The big AI job swap: why white-collar workers are ditching their careers

California-based Jacqueline Bowman had been dead set on becoming a writer since she was a child.

At 14 she got her first internship at her local newspaper, and later she studied journalism at university.

Though she hadn’t been able to make a full-time living from her favourite pastime – fiction writing – post-university, she consistently got writing work (mostly content marketing, some journalism) and went freelance full-time when she was 26.

Sure, content marketing wasn’t exactly the dream, but she was writing every day, and it was paying the bills – she was happy enough. “But something really switched in 2024,” Bowman, now 30, says.

Layoffs and publication closures meant that much of her work “kind of dried up.

I started to get clients coming to me and talking about AI,” she says – some even brazen enough to tell her how “great” it was “that we don’t need writers any more”.

She was offered work as an editor – checking and altering work produced by artificial intelligence.

The idea was that polishing up already-written content would take less time than writing it from scratch, so Bowman’s fee was reduced to about half of what it had been when she was writing for the same content marketing agency – but, in reality, it ended up taking double the time. “I now had to meticulously fact-check every single thing in the articles.

And at least 60% of it would be completely made up,” she says. “I would just end up rewriting most of the article.

So something that would take me two hours when I was writing it by myself now took me four hours, making half the money.” To add insult to injury, Bowman’s few remaining clients have sometimes accused her of using AI to create her work. “I never use AI to write anything,” she says, but she has noticed that AI-produced copy can sometimes seem eerily similar to her own writing – which she suspects is due to large language models being trained on some of her previous work.

She can’t afford to take any of the Silicon Valley giants to court – though she is grateful for the authors, like George Saunders and Ta-Nehisi Coates, who have done so.

By January 2025, Bowman was no longer able to afford her own health insurance, which hammered home what she had already begun to suspect: “Writing is not going to work out for me any more.” She decided to bring her wedding forward (she and her partner are still going ahead with their planned celebration in March, but last year obtained a marriage certificate from their local courthouse) so she would be eligible to join her husband’s health insurance plan.

But she knew a more drastic change would be needed before long.

She remembered a psychology elective she had enjoyed in college, and wondered if she might be able to make a more secure living by becoming a therapist. “It’s not AI-proof” – Bowman admits that some people will be happy to use AI-powered therapy services, which already exist. “But there’s another subsection of people who are going to say: ‘Hey, AI took my job, AI ruined my life.

I’m not going to go to an AI therapist,’” she says. “So in that way, I do think that there’s still going to be an audience who wants a human therapist.” Bowman decided to take action and retrain, “while I still do have a little bit of work”, and is now back at university studying to become a marriage and family therapist.

She counts herself “incredibly lucky” because she is able to rely on her husband, and on any writing work she can still get, to make ends meet – but has still had to take out loans.

She’s enjoying the course, and is “glad she has the opportunity to do it”, but it is not something she would have considered if her writing work hadn’t become untenable.

Janet Feenstra, an academic editor turned baker based in Malmö, Sweden, also has mixed feelings about her career change, a choice she similarly made because of fears that AI would make her old job void. “It’s complicated because, in a way, I maybe should be grateful to AI for prompting this change,” she says.