Demon of Mechanism

 Well, hello there, Zeitgeist—fancy meeting you here! No sooner had I turned in my recent post about AI job loss, than an op-ed appears in the New York Times suggesting that it is already upon us. The piece, written by a LinkedIn executive, argues that entry level jobs in white collar and professional services may already be disappearing into the maw of automation. 

In short, we are back in the same situation that Thomas Carlyle diagnosed two hundred years ago—as I discussed a few posts back: "The huge demon of Mechanism smokes and thunders, changing his shape like a very Proteus; and infallibly at every change of shape, oversetting whole multitudes of workmen [...] hurling them asunder, this way and that[.]"

Such has always been the story with technological displacement—from the Silesian weavers Heine and Hauptmann wrote about in the 19th century up to the present. 

But, so too, the defenders of automation can always argue (somewhat accurately) that each new wave of technological innovation ultimately creates more jobs than it displaces. Carlyle wrote that this is what the defenders of the steam engine—the lecturers in political economy—told people in his era as well: "the Steamengine always in the long-run creates additional work."

The New York Times piece nods to the same idea: "Eventually, A.I. will create plenty of jobs," the author writes; "The World Economic Forum predicts that number could be as high as 78 million more jobs, even after predicted job losses."

But—the author then acknowledges—this new "eventual" job creation may prove cold comfort for the people who are displaced in the immediate term. It may come too late for them. If they miss the chance to get on the career ladder at the bottom rung early on, right after graduating from college, they may never have a chance to do it later. 

The Times opinion author notes: "getting a late start can slow down workers’ careers for decades. The Center for American Progress found that young adults who experience six months of unemployment at age 22 can expect to earn approximately $22,000 less over the next decade."

Indeed, the problem with automation has never been that it reduces jobs in aggregate. I'm perfectly willing to believe the economic theorists who reassure us that human wants are infinite, technological change makes firms more efficient, and so—in the long run—increased productivity will always create several new jobs for each one it eliminates. 

But the problem has always been how to match these new jobs to the actual human beings who lose their livelihoods in the meantime—and who, like all of us, are not infinitely flexible, infinitely mobile. The pace of technological innovation is truly "protean"—as Carlyle puts it. But human beings, brains, and skills are not nearly as protean—not changeable fast enough to match it. 

Carlyle notes—in response to the people who say "the Steamengine always in the long-run creates additional work"—that, well, this may be true—but where does it create that work? Well: "somewhere in Europe, Asia, Africa or America, doubt it not," writes Carlyle. 

His point being: how are the displaced workers supposed to find it? How are they to know where these new jobs are going to be located? And how are they to get there once they identify them? 

As Ernest Gellner notes in his Nations and Nationalism, modern societies have come to expect a level of constant economic growth that requires continual technological change. And this, in turn, requires individuals to be ever more "protean" within their lifespans—in order to keep up with the protean pace of economic innovation. 

This, in Gellner's view, is responsible for many of the pathologies of our time—nationalism especially. People who are expected to be ever-protean in their vocation and location—willing to move "somewhere in Europe, Asia, Africa or America," to chase all those new jobs that are "eventually" being created—they cling for something solid on which to ground their collective identity. 

Hence arises the disease of nationalism—the in-group/out-group thinking so observable in our times. Because people no longer have real community or stability to ground them, they seek it under the fictive guise of the state, the "nation," the "race," and other dangerous chimeras. 

We can expect already, then, that our AI future, even if it comes with new jobs, will not resolve the social pathologies of our time. The desire for economic growth will keep demanding protean technological change, which will require us to be ever more protean in our life curricula. And people will then resort to some very bad ideas to generate a false sense of stability. 

What is the alternative, though? Could we perhaps accept a bit less economic growth in exchange for greater prosperity? Some societies have made this choice in the past. And, of course, it comes with trade-offs. I would not deny them. But they are at least worth considering. 

Perhaps we should actually consider the collective option to deliberately slow down the pace of technological change through regulation. It's not an absurd idea. As Robert Frost once argued, we have had to curb ambition in other domains—political power, the concentration of wealth, etc.; so why not in this one? As Frost put it: 

Even while we talk, some chemist at Columbia
Is stealthily contriving wool from jute
That when let loose upon the grazing world
Will put ten thousand farmers out of sheep. 

[...]
None should be as ingenious as he could,
Not if I had my say. Bounds should be set
To ingenuity for being so cruel
In bringing change unheralded[.]