Chances are you're not too worried about a robot taking your job. After all, when we picture out-of-control automation, we imagine that the blue-collar folks who work with their hands in the factories or checkout counters will be pushed aside by a collection of chips, software, and servo-motors.
But don't be smug, techies. An intriguing new research paper from two major universities posits that even software developers, the very people who program those machines, could face a future in which their skills are no longer needed and their incomes drop precipitously as smart machines reduce the need for human-produced software to mere maintenance operations.
If that happens, wages will drop, demand will decline, investors will make smaller returns, and the capital needed for future tech booms won't accumulate. "In other words, technological progress can be immiserating," say the authors of "Robots are us: Some economics of human replacement."
The researchers -- Boston University's Seth Benzell, Laurence Kotlikoff, and Guillermo LaGarda, and Columbia University's Jeffrey Sachs -- aren't predicting some silly, Terminator-like robot apocalypse.
What they are saying is that our economy is entering a new type of boom-and-bust cycle that accelerates the production of new products and new code so rapidly that supply outstrips demand. The solution to that shortage will be to figure out how not to need those hard-to-find human experts. In fact, it's already happening in some areas.
After the developer boom comes the developer bust
In the past, technological changes have generally spared the most highly skilled members of the workforce. But as we slide down the backside of the cycle, the pain will be shared quite broadly. "The long run in such a case is no techno-utopia," the authors say. In that future, "code is abundant. But capital is dear. And yes, everyone is fully employed. But no one is earning very much."
In the beginning of the cycle, demand for people who can write code soars, so wages go up, say the researchers.
That's exactly the situation we're in now, when demand for software developers is at an all-time high. In the last three months, for example, employers have posted more than 100,000 openings for software engineers and developers on the Glassdoor site, though not all the postings are still current.
Wages for tech workers, including highly paid executives, are double -- on average $100,400 a year versus $49,600 -- that of workers in the rest of the economy, according to a recent study by CompTIA, an industry group that issues certifications and follows tech-job trends.
So far, so good -- but over time, more and more code is produced. Some of it enables smart machines to actually learn to perform new tasks or become so adept at their tasks that there's no need to spend the money to make them even smarter.
"Take Junior, the reigning World Computer Chess Champion. Junior can beat every current and, possibly, every future human on the planet. Consequently, his old code has largely put new chess programmers out of business," say the researchers.
In the case of chess, that's a tiny number of techies who are displaced, of course, but it illustrates how the cycle could shift to the downside. Eventually the demand for new code shrinks, coders lose their jobs or are reduced to performing routine maintenance, updates, and repairs.
Job-stealing robots are already here
The "Robots are us" title of that research paper sounds extreme, but there already signs that even highly skilled work may be ceded to machines -- and not only for chess programming.
Consider a company call Narrative Science founded by researchers from Northwestern University. They discovered that they could develop software that can write basic sports stories. I've read a few of those stories, and although they're not brilliant, they're serviceable -- probably good enough to cover small-town sports.
In 2013, Narrative Science's software produced nearly 400,000 accounts of Little League games. Last year it was expected to crank out 1.5 million, according to an interesting piece in Wired. The software also produces stories on corporate earnings -- usually quite basic, but good enough to give investors a quick read on the market and do it much faster than even an experienced business writer can.
There's historical precedent for technology displacing highly skilled workers as well. Like a modern entrepreneur, Gutenberg began with an inspiration (movable type), raised venture capital, built a team, and eventually disrupted the market for books produced one at a time by scribes.
Prepare yourself for "do you want fries with that?"
In the world that's emerging, we're talking huge numbers of people who will be affected: Last year, there were 670,800 jobs for software applications developers, plus an additional 387,400 for in systems software. Tech employment as a whole was about 6.5 million, according to U.S. Labor Department statistics cited in the CompTIA report.
Naturally, wages will drop, and some high-tech workers will get pushed into other jobs, particularly in the service sector. Former coders won't make $100,000 teaching math, slinging hamburgers, or driving for Uber. The influx of smart new employees will push wages down even further.
Aside from the personal misery of people living on reduced incomes, the entire economy will suffer as people have less money to spend, which reduces the demand for goods and services. Less demand means less capital accumulates, so investment that would have funded the next boom stagnates -- which is what the researchers mean when they say "when smart machines replace people, they eventually bite the hands of those that finance them."
The researchers concede that the rise of the smart machine may have an outcome much better than the worst case they suggest. Society could take steps that could minimize that impact.
One possible remedy is a tax (really, a forced rainy-day savings) on workers who benefit from a technological breakthrough during a boom. That money could be set aside and used to pay a stipend during a bust, keeping them afloat and ensuring a steady supply of capital.
"Absent appropriate fiscal policy that redistributes from winners to losers, smart machines can mean long-term misery for all," the authors conclude.
Think of that the next time you program a smart machine.