HUndreds of thousands of tech workers are facing a harsh reality. Their well-paying jobs are no longer safe. Now that artificial intelligence (AI) is here, their futures don’t look as bright as they did a decade ago.
As US tech companies have ramped up investments in AI, they’ve slashed a staggering number of jobs. Microsoft cut 15,000 workers last year. Amazon laid off 30,000 employees in the last six months. Financial-services company Block eliminated more than 4,000 people, or 40% of its workforce, in February. Meta laid off more than 1,000 in the last six months, and, according to a Reuters report, may cut 20% of all employees in the near future. Just this week, the software giant Oracle laid off thousands of workers. Smaller players like Pinterest and Atlassian also made recent cuts, culling about 15% and 10% of their workforces, respectively. Estimates put the total number of tech layoffs in the past year at more than 165,000, according to the tracker Layoffs.fyi.
“At no point in my career have I ever been this pessimistic about the future of careers in tech,” said a tech employee, who has worked at big tech companies for decades and requested anonymity for fear of retribution. “And that’s really sad because I love tech.”
The anxiety extends beyond Silicon Valley. Because tech companies are seen as innovators of the corporate world, as they reduce their headcounts – in anticipation of AI efficiency gains, or to prioritize AI investments – the moves could set a precedent for other businesses to make similar cuts.
But even though AI has helped to accelerate coding, analyze large datasets and aid with research, many AI experts say we’re still a long way from AI being able to replace large swaths of the workforce, if it ever can. So what is really going on?
In interviews over the last month, AI researchers, economists and tech workers said that essentially, we’re all living through an experiment. Over the next few years, tech companies’ experimentation with AI will probably lead to several critical outcomes: more job cuts across industries, unseen consequences from overreliance on AI and a fundamentally different model of work.
“The maximum hype you have right now, which is that AI is replacing people, is not true,” said Ethan Mollick, an associate professor at the Wharton School of the University of Pennsylvania who studies AI. “But it’s also not true that AI will never threaten jobs. It’s going to be complicated.”
Reshaping jobs
OpenAI, Anthropic and Google have promised that their generative AI tools, such as ChatGPT, Claude and Gemini, will change the way people do their jobs, automating time-consuming tasks and shifting humans to more complex work. Agentic AI, or bots that complete tasks without human intervention, takes that promise further, potentially automating entire roles or business functions.
On the ground floor, tech workers are facing the first phase of the AI experiment, as they’re pushed to use the tech more often. But the outcomes don’t always align with leaders’ expectations.
For technical workers, using AI has become a baseline expectation for employers across the tech industry, said a former Block engineering supervisor who got laid off in February.
AI helps generate code faster, but this makes keeping up with code reviews more difficult, he said. Human reviews are important to think through any potential conflicts the code may have with other parts of the system and spot bugs that AI makes look legitimate, he added.
“Now there’s three times as much code because it’s producing faster,” he said. “We were falling behind on reviews.”
A recently laid off senior user-experience designer at Amazon Web Services, who asked to remain anonymous for fear of retribution, said his team was experimenting with two internal generative AI tools core to their jobs, both of which were in early testing phases. Neither was fully functional or useful for workers’ jobs yet, he said. So when cuts hit his team, he was surprised and confused.
“It felt like, ‘None of this is ready yet,’” he said. “How is all this work going to get done?”
Amazon employees felt a veiled threat that if they did not use AI, their jobs could be next, he said, echoing earlier reporting from the Guardian that employees say the tech company pressures them to use AI even when it slows them down. Amazon stressed in previous statements that AI use was not mandatory.
As more tech workplaces center AI and urge employees to embrace it – sometimes that push comes with surveillance and enforcement.
A former worker at Microsoft said when it came to his and his colleagues’ AI use, he had the “feeling of being watched” and felt pressure to “adopt the tech whether we like it or not”. He also requested anonymity for fear of retribution. He felt he could voice concerns about AI at work if it helped protect the company from a bad outcome, but larger societal concerns were less welcome.
“I can’t bring up environmental or job concerns,” the worker said. “You don’t want to be known as the person against AI.”
Microsoft said it maintained system‑level oversight of AI usage for security and risk but didn’t use individual usage as a performance metric. The company also said it offered multiple channels for employees to anonymously raise concerns about how the tech was used.
The power of AI
Some companies are already touting the gains they’ve seen from AI. Google, for example, credited AI for 50% of its code in its latest earnings report. Block’s head of engineering, at the company’s November investor day, said 90% of the company’s code submission was authored “partially or fully with AI support”.
