The Bogus Four-Day Workweek That AI Supposedly 'Frees Up'
Business leaders are enthusiastically promoting artificial intelligence as a gateway to shorter workweeks and improved work-life balance. However, without significant shifts in power dynamics, workers are unlikely to reap the benefits of these technological advancements.
The Utopian Vision of AI
Recent headlines, such as those in the Washington Post, have breathlessly declared that AI is key to implementing four-day workweeks. Executives from companies like Zoom, JPMorgan Chase, and Microsoft, including figures like Eric Yuan, Jamie Dimon, and Bill Gates, have rhapsodized about AI freeing up time, with some even speculating about workweeks as short as two or three days. Elon Musk has taken this idea to the extreme, predicting that within 10 to 20 years, working could become optional, with universal high income eliminating poverty.
Yet, this vision is largely unrealistic. Despite massive investments in generative AI, studies, including one from MIT, show that 95% of organizations are seeing zero return on their investments. Even if AI does boost productivity, history suggests that workers may not see the gains.
The Reality of Productivity and Wages
Worker productivity has been rising for years, yet median wages have stagnated when adjusted for inflation. If AI leads to a shorter workweek, it is more likely to come with reduced pay—four days of work for four days' worth of wages, three days for three, and so on. This could leave many workers poorer or forced to take on additional jobs to maintain their current income levels.
In his 1930 essay, Economic Possibilities for Our Grandchildren, John Maynard Keynes predicted that by 2030, technology would create an age of abundance, freeing people from economic worries and allowing them to focus on leisure. However, with just five years to go, this prediction seems wildly off the mark. Instead, new technologies have contributed to a two-tiered society, with a few accumulating extraordinary wealth while many struggle to make ends meet.
The Distribution Dilemma
The core issue is how productivity gains are distributed. When AI enables more to be done by fewer people, the question of who gets paid what becomes critical. Without bargaining power, profits are likely to concentrate among a small circle of owners, leaving the broader population with less money to purchase the very products and services AI creates.
Imagine a futuristic device, an iEverything, capable of producing any desired item or service. While this sounds wonderful, it highlights a real dilemma: if no one has the means to earn money because AI handles all work, who can afford such innovations? This scenario underscores the urgent need to address distributional fairness.
Power and Politics: The Path Forward
For workers to share in AI's productivity gains, they need bargaining power. Labor unions, which once represented over a third of the private-sector workforce, now cover only 6%, offering limited leverage. This leaves politics as a potential avenue for change.
Will average working people gain the political muscle to demand fair distribution through measures like wealth taxes funding childcare, elder care, and healthcare? Alternatively, could a new workers' party emerge to champion these causes? The answer remains uncertain, but without such shifts, AI risks widening inequality further.
In the meantime, it is crucial not to fall for the breathless rhetoric about AI 'freeing up' employees' time. The real question is whether AI's productivity gains, if they materialize, will be shared with workers. The truth is, employers are unlikely to share these gains unless compelled to do so through collective action or policy reforms.
