Scared tech workers are scrambling to reinvent themselves as AI experts

While tech workers are dealing with pay stagnation, layoffs, and generally less demand for their skills than they’d enjoyed for the past decade, the artificial intelligence specialist has become the new “it” girl in Silicon Valley.

“All of the products that we’re working on, that we’re seeing today, are shifting toward that AI-powered type of operation,” said Zac Brown, founder of the AI startup NonprofitsHQ. “This is a rough time to be a regular software engineer.”

When Brown was looking for jobs last year, he hadn’t updated his resume to focus on all the work he’d done with AI in his previous roles. Previously, the 28-year-old had been used to companies tripping over themselves to talk to an experienced software engineer, but all of a sudden, he wasn’t seeing the same interest.

“I was a software engineer, because that’s what I always was,” Brown told Vox. “I realized this last time I was looking, you have to highlight the AI that you’ve got, because that is what’s attractive to companies right now.”

While tech companies and investors pull back seemingly everywhere else in tech, money is still flowing into AI, which the industry sees as the next big thing. That’s meant outsize demand, pay, and perks for people who can facilitate that kind of work. This situation is incredibly attractive to people who’ve recently been laid off in tech or who worry that their tech jobs don’t have the upward mobility they used to. To capitalize on this, people in adjacent tech careers are attempting to reposition themselves where the good jobs are. Short of getting another degree, many are hoping to do so with on-the-job training, boot camps, and self-education.

“If you take a look at job openings right now on job boards, the job listings are more emphasized on software engineers who have a background in AI,” technical recruiter Eddiana Rosen told Vox. (Data from the salary site Salary.com showed that in the 12 months ending May 31, 2023, employers sought AI-related skills in 1.1 million job postings, more than twice the amount for the same period a year earlier.)

“On top of that, when it’s time to negotiate for a higher salary, those people will have more advantages and more leverage.”

People with AI skills are paid on average 27 percent more than typical tech workers, according to data furnished by the compensation software company Payscale. The median annual salary for an AI engineer was $243,500 in May, according to data by the tech career comparison site Levels.fyi, compared with $166,750 for non-AI engineers. And their pay is growing at a faster rate. Comprehensive.io, which tracks compensation across more than 3,000 tech companies, found that pay for senior software engineers who specialize in AI and machine learning grew 4 percent since the beginning of the year, while pay for senior software engineers overall stayed flat. A.Team, a firm that connects groups of tech talent with companies looking to hire their services, said 30 percent of their new pipeline over the last month was AI-related, a fivefold jump over the previous three months.

Big tech companies are scouting AI talent from universities, even while rescinding offers for non-AI talent, says Zuhayeer Musa, co-founder of Levels.fyi, which also helps candidates negotiate offers. Those companies are also trying their best to hold on to the talent they have, offering key AI engineers multimillion-dollar retention bonuses lest they leave for more exciting opportunities at other firms, especially smaller ones where the work might be more interesting and the potential for growth, both financial and technical, higher.

“It’s kind of a bonanza,” Musa said. “We’re seeing people go from everywhere to everywhere.”

In contrast to crypto or web3, few people think AI will be a flash in the pan. Just how prevalent it becomes, of course, will depend largely on how profitable business use cases for it are. Already, tech workers are losing jobs to AI, so many figure they might as well get ahead of it and get in on the action. They’re turning to communities on Reddit as well as to friends and colleagues already in the field to find out how they can pivot to lucrative jobs in AI rather than having their jobs replaced by AI.

For Brown, who is no longer looking for work since he founded his own AI startup, selling himself was a matter of advertising skills he already had and that he’d picked up on the job. A previous employer gave him the opportunity and the time to work with another team that was working with machine learning and AI, so that he could fill in his skills gap. Brown said that while a boot camp or online course might provide a good introduction to the skills needed for AI, the best instruction comes from working on it yourself.

“A lot of that is going to come in actually doing it, actually working on these systems, messing it up, making mistakes, learning from those, and pushing forward,” he said.

That’s not always easy to do, especially in an economy where even tech workers are being laid off and where companies are better positioned to hire people who already have those skills.

Taylor, a software developer in North Carolina who asked that we not use his full name so as not to jeopardize his employment, plans to moonlight at a friend’s AI startup, where he’s hoping he’ll be able to learn from a back-end engineer there who’s experienced in machine learning. The goal isn’t to become an AI engineer per se, but rather to be able to do his job better, since he believes AI will creep into regular software development work in the near future.

“It’ll either help me more in my current job or help me get the next job,” Taylor, 41, said.

Eric Lamy, a customer data product manager who typically works alongside engineers, is trying to develop his knowledge around AI governance because he sees it as a future need at his current job, where he sits on a corporate cybersecurity task force and where there’s a lot of interest in how to responsibly deploy AI. To up his understanding of the new technology, the 37-year-old is using a body of knowledge document released by the International Association of Privacy Professionals, which will soon provide certification and training for AI governance professionals, to guide his independent study.

“It’s not so much that the transition is relying on a gap that’s in my current career; it just doesn’t really exist yet as a function,” Lamy said. “I see this as a place where people, who are able to get on board early and learn some of these frameworks and apply this governance mindset, have an opportunity to really do some good work.”

Nitin Pathak, a data scientist at Microsoft, recently got a six-month professional certification in machine learning and AI from Berkeley so he could perform better at his job. “It really helped me draw those connections between machine learning and AI concepts, and which models would make sense for different business situations,” he said.

“I’ve been working in technology for several decades, and in today’s world, it’s so clear that we’ll all have many careers,” he added. “I love technology and I don’t want to be obsolete. I want to be on the cutting edge.”

Even those who studied AI as part of an advanced degree are always working to stay abreast of the tech.

Nicole Hedley, a 30-year-old fullstack product engineer who runs her own consulting company that specializes in generative AI, had taken artificial intelligence and machine learning courses while getting her computer science degree. She’s also had plenty of hands-on exposure to new AI projects, especially in the last year when they’ve dominated her workload. Even still, she’s constantly trying to keep up with new advances in AI.

“Because there are so many recent developments, it’s a constant learning process,” Hedley, said.

Of course, just because you become fluent in AI doesn’t mean your career is bulletproof.

Alexander Whedon, a software engineer who specializes in AI, was laid off from Meta earlier this year, despite his skills. But now he considers that loss a “blessing in disguise.” As a freelancer, Whedon, also 30, gets to work on a wider variety of projects for a wide variety of companies.

“I enjoy this work so much more and I honestly make more now,” Whedon, who advocates for trying to build your own AI projects rather than going through boot camps, said.

“The future of any company isn’t sure,” he added. “But the future potential of AI I think is very potent.”

A version of this story was also published in the Vox technology newsletter. Sign up here so you don’t miss the next one!