Free beta access. 50 free AI tokens each week. Optional top-ups available. See what is included
Opening Word document…

Rapid automation causes inequality---that triggers global wars from authoritarian populism

Aaron James 20. PhD from Harvard and is professor and chair of philosophy at the University of California, Irvine, September 2020. “Planning for Mass Unemployment: Precautionary Basic Income,” Ethics of Artificial Intelligence, https://academic.oup.com/book/33540/chapter/287905325

Rapid automation leads to unemployment on a mass scale—not as a temporary shock, but as a new normal. Such was once the stuff of dystopian fiction. Now it could be our future, especially given amazing recent advances in machine learning and robotics. Not this year or this decade, necessarily. But down the road. After all, it’s only dawn in the new machine age.

Keynes warned in 1930 of a “new disease” that he called “technological unemployment,” by which he meant “unemployment due to our discovery of means of economizing the use of labor outrunning the pace at which we can find new uses for labor.” Alas, his fears didn’t materialize in much of the twentieth century. But by the close of the millennium there were signs that it’s no iron law of economics that jobs destroyed will invariably bring new opportunities for human work. As Erik Brynjolfsson and Andrew McAffee explain, citing trends in computing capacity, digitization, and the power in recombining older technologies, “Employment grew alongside productivity up until the end of the twentieth century. [But then] job growth decoupled from productivity in the late 1990s.” Even as productivity rose, employment fell off. The long trend didn’t hold. And so, in due course at least, artificial intelligence (AI) really might cause lasting structural unemployment on a mass scale.

Imagine the scene: jobs are steadily automated, year after year. In the old days, for every job destroyed, a new one was eventually created, leaving total employment more or less unchanged. Now deep-learning machines, aided by clever entrepreneurs, race ahead and do the new tasks as well. The routine tasks, both manual and cognitive, have been mostly automated away. The best workers are those that work best with machines, in some cases besting both person and machine alike. A fair number still do what computers still can’t (as of yet). But many workers—maybe most workers—are left to do the relative few manual or service jobs left. Wages have long stagnated or slipped. Joblessness used to be temporary for most; it wasn’t fun, but the hand of technological fate was relatively arbitrary, spreading disruptions relatively evenly across the population, so that no one took the losses over and over again and nearly everyone shared the rising average standard of living. Now the lag between jobs grows ever longer. Many people—most people, even in their prime years—simply can’t find tolerable work and stop looking. You wouldn’t bet on mortgages, vacations, and college, rising wages over a lifetime, or better prospects for the children. It’s enough just to try to get by. Reared in the Protestant Work Ethic, people are frustrated, distrustful, and angry, nursing contempt for the institutions that promised steady rewards for “hard work” and then failed to deliver decade after decade. With people driven ever farther apart by social media, and public discourse long since polluted, there’s ever less scope for addressing common problems. Democracy, after its dramatic rise in the twentieth century, has become but a shell for corrupt authoritarianism, as hungry political entrepreneurs conspicuously feast on the unraveling.

It’s an awful scenario. Relentless wage stagnation, the decline of trust, the hollowing out of democracy—that’s all bad enough by itself, even without mass unemployment. Add millions of anxious souls out of work, and it’s a terrible state of society—but now seemingly possible.

We certainly don’t know work will be automated faster than reemployment any time soon. The economic historian Robert Gordon even calls the prospect “unlikely” (of which more later). Yet he has no doubt that a different slow-moving automation crisis is already upon us: the quality of the jobs there are has markedly declined. Among the many factors (including trade, the decline of unions) that have brought stagnating wages since the mid-1970s, automation is an important part of the story. He writes, “The problem created by the computer age is not mass unemployment but the gradual disappearance of good, steady, middle-level jobs that have been lost not just to robots and algorithms but to globalization and outsourcing to other countries, together with the concentration of job growth in routine manual jobs that offer relatively low wages.”

If nothing else, the idea that a mass unemployment crisis could eventually have awful social and political consequences, such as the rise of authoritarianism and hollowing out of democracy, shouldn’t seem fanciful. The long trend of bad jobs and stagnant wages has arguably already had social and political consequences, in part for aiding the rise of authoritarianism and the hollowing out of democracy. Gordon thinks “this time is not different” as regards mass unemployment. But we’re already in a sort of employment crisis that may warrant a precautionary basic income—precautionary, because things might get even worse.

And the different employment issues aren’t in competition. The onset of mass technological unemployment would vastly worsen the bad jobs crisis we’re in. We don’t know we are proverbial frogs in boiling water, but caution seems to be in order, if only because of how bad things could become, if not next year or decade, then in the further future. By now it is not hard to imagine an upheaval so radical as to frighten anyone conservative enough to value advanced peaceable civilization as we have come to know it (including those on the left who value what is by now a long history of progressive achievements). The risks to democracy alone are already grave enough. Recall that the historical rise of democracy is associated with the “long peace,” a period of markedly reduced war and violence amid unprecedented prosperity. With the sunset of democracy, that trend might end.

Sign in to recut, update, or save this card.