The central argument is that the huge spending on artificial intelligence may not end in a financial bust, even if the revenue required to justify it seems implausibly large. Earlier generations of workplace technology were paid for from I.T. budgets and helped employees do their jobs; A.I. can perform tasks itself, putting it in competition with corporate payrolls. A survey of large companies found that A.I. budgets had tripled in a year and that only 15 cents of every dollar came from I.T. departments. While few executives say A.I. has increased sales so far, many expect it to do so soon, and nearly half anticipate reduced employment.
If companies fund A.I. by shifting money from wages to computing, the investment boom could succeed while workers bear the cost through fewer openings, slower wage growth and job losses. One economic scenario estimates A.I. could add $2.8 trillion to U.S. output by 2030, but also leave knowledge workers more likely to be unemployed and reduce labor’s share of national income. The piece invokes the 19th-century “Engels’ pause,” when productivity rose long before wages caught up, and research linking automation in Dutch firms to lower earnings and harder job searches. New A.I.-related occupations may eventually emerge, but too late or far from displaced workers. Because labor-market damage may unfold quietly rather than through a dramatic crash, governments and society should prepare rather than trust markets to close the gap quickly.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.