Recent debates about the impact of artificial intelligence (AI) on employment have oscillated between the poles of apocalypse and utopia. In an apocalyptic scenario, inequality will be greatly exacerbated as AI takes over the majority of all jobs and small capital-owning classes capture the production surplus previously shared with human workers. .
Oddly enough, the utopian scenario is the same, except that the very rich are forced to share their winnings with others through a universal basic income. Everyone will enjoy abundance and freedom, and ultimately achieve Marx's vision of communism. The common premise of both scenarios is that AI will increase productivity so much that high-paid doctors, software programmers, and even airline pilots will have to work alongside truck drivers and cashiers. It means that it must be done.
Click here to follow our WhatsApp channel
Both dystopian and utopian outcomes make AI a political problem. It's about whether marginalized people can force the AI bigwigs to share their wealth. There are reasons to be optimistic. First, the benefits from AI under this scenario are so enormous that the super-rich may not mind parting with a little money to appease their conscience or buy social peace. Second, the growing number of marginalized people will include highly educated, politically engaged people who join traditionally marginalized groups in calling for redistribution.
But there are also deeper questions. How do people react psychologically and politically to the realization that they can no longer contribute to society by engaging in paid work? Male labor force participation has already declined significantly since the 1940s. Although women only entered the labor force in large numbers in the 1970s and 1980s, their labor force participation rates have also begun to decline. This may reflect the tendency for those at the bottom to lose the ability to convert their labor into compensable value as technology advances. AI could accelerate this trend and attack those in the middle and even at the top.
If social surplus is widely shared, one might ask, “Who cares?” In the past, upper-class people avoided working and looked down on those who did. They appear to have been quite content with their situation, filling their time with hunting, literary activities, parties, political activities, and hobbies. Modern economists tend to think of work in the same way. It is simply a cost ('c') of putting people to work that must be offset by higher wages ('w'). Social welfare is maximized through consumption, not through the acquisition of “good jobs.” If this is correct, we can compensate people who have lost their jobs simply by giving them money.
Perhaps human psychology is malleable enough to see a world with plenty of jobs and few or no jobs as a boon rather than an apocalypse. If the aristocracy of the past, the retirees of today, and the children of all eras can fill their time with play, hobbies, and parties, perhaps the rest of us can too. However, research shows that the psychological toll of unemployment is significant. Even after controlling for income, unemployment is associated with depression, alcoholism, anxiety, social withdrawal, worse child outcomes, and even early death. Recent literature on “deaths of despair” provides evidence that unemployment is associated with increased risk of suicide and overdose.
So the long-term challenge posed by AI may be less about how to redistribute wealth and more about how to maintain jobs in a world that no longer values human labor. I don't know. One proposal is to tax AI more than labor. Meanwhile, another proposal, recently put forward by MIT economist David Orter, is to leverage government resources to shape the development of AI so that it complements, rather than replaces, human labor.
Neither idea is promising. If the most optimistic predictions about future productivity gains from AI are accurate, very high taxes will be needed to have any impact. Even if taxes and subsidies could keep jobs that are less valuable than their AI replacements alive, it would simply postpone the day of reckoning. People who derive self-esteem from their work do so because they believe that society values their work. Once it becomes clear that their work can be done better and cheaper by machines, they will no longer be able to maintain the illusion that their work is important. Even if humans are able to adapt to a life of leisure in the long term, the most optimistic predictions of AI productivity could result in massive short-term disruptions to the labor market.
The author is a professor at the University of Chicago Law School ©Project Syndicate, 2024
