Not long after the terms “996” and “grindcore” entered the popular lexicon, people started talking to me about what was happening at startups in San Francisco, the hub of the artificial intelligence economy. There was also an article about a founder who hadn’t taken a weekend off in over six months. A woman joked that she gave up her social life to work at a top AI company. Or employees who have started taking off their shoes in the office. After all, if you’re going to be there for more than 12 hours a day, six days a week, you might as well wear slippers, right?
“If you go to a cafe on Sunday, everyone is working,” says Sanju Lokhitige, co-founder of pre-seed stage AI startup Mithril. He moved to San Francisco in November to be closer to the action. Lokuhitige says he works seven days a week, 12 hours a day, but holds several carefully selected social events each week where he can network with other people at the startup. “Sometimes I’m coding all day long,” he says. “I don’t have a work-life balance.”
An employee at another startup who came to San Francisco to work at an early-stage AI company showed me a harrowing photo from his office. It’s a two-bedroom apartment in Dogpatch, a neighborhood popular with tech workers. The founders of his startup live and work in this apartment. From 9 a.m. to 3 a.m., the only breaks are to DoorDash to eat or sleep, and the only time I leave the building is for a cigarette break. The employee, who asked not to be named because he still works for the company, described the situation as “horrible.” “I heard about 996, but they don’t even do 996,” he says. “They work 16 hours a day.”
Startups have never been particularly glamorous. Ten years ago, when I started reporting on the industry, people were making a lot of money in the new mobile app economy and programmers were drinking Soylent to stay at their desks longer. Startups back then were also defined by a hustle culture, high energy, and the pursuit of growth at all costs, and to some extent these ideas remain in the industry’s bloodstream.
But over the past year, as the magical dust of artificial intelligence has taken root in San Francisco, the mood among those working in the tech industry certainly seems to be changing. The excitement about a new era of technology and all the money that comes with it is currently tempered by worries about the industry and the economy. While some workers are fully committed to AI, some have doubts about whether it’s all good for the world. Some people are effectively training machines to do their jobs better than they are capable of doing. And many of those same workers struggling to build a future are now wondering if there is a place for them in the future they are building.
While the rest of us may be dimly aware of these anxieties, they are already tangible and keenly felt within the tech industry. Even the biggest technology companies, once known for pampering their employees with on-site massages and barbershops, have scaled back their perks as employee expectations have increased. Mark Zuckerberg and Elon Musk have each been outspoken in their predictions that some junior and mid-level engineers at their companies will be replaced by AI, and each called for their workforce to become more “efficient” and “very hard-core” as waves of layoffs put workers at risk. According to a report published by RationalFX, technology companies will lay off around 250,000 employees worldwide in 2025. AI is cited as a key factor in many of these layoffs, even though the full reason for the layoffs is often more complex.
“If you were a software engineer five years ago, you could write tickets,” says Mike Robbins, an executive coach who has worked at companies like Google, Microsoft, Salesforce, and Airbnb. The balance of power is now shifting away from technology workers, many of whom feel insecure about their job performance. “If companies are less afraid of losing employees, they will be a little more forthright about what they want and a little more demanding.”
Robbins, who wrote the book “Bring Your Whole Self to Work,” has been asked to speak to companies and their leaders about topics such as employee burnout, health and belonging, which are top priorities during and in the years immediately following the pandemic. “Frankly, we stopped talking about it,” he says. Business leaders are now seeking advice on topics such as change, disruption, and uncertainty in the workplace.
These themes, such as change, disruption, and uncertainty, are driving technology workers to spend more time and at higher intensity. Investment in artificial intelligence companies will reach an all-time high in 2025, but workers are feeling the scarcity like never before.
“It’s definitely on everyone’s mind,” said Kyle Finken, a software engineer at Mintlify, which makes AI tools for developers. “I think a lot of people are worried about, ‘Oh, will I have a job in three years?'”
Despite his concerns, Finken, like many other startup employees I spoke with, is encouraged by the “tremendous innovation” happening in artificial intelligence and believes there are still plenty of jobs for software engineers in the future, even if those jobs differ from today’s pure coding roles. He and other tech insiders characterize this as a particularly creative and productive time in the tech industry, with people putting in extra hours at work, not because their employers demand it, but out of sheer interest in new tools and capabilities. For example, Garry Tan, head of the famous startup accelerator Y Combinator, recently boasted that he “stayed awake for 19 hours” tinkering with Claude Code.
Even those who felt excited about the pace of change acknowledged that AI was rapidly augmenting work in ways that could have uncertain consequences for the future of work. “This is definitely not the time for complacency,” Finken says.
One of the reasons we work such long hours is to keep up with tools and technology that change almost every day. If you take a weekend off, you may miss out on key developments and it will be difficult to keep up with your competitors. Another reason is to have something to show future employers, especially as junior-level jobs tend to be replaced by AI.
“No one hires junior developers anymore,” says Mythril co-founder Lokuhitige. Now, to get a job, he says, you have to “do something great,” such as developing a new product or solving a problem that will be recognized as useful by a large company. According to Indeed’s Career Lab, job openings for entry-level technical roles will decline by a third from 2022, while jobs requiring at least five years of experience will increase. If you are not working hard at a startup, you are missing the prerequisites to get hired in the future.
What this means for the rest of us
Although economists disagree on whether AI will replace or simply change most jobs, they seem to agree that AI is already reshaping many entry-level jobs and will continue to do so. A paper published in November by researchers at Stanford University found that industries exposed to AI have experienced “substantial employment declines for early-career workers,” suggesting that sectors that are already undergoing change could become “canaries in the coal mine” for other parts of the economy. Anthropic CEO Dario Amodei suggested that AI could eliminate around half of humanity. all I plan to move into an entry-level position in a white-collar industry within the next five years.
The head of the International Monetary Fund recently predicted that 60% of jobs in developed countries will be eliminated or transformed by artificial intelligence, “like a tsunami hitting the labor market.” We’re already seeing early signs of this in San Francisco, where Uber drivers compete with self-driving Weymos and baristas are replaced by robot coffee bars. Professional business services that support the technology industry have also been negatively affected by layoffs. The pressure to sharpen up in the tech world could be an early sign of what many other industries will soon feel.
Robbins, the executive coach, said companies used to look to Silicon Valley as a model for how they operated, copying policies like unlimited vacation time and adopting perks like free lunches in the office.
“The business world has long idealized technology and Silicon Valley. Some of that has changed,” he says. “Now, like 10 years ago, people aren’t looking to me to tell them what’s going on in the Valley and help them accept it.”
Rather than being a model for how we should all work, the tech industry may be a harbinger of the insecurities that will come to us all and our attempts to compensate.
