AI will not give you Fridays off

John Maynard Keynes predicted in 1930 that future generations would work no more than 15 hours a week thanks to technology and rising productivity. However, his forecast proved inaccurate. Americans averaged 38 hours a week in 1950 and now average 34. Despite nearly a century of technological progress, from electrification to the internet, work hours have only decreased by four hours.

Research on the four-day workweek is seven years old, with studies conducted in Iceland, Japan, the UK, Australia, and the U.S. The findings show that 45% of workers believed they could complete their tasks in five hours a day without interruptions, yet many worked over 40 hours a week, with the U.S. leading in overtime. Workers burned time, managed appearances, and absorbed inefficiencies that long workweeks encouraged.

Jamie Dimon, Anthony Scaramucci, and Bill Gates have all expressed optimism about AI ushering in shorter workweeks, but based on seven years of data and a century of history, it is unlikely. While a shorter workweek can benefit both companies and employees, the path to achieving it is not straightforward. Microsoft Japan reported a 40% productivity boost after moving employees to a four-day schedule, and a landmark trial across 33 businesses resulted in workers rating the experience 9.1 out of 10, with 97% wanting to continue.

However, the issue lies in the system that must support these changes. In American capitalism, efficiency gains tend to go to companies, not workers' calendars. The PC and internet did not shorten the workweek; instead, they extended workday into evenings and weekends. The problem is not technology but policy. Federal legislation mandating or incentivizing a four-day workweek does not exist in the U.S., and without that foundation, the optimistic predictions about AI's impact on working hours are unlikely to materialize.