18 November 2023

The Jobless AI Future Is Still a Long Way Off

MICHAEL R. STRAIN

WASHINGTON, DC – In a recent discussion with British Prime Minister Rishi Sunak, tech billionaire Elon Musk prophesied that there “will come a point where no job is needed,” owing to advances in artificial intelligence. “You can have a job if you want a job,” continued the CEO of Tesla and SpaceX, “but AI will be able to do everything.”

The future is long, and Musk did not specify when exactly this point will come. But, at least for the next several decades, the odds of AI causing a jobs apocalypse are vanishingly small.

Fear of technological unemployment is nothing new. In the early nineteenth century, a group of English textile workers known as the Luddites smashed labor-saving machinery to prevent its use. Yet even though technology has leapt forward in the two centuries since then, businesses continue to employ workers.

Much of the concern about technological advances eliminating the need for human workers is rooted in a zero-sum mentality that fundamentally misunderstands how economies evolve. Yes, new technologies will be able to perform some tasks relatively better and at lower cost than humans. Yes, this will lead businesses to use technology, not workers, for those tasks. But the process of creative destruction creates as well as destroys.

New technology will make many workers more productive and thus of greater value to firms, which will compete more aggressively for them in the labor market, driving up their wages and incomes. Higher incomes will increase overall demand for goods and services in the economy, which in turn will increase the need for workers. This dynamic process allows an economy to avoid higher structural unemployment. Moreover, new technology creates novel goods and services, which also increases demand for workers.

This is not just a theory. Consider the remarkable advances in information and communications technology and robotics over the past five decades. These breakthroughs have had profound effects on the labor market – for example, by substantially reducing the employment share of manufacturing and clerical occupations – and the broader economy. But it has not become more difficult for workers to find jobs. There has not been an upward trend in the unemployment rate.

Looking ahead to the next several decades, my main concern is not too many workers, but too few. Falling fertility rates and rapid population aging will reduce the rate of workforce growth in the United States and across much of the developed world. These countries could rely on larger inflows of immigrants to make up for the shortfall, but the political winds are currently blowing in the opposite direction.

In the US, the nonpartisan Congressional Budget Office forecasts population growth of 0.3% per year over the next three decades, or roughly one-third the pace from 1983-2022. In this environment, workers who want jobs should be able to find them.

Of course, advances in AI will certainly be disruptive, but not by eliminating – or even substantially reducing – the need for workers. Instead, AI will change what many workers do. Again, this is not new. While technological unemployment may have failed to materialize over previous decades, the structure and nature of employment were utterly transformed. The MIT economist David Autor and coauthors found that the majority of current employment is in job specialties introduced after 1940.

The further into the future we look, the harder it is to rule out the possibility that Musk’s prediction could come true. But a world in which AI eventually replaces all human workers would look a lot different from ours. While one of today’s fundamental economic problems is how to make the best use of scarce resources, Musk’s future is one of abundance, in which technology meets all our needs and inequality as we currently understand it no longer exists. Why accumulate wealth in a world of abundance? On the other hand, such a world could also exacerbate inequality, particularly if a relatively small number of people own the machines that are generating all the income.

How would humans engage with this world? I receive a great deal of fulfillment and satisfaction from my professional activities, but the primary reason I get up and go to work each day is to provide for my family. Without the need to put food on the table and money in the retirement accounts, would we get up each day with the goal of bettering ourselves and our communities? Or would we fall to the darker angels of our nature, with unoccupied time leading to boredom and dystopia?

There are no easy answers to such questions. Fortunately, it’s more likely than not that we won’t have to answer them. Technological change has not eliminated the need for human workers in the past, and it likely won’t in the future – at least not within any timeframe that is relevant to workers and policymakers today.

No comments: