correlations themselves as the finding of interest.

Now, one has to be cautious about such revolutionary proclamations; several years ago similar things were said about Gary Wolfram’s arguments about the relation between computation and science. But, like Mr. Wolfram, Mr. Anderson is a very smart man. His most recent book, The Long Tail, arguing that the digitization of knowledge opens up access to small-market cultural forms that were previously unprofitable to distribute, decisively changed the way I thought about globalization and culture in The Rise of the Anti-Corporate Movement.

But whether science is about to change or not, translated into moral/political philosophy the Google approach is disastrous. Governments – particularly the American one – start from philosophical premises and then design the government accordingly. It is unacceptable in the traditional American way of thinking to start with desired results, perhaps driven themselves by a crude utilitarianism, and then to increase government power accordingly. There are both philosophical and practical/utilitarian reasons to be concerned.

Suppose, for example, that someone discovers a high rate of correlation between legal gun sales and violent crime, thus leading to the conclusion that ending the sales drives down the crime. The first objection is philosophical – there is a right to self-defense that morally precedes the right of the state to limit gun trafficking in its assertion of the public interest. The second objection is that the political process needs a hypothesis, and may get it wrong – legal gun sales may be high precisely because crime was already high. So could we trust the political process to make use of the available information in the most socially responsible way? Unlikely. Government scientists, and the legislators enabled by their findings, will naturally tend toward explanations that lead to more control and regulation – to, at best, try and solve the problems that we can see, heedless of the ones we can’t see and the ones we create out of our solutions. At worst, the exploration for correlation is used simply to enhance the ability of the state to control us. (I leave it to the reader to decide whether the NASA scientist James Hansen’s hysterical assertion that oil company executives should be put on trial for crimes against humanity is an example of the former or the latter.)

Al Gore once told The New Yorker that he saw the U.S. as a giant parallel computing machine, where ordinary people identified problems through their scientific work, political activity, etc., and then funneled the results up to Washington, which magically solved the problems discovered. This kind of thinking, presumably, is how he assumes that we can (or doesn't care that we can't) with relative ease and little corruption or loss of freedom shift the entire electricity grid to non-carbon sources in ten years. Google famously claims that its technology is profitable without “doing evil.” Government by Google, alas, would not end nearly as well.