Productivity up, burnout up: The AI productivity paradox

American workers have experienced a significant surge in productivity over the past five years, with output for non-farm businesses increasing by 2% annually, according to data from the Bureau of Labor Statistics cited by The Economist. Federal Reserve Chairman Jerome Powell has acknowledged being surprised by this trend, stating in a March press conference that he never expected to see so many years of really high productivity. While artificial intelligence has been widely credited as the driving force behind such workplace gains, Stanford economist Nicholas Bloom argues the real catalyst is the shift toward work-from-home policies that accelerated during the pandemic. Bloom points to trial data showing that remote work increases productivity by eliminating commuting time and reducing general office inefficiencies, while also expanding the labor supply by enabling more people to participate in the workforce. However, the integration of AI into daily work life has brought unexpected consequences for many employees, particularly in high-adoption countries like South Korea, where 51.1% of surveyed workers reported using AI on the job, the highest rate among countries studied. Workers across multiple industries describe AI as having accelerated the pace of work to exhausting levels rather than easing their burdens. Tasks that were once simple and time-consuming have been automated, but workers now report spending their days on more mentally draining activities like reviewing AI-generated output, coordinating across departments, and validating results. A data engineer at a major Korean search platform described how expectations shifted immediately once everyone began using AI tools, noting that easier work disappeared while more difficult work multiplied. The productivity gains promised by AI have not translated into lighter workloads for many employees. A recent Upwork Research Institute survey found that 81% of managers in the US and UK reported that AI increased their expectations for worker productivity, while 71% of employees reported experiencing burnout. The Boston Consulting Group has coined the term "AI brain fry" to describe the cognitive overload from managing multiple AI systems simultaneously, noting that productivity gains plateau or decline when workers juggle four or more systems. Researchers from the University of California Berkeley Haas School of Business found in an eight-month study that employees using AI ultimately worked faster, longer, and handled more tasks. Harvard Business Review has warned that such sustained increases in productivity pressure could eventually lead to declining work quality, higher turnover rates, and deteriorating decision-making. Meanwhile, companies including Home Depot, Amazon, Instagram, and Stellantis have moved to require employees to work in-office five days per week, with more than half of Fortune 100 companies now mandating full in-person attendance, a trend Bloom describes as counterintuitive to productivity goals.