After Testing Four-Day Week, Companies Say They Don’t Want to Stop

Want to try a four-day workweek? Put this on the boss’s desk.

A large majority of U.K. companies participating in a test of a four-day workweek said they would stick with it after logging sharp drops in worker turnover and absenteeism while largely maintaining productivity during the six-month study.

In one of the largest trials of a four-day week to date, 61 British businesses ranging from banks to fast-food restaurants to marketing agencies gave their 2,900 workers a paid day off a week to see whether they could get just as much done while working less, but more effectively. More than 90% said they would continue testing the shorter week, while 18 planned to make it permanent, according to a new report from the study’s organisers.

The idea of working less than the conventional 40 hours over five days a week has been discussed for decades. The concept has gained new momentum recently as employers and employees seek new and better ways to work. The Covid-19 era ushered in broader acceptance of remote and hybrid work arrangements. Now, some employers, as well as policy makers, are exploring whether a shorter workweek can improve employee well-being and loyalty.

“At the beginning, this was about pandemic burnout for a lot of employers. Now it’s more of a retention and recruitment issue for many of them,” said Juliet Schor, an economist and sociologist at Boston College. Her team helped conduct the study with the nonprofit advocacy group 4 Day Week Global; U.K.-based think tank Autonomy, which focuses on issues including the future of work and climate change; and researchers at Cambridge University.

Global tests

Companies in the U.S. and Canada recently concluded a smaller pilot of a four-day week led by the U.K. study organisers, and similar trials are in the works in Australia, Brazil and elsewhere. Consumer-goods company Unilever PLC recently tested the concept in its New Zealand offices, while Spain’s government plans to pay companies to experiment with a four-day week. In a study in Iceland involving more than 2,500 employees across industries, researchers found most workers maintained or improved their productivity and reported reduced stress.

Widespread adoption faces a number of obstacles. Most companies that have experimented with a four-day week are small employers. Many larger companies haven’t embraced the concept. And at some companies trying four-day weeks, some workers have reported struggling to get everything done in that time.

In the U.K. study, which ran from June through November, most employees didn’t work more intensively, researchers say. Rather, they and their bosses sought to make work days more efficient with hacks such as cutting back on meetings and ensuring employees had more time to focus on completing tasks.

On a scale of 0 (very negative) to 10 (very positive), employers on average scored their productivity and performance over the six months at 7.5. A survey conducted halfway through the trial found 46% of companies said their business productivity had remained about the same, while 34% reported a slight improvement and 15% a significant improvement.

Meanwhile, 39% of employees said they were less stressed than before the pilot program started; about half reported no change. Nearly half observed improvement in mental health, and 37% also noted an improvement in physical health.

Zapping meetings

Claire Daniels, chief executive of Trio Media, a 13-employee digital-marketing agency based in Leeds, England, said she joined the trial to see whether a more effectively structured week could improve her business’s productivity. Before starting, she and her staff tracked and analyzed their workweek and concluded 20% of it was wasted in unessential meetings, business travel and other inefficiencies.

“So immediately, we knew we weren’t having to cram extra work in the four days,” she said.

Staggering everyone on Monday-Thursday and Tuesday-Friday schedules—with each employee having a partner to cover the day they were off—Ms. Daniels said she and her staff stopped holding marathon daily team meetings. And in longer meetings involving clients and multiple presentations, employees would drop in for portions and leave again, depending on how necessary their attendance was.

The hardest part, she said, was for staff to make sure they didn’t slip back into old work mind-sets or habits. On the whole, productivity was the same, or slightly improved, and revenue rose 47% compared with the year-earlier period, she said.

Ms. Daniels said she wants to continue the trial another six months before making a permanent change, “but I don’t see us going back to a typical five-days-a-week model.”

There is ample and alarming evidence that artificial intelligence can help humans do bad things, such as committing cyberattacks, building weapons and even killing themselves or others. The Hugging Face hacking episode—and a growing list of others by poorly constrained swarms of agents—indicate just how powerful and potentially dangerous AI has quickly become.

Yet many inside the U.S. AI industry insist far worse is coming, on account of the imminent arrival of AI with superhuman and self-improving abilities.

Plenty of experts, including many who study AI harms for a living, are skeptical. The so-called doomers’ assertion that AI might decide to wipe out all of humanity—or even “just” topple human civilization—is contingent on it achieving a pace of development not yet seen.

And if the assumptions behind this global-doomsday scenario are wrong, it could lead us to curb or regulate AI in ways that don’t address its real harms.

At the center of this debate is the claim that current AI systems might take over the job of training their next versions, a process called “recursive self-improvement.” Think of it like evolution on steroids—billions of years happening at light speed within vast AI supercomputers. Anthropic Chief Executive Dario Amodei recently proposed a global agreement to slow down the pace of releasing new AI models, with the goal of delaying the arrival of recursive self-improvement.

“I don’t think we’re anywhere near ‘artificial general intelligence,’” says Melanie Mitchell, a professor at the nonprofit research group Santa Fe Institute who studies AI. She said AI is making impressive strides but thinks claims by engineers that they’ve achieved recursive self-improvement don’t stand up to scrutiny.

