Tech leaders say AI means less work - their staff say they work up to 90 hours a week
Despite tech leaders claiming AI will lead to shorter workweeks, employees at major AI companies like OpenAI, Anthropic, and Meta report working up to 90 hours a week.
Intelligence analysis by Gemini 2.5 Flash
Executives at leading AI firms advocate for a four-day work week, citing AI's potential to reduce human labor. However, current and former staff at these very companies reveal a demanding culture with frequent 'sprints,' mandatory reassignments to urgent AI projects, and workweeks often exceeding 70 hours, directly contradicting the public narrative.
Imagine grown-ups who make super-smart computer brains (AI) say that these brains will soon do so much work that everyone can have a shorter week, maybe even a four-day weekend! But the people actually building these smart brains say they are working super long hours, sometimes 90 hours a week, just to make them. It's like the people making the fastest race cars are too busy working on them to ever take a break and drive one themselves.
Analysis
The narrative surrounding artificial intelligence often paints a picture of a future where human labor is significantly reduced, potentially leading to a four-day work week. This vision is frequently championed by executives at companies heavily invested in AI development, such as OpenAI and Meta. However, a closer look at the internal workings of these organizations reveals a stark contrast to this optimistic outlook, with employees reporting grueling schedules and immense pressure.
OpenAI
OpenAI, a prominent AI research and deployment company, publicly urged other firms to trial a four-day work week, suggesting that AI would soon accelerate human labor sufficiently to warrant such a shift. Yet, a former technical employee who departed the company last year disclosed to the BBC that OpenAI itself never implemented this policy internally during their tenure. Instead, the individual described a demanding environment characterized by frequent 'crisis meetings,' mandatory weekend work, and a 'super cut-throat' performance review system that often resulted in sudden dismissals. This internal reality of 70-90 hour workweeks stands in direct opposition to the company's external advocacy for reduced working hours.
Meta
Similarly, Meta's CEO Mark Zuckerberg has spoken about AI dramatically changing work, enabling fewer employees to achieve more. However, current and former Meta employees describe being 'drafted' into urgent AI projects without choice, facing long hours, working into the night and on weekends, and feeling constantly 'on call.' This aggressive push into AI development, which includes creating tools for software engineering and building infrastructure to replicate human tasks, has coincided with significant layoffs, further intensifying the workload for remaining staff. The company's approach has created a culture where refusing AI-related assignments could lead to termination, despite recent attempts to soften this stance.
Amin Shali
The impact of AI development on employee workload extends beyond those directly building the tools. Amin Shali, a former Google employee, left the company due to the negative effects AI had on his job. He reported frequently working late nights because internal engineering functions were failing, a consequence of Google redirecting crucial resources like processing units and memory storage towards its burgeoning AI projects. Shali's experience underscores how the prioritization of AI can strain existing infrastructure and increase pressure on non-AI engineering teams, leading to burnout and a detrimental work culture. His improved health and sleep since leaving Google highlight the severe personal cost of this intense work environment.
Key points
- Tech executives claim AI will lead to a four-day work week and reduced human labor.
- Employees at companies like OpenAI, Anthropic, and Meta report working 70-90 hours a week, often in 'sprints' or on urgent, mandatory AI projects.
- A former OpenAI employee stated the company never trialed the four-day work week it advocated for others.
- Meta employees describe being 'drafted' into AI teams, with long hours and feeling constantly 'on call'.
- Even non-AI engineers, like a former Google employee, experienced increased workload and system failures due to resources being diverted to AI projects.
If the promises of AI to automate tasks and increase efficiency are fully realized, it could eventually lead to the reduced workweeks and improved work-life balance that tech leaders envision, freeing up human workers for more creative or less demanding roles. The technology's potential to streamline operations could ultimately benefit employees by reducing mundane tasks and allowing for greater flexibility.
However, the current reality suggests AI development is intensifying workloads, with employees facing extreme hours and pressure, potentially leading to widespread burnout and health issues. This could also result in significant job displacement as AI becomes more productive, creating economic instability and a worsening of work culture if companies prioritize AI integration over employee well-being.
