Interview: Architect Koichi Takada

Luke Michel has already lived through two technology overhauls in his career, first desktop publishing in the 1980s and online publishing later on. But AI? He’s had enough. 

So when his employer, the Dana-Farber Cancer Institute, made an early-retirement offer to some staff last year, the 68-year-old content strategist decided to speed up his exit. Before, he had expected to work a couple more years. 

“The time and energy you have to devote to learning a whole new vocabulary and a whole new skill set, it wasn’t worth it,” he said. 

It isn’t that he’s shunning artificial intelligence—he is learning Spanish with the help of Anthropic’s Claude. But, at this point, he’s less than eager to endure all the ways the technology promises to upend work. 

“I just want to use it for my own purposes and not someone else’s,” he said. 

After rising for decades and then hovering around 40% in the 2010s, the share of Americans over 55 years old in the workforce has slipped to 37.2%, the lowest level in more than 20 years.  

The financial cushion of rising home equity and stock-market returns is driving some of the decline, economists and retirement advisers say. 

But for some older professionals, money is only part of the equation.  

They say they don’t want to spend the last years of their career going through the tumult of AI adoption, which has brought new tools, new expectations and a lot of uncertainty.  

Many people retire when key elements of their work lives are disrupted at once, said Robert Laura , co-founder of the Retirement Coaches Association and an expert on the psychology of retirement. 

“Maybe their autonomy is being challenged or changed, their friends are leaving the workplace, or they disagree with the company’s direction,” he said.  

“When two or three of these things show up, that’s when people start to opt out.”  

“AI is a big one,” he adds. “It disrupts their autonomy, their professionalism.” 

Michel, whose work required overseeing and strategizing on website content, has been here before.  

When desktop publishing arrived in the 1980s, he was a graphic designer using triangles and rubber cement.  

The internet’s arrival changed everything again. Both developments required new skills, and he was energized by the challenge of learning alongside colleagues and peers. 

It felt different this time around. “Your battery doesn’t hold a charge as long as it used to,” he said. 

He would rather spend his energy volunteering, making art, going to operas and chairing the Council on Aging in North Andover, Mass., where he lives. 

In an AARP survey last summer of 5,000 people 50 and over, 25% of those who planned to retire sooner than expected counted work stress and burnout as factors.  

About half of those retired said they had left work at least partly because they had the financial security to do so. 

In general, older Americans are less likely than younger counterparts to use AI, research shows.  

About 30% of people from ages 30 to 49 said they used ChatGPT on the job, nearly double the share of those 50 and older, according to a 2025 Pew Research Center survey of more than 5,000 adults. 

Baby boomers and members of Generation X also experienced the sharpest declines in confidence using AI technology, according to a ManpowerGroup survey of more than 13,900 workers in 19 countries. 

“We as employers aren’t doing a good enough job saying (to older workers), we value the skills that you already have, so much so that we want to invest in you to help you do your job better,” says Becky Frankiewicz , ManpowerGroup’s chief strategy officer. 

Jennifer Kerns’s misgivings about AI contributed to her departure last month from GitHub, where the 60-year-old worked as a program manager.  

Coming from a family of artists, she said, it offends her that AI models train on the creative work of people who aren’t compensated for their intellectual property. And she worries about AI’s effect on people’s critical-thinking skills. 

So she was dismayed when GitHub, a Microsoft-owned hosting service for software projects, began investing heavily in AI products and expecting employees to incorporate AI into much of their work. In employee-engagement surveys, the company had begun asking them to rate their AI usage on a scale of 1 to 5. 

When it came time to write reports and reviews, colleagues would suggest that she use ChatGPT.  

“I’d be like, ‘I have no idea how to use that and I have no interest in using AI to write anything for me,’” she said. 

It would have been more prudent to work until she was closer to Medicare eligibility, she said. But by waiting until her children were out of college and some of her stock grants had vested, the math worked. 

Her first act as a nonworking person: a solo trip to Scotland, where she took a darning workshop and learned how to repair sweaters.  

“The opposite of AI,” she said. 

Employers already under pressure to cut workers—such as in the tech industry—may welcome some of these retirements, said Gad Levanon , chief economist at Burning Glass Institute, which studies labor-market data. 

“The more people retire, the fewer they have to let go,” he said. 

Some of the savviest tech users are also balking at sticking around for the AI upheaval. Terry Grimm, who worked in IT for 40 years, retired from his senior software consultant role at 65 last May.  

His firm had just been acquired by a bigger firm, which meant learning and integrating the parent company’s AI and other tech tools into his work.   

Until then, Grimm expected he might work a couple more years, though he felt that he probably had enough saved to retire. 

“I just got to the point where I was spending 40 hours at work and then 20 hours training and studying,” said Grimm, who has since moved with his wife from the Dallas area to a housing development on a golf course in El Dorado, Ark.  

“I’m like, ‘I’ll let the younger guys do this.’”