However, in its current form, AI is not as capable as some of the hype suggests, said Stephan Rabanser, a post-doctoral researcher at Princeton University who has co-written a white paper about the reliability of AI agents. While the output of generative tools has been improving over the years, the tech still has problems consistently producing the same correct answer, even when the same prompt is used. That especially gets messy when there are different users or conditions, Rabanser said.
“This is the barrier to job transformation,” he said. “Reliability will be a key limiting factor.”
More companies will probably experience failed AI deployments or problematic results, Rabanser said.
AI systems need huge amounts of data to become even acceptably good at a task, said Stuart Russell, a University of California, Berkeley, professor and an AI researcher, and high-quality training data is becoming scarce. Often, even when a chatbot lacks the necessary data, it will respond confidently anyway, producing wrong answers that can lead to faulty transactions and deleted databases, he added.
AI also struggles to learn continuously and remember what it did previously, Mollick, of Wharton, said. Nevertheless, some companies are already adopting advanced-use cases, relying on AI to write all their code and then shipping those products without human review, despite the risk from AI’s limitations, he said. He called them “dark factories”, since they operate largely without human supervision.
Betting on AI like this is risky. It creates exposure to financial losses, reputational harm, and negative customer or client outcomes, according to AI and business experts.
In some cases, over relying on AI can cause critical consequences far beyond the business. “We don’t want to move fast and break things in high-risk situations, like in healthcare or judicial fields,” Rabanser said. “There are high stakes involved” that in some cases could mean life or death, he added.
The truth behind the cuts
While the drumbeat of companies that say AI will help them do more with less is getting louder, it’s unclear whether AI is actually driving cuts. Some companies may be “AI-washing” layoffs, using the technology as a convenient excuse for a slowing labor market, lagging consumer demand or rising costs, researchers and AI experts said.
Just this week, the prominent venture capitalist Marc Andreessen, a bona fide AI booster who has written that “AI will save the world,” said on a podcast that large tech companies were culling workers because they were overstaffed, and “now they all have the silver-bullet excuse: ah, it’s AI.”
“It’s easy to confuse the effects of something like generative AI with a weakening of the labor market,” said Ryan Nunn, director of research at Yale University’s Budget Lab, which researches AI’s impact on jobs. “We really don’t see anything differentially happening with the AI-exposed labor market.”
If a company is struggling financially, saying AI drove cuts definitely makes for a better story, said Thomas Malone, professor of information technology at the Massachusetts Institute of Technology’s Sloan School of Management.
There’s also a long history of overshooting predictions of the impact and adoption rate of new tech, he said. It happened in the dot-com era and with autonomous driving.
“I do think many people are overestimating the rate at which jobs will change,” Malone said about AI projections.
When Pinterest announced an almost 15% cut of its workforce in January, it cited reasons including reallocating resources to teams focused on AI and prioritizing AI‑powered products and capabilities. But a Pinterest employee, who asked for anonymity because she was not authorized to speak to the press, said she believed the layoffs were more about fixing the company’s business than anything else.
“While I know that AI was one of the reasons cited, I don’t think it was the real reason,” she said, adding that cuts were related to optimizing operations. “They did a thorough review of the entire business, and what you see now is a sort of leaner, meaner Pinterest.”
Pinterest called this a mischaracterization.
The potential savings and competitive advantages of AI are compelling for Wall Street investors. Headcount reductions can imply greater productivity per employee, which then leads to higher profits, said Joseph Feldman, analyst at Telsey Advisory Group.
After Jack Dorsey, Block’s CEO, connected his company’s layoffs directly to AI productivity gains, the company’s stock price increased by 20%.
But cuts alone don’t always satisfy the market, which is also watching for signs of sustainability, several analysts said. Two weeks after the initial pop in price, Block’s stock was down 6%, signaling that the market recognized the execution risk, said Matthew Coad, analyst at Truist Securities.
“A big part of it is the uncertainty around, ‘Did (Dorsey) cut into bone?’” Coad said, referring to the engineering staff.
And in the day after Oracle’s layoff news, the company’s stock popped up by 7.5%. But the boost was short-lived, as days later the stock had retreated to near pre-layoff levels. Amazon similarly experienced a stock pop after its latest cuts in January, though stock has since dropped in the months following as the market questions its AI spending plans.
Even the markets are trying to make sense of the hype surrounding AI. For those seeking a clear answer on exactly how this tech will transform work and the economy, the answer is yet to be determined. This tech is changing some jobs, but the greater impact will take years to play out.
“We will see changes over the next couple of years as a result of AI,” Mollick said, referring to anticipated improvements in the tech. “It’s already changing programming. So it will change jobs and transform them, but we just don’t know the job consequences yet.”
