Too Old for Silicon Valley? Think Again. AI Is Changing the Math

Perhaps his most prestigious gig was at Palm, Inc., the consumer electronics and software pioneer.

The perks weren’t as lavish as at Google during the “glory days” of Silicon Valley employment, but there was a ping pong table, and the cafeteria was “fantastic,” which meant a lot to the young foodie.

He eventually pivoted to academia, feeling he could be “paid to indulge my curiosity and teach, which are two things that I would do all the time if I could,” pursued a Ph.D. in cognitive science and computer science at Indiana University, and became a computer science professor at Cal Poly Humboldt.

But when Kovitz decided last year to return to full-time work in Silicon Valley, he discovered a labor market dramatically reshaped by artificial intelligence.

He’d reinvented himself before. The question was whether the industry would let him do it again.

It turns out, yes. For all of Silicon Valley’s recent mass layoffs and historic ageism, experienced engineers like Kovitz who can confidently work with AI may have a better chance of riding tech’s next wave than the news headlines suggest. But to survive the brutal hiring gauntlet, they just might need to invest in help — both human and AI.

Is the market leaning in favor of experience?

Overall, tech jobs are surging. Online employment marketplace ZipRecruiter, which tracks job postings in IT and computer science, reports they were up 16.7% year-over-year nationally in May.

But that rosy picture looks different depending on your experience level. The share of senior-level job postings has risen to 43.1%, up from 38.8% a year ago. At the same time, the share of entry-level job postings has fallen slightly, from 8.1% to 7.4%.

“A key challenge within this job market is a distinct preference for senior or highly-skilled talent over entry-level hires,” ZipRecruiter labor economist Nicole Bachaud wrote to KQED.

That’s not great news for newly minted college graduates, but it is a hopeful sign for Kovitz and others with decades of experience.

Meanwhile, more than a quarter of employers now ask for AI skills. That’s nearly double the share a year ago.

“This surge is primarily fueled by businesses expanding their AI integrations, products, and services, creating high demand for workers to implement and deploy these new tools,” Bachaud wrote.

The fastest-growing roles in software engineering aren’t traditional coding jobs, according to Kory Kantenga, head of economics for the Americas at LinkedIn. They’re positions like “forward-deployed engineer,” a title pioneered by companies like Palantir to describe engineers who embed directly with clients to install, customize and troubleshoot AI tools on-site.

LinkedIn has seen an 18-fold increase in such roles.

“Those are the roles that we see have a lot of momentum,” Kantenga said.

How Ben Did It

While shy about sharing his exact age, Kovitz will say he’s old enough “to have watched moon shots on television” as a boy in the early 1970s, and to have programmed computers with punch cards.

He’s also old enough to know the Silicon Valley job market is vastly different from the one he learned to navigate at the start of his career. Today, AI defines not only the jobs, but the job search as well.

Applicants use AI to fire off applications while employers use AI to screen them. The impacts are especially acute in Silicon Valley, where the internet has turned many job openings into a worldwide competition with thousands of well-qualified applicants.

“I watched a friend struggle for a year after he finished his master’s degree,” he said, “and filling out applications every single day, hours a day. It took him a year to get a job.”

Kovitz decided he needed to pay for professional help to reduce the strength and duration of the struggle ahead of him.

“Reducing that by a few months and also reducing the sheer pain of it, that is easily worth it,” Kovitz said.

He signed up with two companies that offer job search support: ApplyPass and Resume Wizard 101, which has recently rebranded to Career Lander.

Resume Wizard assigned Kovitz a “reverse recruiter” who sent him new prospects daily, including many that Kovitz might never have considered on his own. The recruiter then filled out about 100 applications a month on his behalf.

ApplyPass used an AI agent to fire off applications at industrial volume — 400 a week.

Given the volume of “cold applications” coming at employers, it’s essential job seekers — or agents acting on their behalf — reach out to the relevant person at the company who has some influence over the hiring process, according to Neil Bhatt, Career Lander’s founder and CEO. “The conversion rate (from application to interview) skyrockets when you actually do some sort of networking.”

Kovitz said Resume Wizard 101’s human touch led to a higher rate of employer responses. But ApplyPass isn’t solely automated. The company provided a coach to help navigate “some of the weird things they do in software engineering interviews today.”

After every interview, Kovitz shared how he responded to certain questions, and the ApplyPass coach suggested alternative approaches.

