“Artificial Intelligence Will Destroy Jobs, Impoverish the Masses, and Enrich the Elite”
The claim that labor-saving machinery makes the average citizen poorer by eliminating their jobs is older than the Industrial Revolution. Daron Acemoglu and James Robinson describe how all the way back in 1589, William Lee created a stocking frame knitting machine that would produce far more textiles than could be made by hand. When he went to London to ask Queen Elizabeth for a patent, she denied his patent and is recorded as saying “Consider thou what the invention could do to my poor subjects. It would assuredly bring them to ruin by depriving them of employment, thus making them beggars” (p. 182-183). The Queen did not realize that increasing productivity is what improves the standard of living of the average citizen. Frédéric Bastiat satirized the claims made by those opposing automation in 1850, writing "'A curse on machines! Every year, their increasing power relegates millions of workmen to pauperism, by depriving them of work, and therefore of wages and bread. A curse on machines!'" (p. 29). One hundred and seventy-five years later, we have more labor-saving technology and yet have more jobs and abundance than in Bastiat's day.
Walter E. Williams writes that the reason we are wealthier today than they were in Bastiat's day is because we use more labor-saving technology, writing that our "…higher standards of living is a result of capital being used to replace labor, thereby freeing up that labor to undertake other tasks" (p. 55). Today, people argue that automation in the form of artificial intelligence poses a unique problem. In the past, machines still needed to be operated by a human, but AI can think and act for itself, which will finally bring about the unemployment apocalypse. Again, these same arguments about automation were made decades ago. For example, Henry Hazlitt writes that in the early 1960s "Automation was discussed as if it were something entirely new in the world. It was in fact merely a new name for continued technological advance and further progress in labor-saving equipment" (Hazlitt, One Lesson, p. 53).
Williams argues that the reason the feared outcomes of automation don't come to fruition on a large scale is that they assume there is a fixed number of jobs and there will be no use for the labor that has been replaced. Humans needs, however, are insatiable and there will always be use for labor to achieve those needs. Saifedean Ammous argues that no matter how much abundance we have in our society, humans can always imagine better alternatives and will labor to achieve them. He writes that "So long as humans are alive and need to decide what to do with their time, the economic problem exists, and humans attempt to solve it by working... There will never be complete satisfaction because human reason will always foresee a better possibility and work toward it" (p. 75).
I wonder if the economic principle of comparative advantage might be applicable in explaining why AI won’t eliminate all work. The principle of comparative advantage suggests that even if one group of people can produce everything better than another, it is still beneficial for that group of people to specialize in what it produces best and then exchange with groups that specialize in what they do best. Milton and Rose Friedman illustrate comparative advantage using the example of a lawyer and a secretary. Suppose a lawyer wants to hire a secretary but the lawyer discovers that he or she can type twice as fast as the secretary. Is it worth hiring the secretary? The Friedmans respond that “If the lawyer is twice as good a typist but five times as good a lawyer as his secretary, both he and the secretary are better off if he practices law and the secretary types letters” (p. 45). I think the same principle applies to AI. Even if AI and autonomous robots can do everything that humans do (or better), it is still limited by time, space, and energy, no matter how efficient it becomes. As it advances, it would still be beneficial to let AI specialize in what it does best and humans can specialize in what they do best.
But what about income and wealth inequality? Won’t the AI revolution primarily benefit the rich and make the rest of us poorer? It is true that those businesses that are first to successfully incorporate AI will likely experience massive profits because of decreases in production costs. Bastiat points out that when one business person saves money by adopting a labor-saving innovation, other businesses begin adopting the innovation, and “…soon competition obliges him to lower his prices in proportion to the saving itself; and then it is no longer the inventor who reaps the benefit of the invention—it is the purchaser of what is produced…” (p, 32). The ultimate outcome of producing more at a cheaper cost is greater abundance and cheaper prices for consumers.
References
Saifedean Ammous (2023). Principles of Economics. The Saif House.
Daron Acemoglu & James Robinson (2012). Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Currency.
Frédéric Bastiat (2011). The Bastiat Collection (2nd ed.). Ludwig von Mises Institute.
Milton Friedman & Rose Friedman (1990). Free to Choose: A Personal Statement. Harcourt, Inc. [originally published in 1979].
Henry Hazlitt (1979). Economics in One Lesson: The Shortest and Surest Way to Understand Basic Economics. Three Rivers Press.
Walter Williams (2011). Race & Economics: How Much can be Blamed on Discrimination? Hoover Institution Press.