AI could force 12 million workers to switch jobs in coming years, causing unprecedented competition

Workers will need to retrain or skill up to stay competitive in the job market in coming years, labor experts told BI.

• AI will spark a massive reshuffling of the job market, labor experts say.
• Researchers predict AI will displace as many as 12 million workers through the end of the decade.
• Those workers could have trouble pivoting to a new career if they don't adjust their skillset, career experts say.

The days of casual job-hopping may be coming to an end.

AI is about to make it a lot harder to get a new gig, especially for people working in industries that are highly exposed to AI disruption, career experts told Business Insider.

The advent of AI is a big warning for customer service reps, cashiers, office assistants, and production workers. "Medium-skilled" workers are the most likely to be displaced by the technology, according to Georgios Petropoulos, a labor market and digital tech researcher from the Massachusetts Institute of Technology.

While high-skilled workers are more likely to implement AI into their existing jobs, and trade workers like plumbers and electricians will see continued demand for services, Petropoulos predicts that medium-skilled workers will be in limbo over the next few decades as AI becomes more mainstream.

"We see a higher polarization of the job market," Petropoulos told Business Insider in an interview.

Though he didn't have an exact estimate, Petropoulos expects higher unemployment among medium-skilled workers as the economy transitions to an AI-run job market. That transition could take as long as 20 years, he added.

"We'll have more high-skilled jobs, more low-skilled jobs, and the middle that drops."

Kweilin Ellingrud, a director at the McKinsey Global Institute, anticipates lower-skilled labor to be in less demand as the effects of AI ripple across the US job market. By 2030, McKinsey estimates 12 million workers will need to find different careers altogether, about the same amount of job-switching that took place during the pandemic.

By the end of the decade, the economy could lose 630,000 cashiers, 710,000 administrative assistants, and 830,000 salespeople, McKinsey estimates. The number of clerks across all industries could plummet by 1.6 million, the firm said in a report last year.

"The typical job will require a higher level of skills than it did before," Ellingrud said of increased competition in the job market. "The simpler work, the more repetitive work that can easily be done by a machine or now a generative AI model, is going to be done by that model."

A split job market

Generative artificial intelligence is largely expected to boost productivity and add more jobs to the economy, but job gains largely lie in areas like healthcare and STEM fields, the McKinsey report found.

Meanwhile, employers outside those areas are already beginning to raise their standards of who they might bring on board. 66% of business leaders say they won't consider someone who doesn't know how to use AI, and 71% report they would prefer a job candidate with less experience and AI knowledge, according to a May survey conducted by Microsoft and LinkedIn.

That's left employees feeling the pressure to level up and make themselves more competitive in the eyes of prospective employers.

Demand for Gen AI courses has soared 1,060% this year on the online learning platform Coursera. It's the equivalent of someone signing up for an AI-related course every 15 seconds, according to Coursera CEO Jeff Maggioncalda.

"Will it make the labor market more competitive? It absolutely is already," Maggioncalda told BI, noting that workers who don't upskill will have trouble in the job market.

Coursera is also putting an emphasis on AI among its employees, Maggioncalda noted. The company now requires some of its workers to take mandatory AI skill classes.

"I often will tell my employees, it looks like you did not actually work with GPT to create this presentation or to write this piece. Run it through the model and sharpen your thinking. It's just not clear enough. My expectation for speed and clarity of thinking is much higher than it was two years ago," Maggioncalda added. "I don't think that there'll be much of a safety net for someone who is, say, a knowledge worker who didn't really learn AI. They lose their job."

It could take years for a worker displaced by AI to retrain and make themselves more competitive in a job market more heavily integrated with AI, Ellingrud speculates. On average, she thinks workers will need three to five years to adjust to the new employment landscape.

"I think workers of tomorrow will need to be more flexible than 10 years ago, and that's primarily because of all this occupational transition," she said. "So you might not want to learn a particular skill or get excited about that, but being flexible to shift to a different occupation, I think will be required more in the future than it was in the past."