“We talked about what went right and what went badly,” Kovitz said.

These services aren’t cheap. ApplyPass’s auto-apply plans range from free to about $199 a month, while Resume Wizard 101’s human-run packages range from roughly $300 for a single résumé revision to $10,000 for full ‘career search management.’

But senior software engineers in San Francisco earn a median total compensation north of $270,000 a year, according to Glassdoor, so Kovitz figured the math was likely to pencil out for him.

He was also surprised to find some prospective employers that want proof a software engineer can work with AI. “Go at it with AI,” he was instructed in one memorable interview. “Just throw everything you’ve got at it. Let’s see how you do it,” he said.

This was new to Kovitz, but he dove in with gusto. “I blasted out 2,000 lines of code and solved the problem,” Kovitz said. “But I also saw that the solution wasn’t that great.”

Kovitz said he thinks getting useful work out of an AI coding agent takes judgment that only comes with experience.

As a senior engineer, Kovitz argued, he was able to look at those 2,000 machine-generated lines of code, recognize their shortcomings, and then know what to do about it.

“A lot of what you do with AI is you work out a plan, a little conversation with the AI, because if you just tell the AI ‘Do this thing,’ with no plan, it is going to make a gigantic mess, and you are going to spend a week debugging it.”

Effective prompts require understanding the code and what the company wants from it at a granular level.

“You need to know how all kinds of things can go wrong. Junior programmers don’t have that. They haven’t seen the development of code over a long period of time,” Kovitz said.

“It’s really a game of marketing at the end of the day,” said Bhatt at Career Lander. He said any employer is going to choose the person who can best explain how they use their experience to help this company. “The better job you’re doing at articulating how you can solve their problems, the easier it becomes to land that position.”

In a survey of 400 engineering leaders in the U.S., India and China by Karat, a company that runs technical interviews for hire, 73% said a strong engineer is worth at least three times their total compensation, something they specifically attributed to AI’s impact on productivity. The same survey found that AI is widening the gap between strong and weaker engineers.

So where did Ben land?

Kovitz landed a job after a six-month hunt — slightly less than average for the information sector, according to the Bureau of Labor Statistics, which tracks tech workers as part of that larger group. After submitting what he estimates was around 3,000 job applications, Kovitz got a software engineering job at Impulse Labs, a San Francisco company that makes high-end induction stoves.

He’s happy with the job and turned other offers down to take it.

“Their engineering culture makes quality paramount even as they move fast,” Kovitz said.

Also, as a foodie, Kovitz loves sitting with a stovetop next to his desk for hands-on testing.

Does he work with AI on the new job?

“Yes. Most of the day, every day,” he said, for a wide variety of things that would otherwise take him much longer. They range from finding the relevant parts of a large program to work on, to writing small bits of code in a language he doesn’t know well, to reviewing code for errors.

“AI is both amazing when it works well, and amazingly unreliable, so I have to keep a close eye on it,” he said.

He’s glad the job hunt is finally in his rearview mirror, but Kovitz is philosophical about the effort.

“While it was grueling, it was also a great adventure,” he said. “I learned all kinds of stuff about industries that I’d never known anything about before.”

When he applied for a job with a robotaxi company, he taught himself the basics of motion planning, the math behind how a self-driving car navigates. For another set of interviews, he picked up some basics about satellites.

Kovitz’s facility with math and science, not to mention his natural curiosity and enthusiasm for intellectual challenges, gave him the kind of career flexibility only the most elite coders have enjoyed in recent decades.

But he also maintained a positive attitude by consciously framing the search as a good thing.

“It was an exciting learning experience,” Kovitz said. “It was part of life’s adventure.”

His advice for others still looking?

1. Get help if you can afford it.
2. Treat the search like an eight-hour job and knock off afterward, so the rejection doesn’t grind you down: “You need to protect your morale.”
3. Don’t go into job interviews with an attitude of desperation. “I’m not there to get the job,” Kovitz said. “I’m there to find out if I want the job.” He added, “It’s wise to be fussy,” and hold out for an employer “that has a culture that you like, where the work is meaningful to you.”
4. Treat this as a numbers game: “You know, you are probably going to roll the dice 50 or 100, 150 times before you actually get a good [position].”
5. Don’t let the perfect be the enemy of the good, as the saying goes: “There are plenty of perfectly good jobs out there for lots of people.”