She is hardly alone. A recent paper by two dozen academics at Princeton, Stanford and other institutions found that even the most cutting-edge AIs are incapable of doing the original research required to advance the AI frontier.

Two of the authors involved in that paper also threw cold water on the idea that the Hugging Face swarm hack by OpenAI agents occurred because of a breakthrough in intelligence. The attack succeeded primarily because of a lack of basic technical guardrails, not an unmanageable explosion in AI capability, they wrote.

Yann LeCun, former chief AI scientist at Meta, posted that this analysis was “a welcome dose of sanity in an otherwise insane debate.”

OpenAI and Anthropic didn’t respond to several requests for comment.

The people who disagree with the doomers still consider AI to be dangerous, and point out that such systems don’t have to be particularly capable to be powerful. Some argue that AI should undergo regular evaluation by outsiders and that the companies that make it should be held responsible when their systems do harm. AI should also be treated the same as airplanes and elevators, and should be designed to do the least harm possible, they say.

“If you believe that technology is powerful enough to create novel, dangerous viruses, or to essentially take over the whole planet for some reason, then it must also be strong enough to create cures for cancer, to cure aging, to fix socio-economic or political problems,” says Christopher Canal, CEO of EquiStamp, a company that helps companies and governments evaluate AIs.

AI has shown an ability to rapidly advance because it is matching the abilities of humans who are constantly feeding it their knowledge. Sometimes it can recombine that knowledge and exceed what people have been capable of, through a kind of post-training known as reinforcement learning, as we’ve seen in mathematics.

Today’s LLMs are “models of knowledge” rather than actually intelligent, wrote Yi Ma, professor of AI at Hong Kong University.

Researchers at universities and commercial AI research labs in China wrote in a recent paper that autonomous, self-improving AI is likely to be a long way off, due to the sheer number of breakthroughs required. They also argue that humans will probably remain in the loop, supervising that process—and gating how fast it can occur.

Vals AI, a company that evaluates today’s AI models, maintains an RSI Index that benchmarks whether models can “do the research that builds the next model.” So far, no publicly released model is even close.

Yet Rayan Krishnan, CEO of Vals AI, says his team projects models will exceed humans’ ability to improve the next generation of AIs by August 2027, or sooner, and at that point could start building their successors all on their own.

“Once we get to recursive self-improvement, the fear is that all bets are off,” he says. “You could end up with a ‘fast takeoff’ situation, where the models quickly acquire skills and eclipse humans across every possible domain.”

In a reply to the resignation tweet heard round the world from Jacob Coxon, another Anthropic engineer declared his belief that those odds were at least 10% over the next decade. Many others in the industry chimed in to say they thought the percentage was even higher.

Some who argue the end is nigh say they calculate their personal p(doom) based on a chain of conditional probabilities—a bit like the Drake equation for calculating the likelihood of intelligent alien life. Since all those probabilities are based on speculation, estimates range from 0% to nearly 100%. Many land around 10%.

“The weird thing is that if you go back and look at the predictions on this over the last 10 years or more, it’s always been 10%—it’s just a nice round number,” says Mitchell. “I think it’s all vibes, and there’s no actual evidence or calculation.”

One argument against worrying about superintelligent AI is that the world is full of unlikely humanity-ending disasters, and trying to avert them all can make it impossible to prioritize, says Canal.

This has led AI experts and the policymakers who listen to them to propose remedies that don’t get at its real and present dangers.

AI companies’ proposals to “pace the frontier” aren’t addressing the problem in the right way, argues Stuart Russell, a computer-science professor at the University of California, Berkeley, and the president of the International Association for Safe and Ethical Artificial Intelligence.

“It’s like saying we’re driving toward the cliff at 60 miles per hour and we’re going to drive toward it at 40 miles per hour instead, and everything will be OK,” he says.

Russell and his peers have proposed that AI companies should have to meet the same standards that govern other areas of everyday life, from air travel and buildings to food and drugs. They highlight the “behavioral red lines” AI should not be allowed to cross. Breaking into other computer systems, stealing information or advising terrorists on how to build biological weapons are all illegal for a human to do, and should be illegal for companies’ AIs as well, he argues.

The challenge for AI companies in such a proposal, he adds, is that it would be a de facto ban on today’s advanced AI systems, since the companies behind them don’t know how to make them respect such boundaries all of the time.

Others have proposed something like the Food and Drug Administration, but for AI, but setting up a new agency has so far been a nonstarter in Congress. And some prominent voices in tech have said such a structure would give up America’s AI edge to China.

Given the bipartisan groundswell of support for curbing AI companies and their creations, however, that may soon change.

“The tech industry has had this mantra for decades that regulation is bad,” says Russell. “They don’t accept the liability for any harm, and they hide behind free speech. That, I think, has to change.”

After Testing Four-Day Week, Companies Say They Don’t Want to Stop

Firms saw productivity hold mostly steady and fewer employees quit

Global tests

Zapping meetings

Could fears of an AI apocalypse be distracting us from the dangers already here? Experts debate whether regulation should focus on speculative existential threats or present-day harms, including cyberattacks, weapons and unsafe autonomous agents. Read more via the link in bio.