Technology company leaders have promoted artificial intelligence as a tool that will significantly reduce working hours for employees, with some predicting four-day work weeks within recent years. Google’s engineering director made such a forecast several years ago, while OpenAI formally encouraged companies to trial shorter work weeks this year, arguing AI would enable substantial productivity gains.
However, current and former employees at these same companies paint a contrasting picture. A former OpenAI technical employee reported working at least 70 hours weekly while at the company, characterized by frequent crisis meetings, weekend work, and intense performance reviews. Workers at Anthropic described sprints lasting many weeks with peaks reaching 90 hours in a seven-day period. At Meta, employees described being “drafted” onto urgent AI projects without choice, facing long hours, overnight work, and feeling perpetually on-call.
Meta, OpenAI, and Anthropic declined to comment on these accounts. A former Google employee left the company in May citing negative impacts from AI implementation, including frequent all-night work sessions as the company redirected resources toward AI projects. Since departing, his sleep and health improved.
Research from UC Berkeley tracking tech workers over eight months found that employees using AI “worked at a faster pace, took on a broader scope of tasks, and extended work into more hours of the day.” MIT innovation scholar Neil Thompson explained that even with time-saving technologies, implementation demands and the need to verify AI outputs consume those savings. Additionally, workers tend to fill any freed time with additional tasks to demonstrate their value.
The contrast between executive promises and worker experience highlights how AI adoption in tech companies has intensified workloads rather than reduced them, with employees taking on expanded responsibilities and longer hours despite tools designed to improve efficiency.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI