Stories

The future began four decades ago

Illustration: Aïda Amer/Axios

If you're an optimist about the robotic future, you likely hear talk that we're all going to lose our jobs or suffer a big pay cut, and tell friends to relax — the new technology revolution is going to turn out like all the others since the dawn of the Industrial Age.

But if history is your best hope, you should probably think again: The pessimists have a strong case.

The big picture: For the last half-dozen years and longer, one of the most hotly debated questions on the planet has been the fate of employment in a future age of AI and robots. It is a central issue in current U.S. and European political campaigns — and was a core factor in President Trump's 2016 election victory.

But Carl Frey, author of "The Technology Trap: Capital, Labor, and Power in the Age of Automation," suggests that the "future" is already here — and has been for three decades. The only question is how much longer it will last.

  • Since 1979, 6.6 million people have lost their middle-income jobs in U.S. manufacturing, a third of the total number in the sector, on top of work eliminated in other skilled and semi-skilled professions.
  • Many took pay cuts, other lower-paid work, or simply left the labor force.
  • These factors, resulting in blight, a hollowing out of the middle class, and politically destabilizing anger, replicate the early Industrial Age.

"If we extrapolate from the past, we shouldn't be assured," Frey tells Axios.

For a look at how much longer this could go on, consider the genesis of the Industrial Age: It unofficially began about 250 years ago, when James Watt invented the first workable steam engine and, in what Frey calls the "classic" Industrial Age, ran seven decades through 1840.

A key thread running through both the classic age and now is the nature of new commercial technology: In both periods, it generally replaced workers, rather than helping them work better. (Example of a helpful technology: the telescope, which put no one out of work. Example of a worker-replacing one: the automatic elevator.)

  • In the steam engine's first factory use, the owners of power looms and spinners hired hundreds of thousands of British children, bypassing adults, since they could pay youths a fraction of the wages and hear fewer complaints about the conditions.
  • Worker wages stagnated, growing by just 12% from 1780 to 1840, while output per worker soared 46%. The top 5% of the population captured 37% of total income by 1867, up 75% from 21% in 1759, a mark of sharply growing inequality.
  • Total company profits and the labor share of the total financial pie only began to grow in tandem around 1840 — about 70 years after Watt's patent.

By that time, most of the original rural artisans and their dependent kin were long dead.

Which is to say that, if this time is turning out like the classic age, we may be in for three more decades of sluggish or flat wage and productivity growth and a shrinking middle class, conditions that are behind much of the populist anger in the U.S. and elsewhere.

What's next: Frey argues that many of the new technology prototypes in the pipeline look like they, too, will be the worker replacement type:

  • Among them: AmazonGo cashierless stores (a threat to the 3.5 million U.S. cashiers); Google's AI Assistant (potentially imperiling 2.2 million call center jobs); and autonomous cars (a future threat to 3.5 million truck, bus and taxi drivers).

The bottom line: Prior to the Industrial Revolution, European royalty routinely killed blockbuster inventions out of fear of destabilizing riots. In retrospect, public officials traded off eventual societal advancement for holding onto their own jobs. This is what Frey calls "the technology trap."

  • Frey argues that we are at severe risk of a new technology trap. Already, the entire post-war Western system is in jeopardy because "governments chose to overlook the costs of globalization and focus on the benefits."
  • “Governments must avoid making the same mistake with automation.”

Go deeper: The problem with automation