America Is Heading Toward an Infinite Workweek
The Atlantic reports on a wave of 'AI brain fry' sweeping knowledge workers tasked with overseeing AI agents. A BCG and UC Riverside study of 1,488 U.S. employees found 14% report mental fatigue from excessive AI oversight, with workers managing three or more agents simultaneously facing the worst effects. Developer and blogger Steve Yegge describes his role as an 'AI babysitter,' while BCG managing director Matthew Kropp likens the variable reward of agent outputs to 'pulling slot machines.' MIT economist David Autor says the change will not produce a 'white-collar apocalypse' but work that simply looks different. The piece argues that as agents now run overnight, pressure to keep pace may be pushing the country toward something resembling an infinite workweek.
What happened
Lila Shroff, a staff writer at The Atlantic, published a reported feature on June 18, 2026, arguing that AI agents are reshaping work in ways that are less about job elimination and more about cognitive overload and always-on availability. The piece draws on a BCG and UC Riverside survey of 1,488 U.S. employees that found 14% reported 'AI brain fry' - defined by researchers as 'mental fatigue that results from the excessive use of, interaction with, and/or oversight of AI tools beyond one's cognitive capacity.' Among developers specifically, 18% reported AI-induced exhaustion, and workers managing three or more agents faced worse effects, requiring 14% more mental effort and experiencing a 12% increase in mental fatigue.
The 'babysitter' dynamic
The central mechanism Shroff identifies is oversight labor. AI coding agents can handle more tasks than a year ago, but still require constant human supervision - asking follow-up questions, needing detailed instructions, and causing failures if left unsupervised. Developer Steve Yegge told Shroff that managing simultaneous agents is like having 'a dozen browser tabs open in my head, all fighting for attention.' BCG managing director Matthew Kropp compared the unpredictable quality of agent output to a variable-reward system: 'It's very akin to gambling,' he said, adding that the dopamine circuit hack keeps workers feverishly rotating among agents rather than taking breaks.
Labor economics angle
MIT economist David Autor offered a longer-term framing, saying, 'People are constantly talking about a white-collar apocalypse. I don't think it's going to look like that.' Instead, Autor expects many jobs to persist but look different - the analogy being a preindustrial cobbler versus his daughter on the factory floor: both made shoes, but under entirely different conditions. AI researcher Andi Peng speculated that within the decade, agents may record conversations and conduct computer work autonomously, potentially eliminating laptops from office work.
The 'infinite workweek' thesis
The piece concludes with the observation that since agents can now work overnight, workers face growing pressure to assign tasks before sleep and review results at dawn. Venture capitalist Marc Andreessen stated on The Joe Rogan Experience that 'the opportunity cost of going to sleep is too high. If you go to sleep, you won't be with your 20 AI coding agents.' Shroff frames this as analogous to how Slack and email made workers feel always-available - but with agents that never stop working, the pressure may be qualitatively different.
Scoring Rationale
Well-reported Atlantic feature grounded in a BCG/UC Riverside survey of 1,488 workers with concrete fatigue data and named expert sources. Relevant to AI practitioners on the human-factors side of deployment, but the story is about labor culture rather than technical advances, placing it solidly in the mid-range.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
