A former AI researcher who worked at both xAI and OpenAI recently announced their departure due to burnout—a scenario becoming emblematic of the AI arms race in Silicon Valley. From the "996" schedule (9 a.m. to 9 p.m., six days a week) to more extreme regimens of 16-hour days and 100-hour workweeks, a culture of overwork is rapidly spreading across the AI sector.
Researchers and industry insiders warn that this high-intensity work culture has taken root at top AI organizations like OpenAI and Anthropic and is now spreading to startups. Nathan Lambert, a senior research scientist at the Allen Institute for AI, stated plainly that this environment comes at a "human cost," including lost time with family, health deterioration, and inevitable burnout.
Meanwhile, the backdrop of this overwork wave is far more complex than individual choice: AI is reshaping the labor market, with a sharp decline in entry-level positions. The intertwining pressures of layoffs and technological anxiety have left professionals grappling with a dual dilemma—creating the future while worrying if there will be a place for them in it. Economists caution that this may only be the beginning of a larger-scale labor transformation.
The discussion was sparked by a widely circulated tweet in AI circles. A researcher who went by the handle hyhieu226 posted on platform X announcing their departure from OpenAI. They described working at both companies as a "once-in-a-lifetime experience," having met "the best people—not just in AI or tech, but the best among humans." They took pride in contributing to the creation of "highly intelligent entities that will tangibly improve human life."
However, they added a sobering note: "But intense work comes at a cost. I never thought I’d say this—I am burned out. The mental health issues I once dismissed are real, painful, frightening, and dangerous."
They plan to step away from cutting-edge AI labs, returning to Vietnam with family to focus on recovery. This restrained yet poignant statement has brought the issue of overwork in AI into public focus and resonated widely among professionals in the field.
The overwork culture in Silicon Valley is not new, but the AI era has pushed it to new extremes. According to reports, AI research scientist Sebastian Raschka observed that the work culture he’s seen differs from "996" but is "moving in that direction." He noted, "You have to deliver continuously; it’s really tough."
Some reports suggest even more extreme conditions: top AI researchers are working up to 100 hours per week, jokingly referring to their schedule as "0-0-2"—working from midnight to midnight with only two hours off on weekends. Sanju Lokuhitige, co-founder of Mythril AI, shared that he works seven days a week, about 12 hours daily, admitting, "Sometimes I code all day; there’s no work-life balance."
Another employee described living in a two-bedroom apartment with a founder, working from 9 a.m. until late at night, sustained by takeout and minimal sleep, stating plainly, "This isn’t 996; it’s a 16-hour shift."
A startup called Sonatic even explicitly required employees to be "on-site seven days a week" in a job posting, offering free housing and dating app subscriptions as compensation. The 21-year-old CEO claimed that when everyone is focused on the same mission, "the probability of success increases," and employees need not worry about housing, food, or socializing.
This wave of overwork is not solely driven by external pressures; internal motivation plays a significant role. Raschka admitted that during his time in academic AI research, he wasn’t forced to work long hours—he chose to. "It’s a combination of passion and competition that drives this grind mentality," he said.
Lambert pointed out that employees at OpenAI and Anthropic accept high-pressure cultures because they genuinely want to do the work, especially programmers. Kyle Finken, an engineer at Mintlify, noted that this is one of the most creative and productive periods in the industry’s history, with many working extra hours not just due to pressure but also out of enthusiasm for the technology.
According to reports, many of these top AI researchers have become millionaires, with some earning seven-figure salaries. "Some are multi-millionaires now, but several say they have no time to spend the wealth," one source noted.
However, the sustainability of this intense model is being questioned. Computer science professor Cal Newport describes it as "pseudo-productivity," where the measure of productivity is not output but visible hours spent at a desk or logged into Slack. In his book, he argues that "doing less is key to producing quality work," citing examples from Newton to Jane Austen.
The costs of intense work are becoming apparent, and worsening external conditions are heightening anxiety among professionals. Lambert emphasized the "human cost": lost family time, isolation from the outside world, and health issues. "You can only keep this up for so long; people are burning out," he said. Raschka shared that skipping breaks led to chronic neck and back pain.
At the same time, layoff pressures are reshaping power dynamics in the labor market. Reports indicate global tech companies laid off approximately 250,000 workers in 2025, with AI cited as a key factor. Tech leaders like Mark Zuckerberg and Elon Musk have publicly discussed AI’s potential to replace junior and mid-level engineers while pushing teams to improve "efficiency" and "resilience."
Executive coach Mike Robbins, who has worked with Google, Microsoft, and Salesforce, noted that the balance of power has shifted: "Five years ago, as a software engineer, you could set terms. Now, companies are more assertive because they’re less worried about losing staff."
Data from Indeed Hiring Lab shows a one-third decline in entry-level tech roles since 2022, while demand for specialists with over five years of experience continues to rise. Finken admitted that many are asking the same question: "Will I have a job in three years?"
Economists warn that the implications of this trend extend far beyond Silicon Valley. A Stanford University study documented a "significant reduction in employment for early-career workers" in AI-affected sectors, describing these industries as "canaries in the coal mine" for the broader economy.
Anthropic CEO Dario Amodei predicted that AI could eliminate up to half of all entry-level office roles within five years. IMF Managing Director Kristalina Georgieva compared AI’s impact on the labor market to a "tsunami," noting that in some advanced economies, up to 60% of jobs could be altered or disappear.
Yet in Silicon Valley, the race continues to accelerate. As one report concluded: overwork is back—at least until these engineers find a way to replace themselves with robots.
Comments