October 19, 2013


Average Is Over--But the American Dream Lives On (ANDREW LEWIS, 10/16/13, American Interest)

One of Cowen's most fascinating projections concerns the proliferation of machine intelligence and its effects on labor market conditions. He sees the mechanization of chess as a small-scale version of how many industries will evolve, following this sequence:

Human is the expert; computer adds little to current product.
Experts continue programming machine, strengthening its ability.
Experts supplement computer with minimal input (correcting obvious miscalculations).
Machine becomes expert; human adds little.
Computers take place of human (middle-wage jobs); only those who can add value to the computer stay on the job.

Cowen's analogizing of the progress of chess programs to broader societal trends is worth delving into with a little more detail. For centuries, chess was a game dominated by human art and skill, with Grandmasters as the foremost artists. In 1990, when a team of IBM programmers (and Carnegie Mellon University graduates) set out to create a machine that could beat these human experts, they were largely scoffed at. But in 1996 the resulting program, Deep Blue, won its first game against the reigning world champion Garry Kasparov (but lost the overall match); in 1997, it defeated Kasparov in a traditional six-game match.

Today, Cowen tells us, an average chess-playing program, running on virtually any piece of consumer hardware, is easily capable of outplaying any human. Many people who compete at the highest levels of chess play what is called freestyle chess, where teams include a computer and a human counterpart. Excelling at freestyle does not require profound skill in chess per se, but rather expertise in working with the computer. The best players are the ones who recognize their limitations and are willing to accept the advice of the computer; those who win most are the ones who design or run the best programs. Cowen predicts that this process will be repeated across many different industries and arenas of human endeavor.

If this process holds, it's not difficult to see why incomes will become increasingly polarized. The top end of the income distribution--which he envisions as the 15 percent, rather than the 1 percent--will be comprised of those who are truly talented or creative in their ability to work with technology. [...]

The question that logically follows from worsening income inequality is how it will affect America's social fabric. Plenty of doomsayers predict that it will lead to a revolution--that the left-behinds will conspire against the new high-earners. But if everyone has the same opportunity to succeed, then how will they feel slighted by the system? Cowen calls these theories of the revolutionary consequences of income inequality "some of the least thought-out and least well-supported arguments with wide currency."

This is largely consistent with his view on the history of income inequality in the United States. He recognizes that the trend is deeply disconcerting to many Americans but also believes it is not the best measure of social inequality. Access to food, modern medicine, and the internet are just a few measures one could cite to show that the average person is better off now than ever before. In a 2011 article for The American Interest, Cowen wrote, "By broad historical standards, what I share with Bill Gates is far more significant than what I don't share with him." Indeed, the advancements of modern society have allowed more Americans to enjoy higher standards of living than at any other time in our nation's history.

A report published in 1997 by the Dallas Federal Reserve Bank, Time Well Spent, substantiates this point. While income disparity has grown, the magnitude of the change does not nearly match that of the rising standard of living. The report examines the cost of goods based on the minutes of work needed, on average, to buy them. It finds that in 1919 it took the average American eighty minutes of work in order to buy a dozen eggs. Today it takes him five minutes.

If this is the case--and by the most advanced methodological standards, it is--then income inequality might not be as devastating an issue as many believe. That's not at all to say that inequality is unimportant, but rather that income inequality is not the best gauge of societal equity. Perhaps opportunity equality should be what we strive for. The book makes a case for this suggestion.  

...will only make the 85% rest of us even more affluent.
Posted by at October 19, 2013 7:20 PM

blog comments powered by Disqus