No one can imagine what work will be like in the 2030s.
getty
In February 1966, Time magazine published an essay predicting that in the near future, “computers will replace not only manual workers but also secretaries and most middle managers. The remaining executives will be responsible for important decisions and long-term policy. By some estimates, only 10 percent of the population will be working, and the rest will have to be paid, in effect, even when they are off.”
Well, something interesting happened on the way to the future. Computers haven’t eliminated jobs. Rather, computers have created new kinds of jobs and opportunities, and many of them. Ask someone who has been working on their computer past midnight to complete their latest project effort. Or maybe you’re looking through a report or spreadsheet while you’re on a plane. Computers are keeping everyone busy and moving far beyond the 9-to-5 work schedule.
Lately, there has been renewed talk about replacing everyone with computers, especially as the wave of artificial intelligence hits. But data and common sense tell a very different story, more akin to the world that evolved after the February 1966 essay.
As an example, David George, a partner at venture capitalist firm Andreessen Horwitz, expects that AI will gradually reveal new roles and opportunities that are beyond our view. Current working conditions are not a “freeze frame” in a movie, he wrote in a recent note. Rather, it is a world that is constantly evolving. Innovation and technological progress will never occur by simply replacing existing practices one-for-one. The assumption that AI will simply replace human jobs on a one-to-one basis is “not only a huge failure of imagination, but a failure of basic observation,” says George.
He backs up his claims with data-rich reports from the U.S. Census Bureau, U.S. Bureau of Labor Statistics, Deutsche Bank, Goldman Sachs Research, Carlyle Analysis, Yale Budget Lab, and others. None show evidence of an impending “apocalypse” of work.
Yes, AI will change tasks and jobs – and that is to be expected, he continues. “As always when a revolutionary technology is unleashed, the shape of the labor market changes. But the claim that AI will create permanent unemployment across the economy is useless marketing, bad economics, and worse history.”
Keep in mind that the author of these statements, David George, is a partner of one of the world’s leading investors in what is now an AI-driven venture, so there is a certain openness to new innovations that involve such a role. Still, his perspective provides a front-line view of the economy for the next decade.
A similar theme was discussed by former Tesla CEO John O’Neill in an interview I conducted last month, but so far only the “first-order effects” of AI have been confirmed. The second- and third-order effects of AI are not yet well understood, but will inevitably become known as people discover new uses and uses that can change industries.
The proof is in the data. Currently, “there is no indication that exposure, automation, and augmentation measures are associated with changes in employment or unemployment,” George said. At the same time, the available data on AI and employment is incomplete.
The Yale Institute for Budget Research, cited by George, said that predictions about how AI will affect the current labor market are “largely speculative.” So far, the data mainly reflects stability and not major disruptions at the economy-wide level. The composition of AI employment appears to be changing more rapidly than in the past, although not significantly so. ”
George attributes much of today’s hysteria about job losses to the “lump of labor” fallacy. This argues that the current level of work remains the same, a zero-sum trade-off, and that tasks can simply be swapped between humans and AI agents. “It assumes zero-sum competition between existing workers and someone or something else who might do the same job, whether it’s another worker, a machine, or in this case, an AI.”
Rather, he points out, increased productivity “should increase the demand for labor because it makes labor more valuable.” Instead of layoffs, there will be more productivity and innovation, resulting in higher demand for talent. For evidence of increased productivity, look at the increase in hiring of both software engineers and product managers, he explains. “If AI were to replace one-to-one thinking, you might plausibly expect that we would need fewer engineers to be PMs, or you could argue that we would need fewer engineers to be PMs. But that’s not what we’re seeing. The key is that people will be able to do more work, so we see demand for both continuing to recover.”
“Automation removes repetitive layers and brings human work up the stack,” he explains. “The reason is simple: humans want to expand. When you move down one tier of scarcity, people move on to the next tier. When food becomes cheaper, they spend more on housing, health, education, travel, entertainment, convenience, pets, safety, beauty, and longevity.”
The same thing is happening in the labor market, he continued. “Human ambition never stops, so new jobs are created, and conquering old frontiers reveals new frontiers to conquer. New business formation is already exploding, and the correlation with AI adoption is pretty good.”
