- AI politics professor Baobao Zhang said trades and care work are safest from automation.
- She argued that roles requiring empathy and hands-on skills, like nurses, are harder to replace.
- Zhang warned young workers to stay flexible, retrain if needed, and use networks to compete with AI.
Generative AI is reshaping industries at breakneck speed, putting many traditional "safe" white-collar jobs at risk. But not every role is on the chopping block.
"There are certainly jobs that will not yet be automated in the next maybe 10 years," Baobao Zhang, an AI politics professor at Syracuse University, said on the London School of Economics' "Ballpark" podcast on Monday.
At the top of her list are tradespeople, including electricians, plumbers, and repair workers, who perform hands-on tasks in messy, real-world environments that machines still struggle to navigate.
She also pointed to care jobs like nursing, primary school teaching, and nursery teaching as roles that heavily rely on empathy, judgment, and social connection — qualities that algorithms can't yet mimic.
Another safe bet, Zhang said, is advanced manufacturing, where specialized roles still require human oversight despite growing automation on factory floors.
"They're not traditionally considered prestigious industries," she said. "But it's these back-to-basics jobs that are harder to automate."
As AI takes over professions, Zhang also warned that adaptability will matter just as much as initial job choice.
She said even students pursuing degrees in more exposed professions should be ready to retrain and lean on human networks when competing for opportunities in an AI-saturated labor market.
"Employers are now flooded with AI-generated résumés, and in response, they have to use AI to filter out these résumés, and so it's this kind of arms race between applicants and employers," she said.
"Unfortunately, I think what gives you an advantage is having that human connection."
Experts are split on which jobs AI will spare
Zhang's assessment is broadly in line with other leading AI voices.
Microsoft researchers recently found that jobs most exposed to AI overlap are those heavy on writing and communication, such as translators and historians, while roles like nursing assistants and hazardous-materials workers are among the least affected.
Geoffrey Hinton, the so-called "Godfather of AI," has also argued that physical trades like plumbing will be safer than "mundane intellectual labor" like paralegal work.
Not everyone is so measured.
AI safety pioneer Roman Yampolskiy has warned of near-total job wipeout by 2030, while Nvidia CEO Jensen Huang has predicted AI will transform jobs rather than erase them outright.
Zhang took a less bombastic view, saying, "I think it's more important to focus on the job reports, the empirical data that's coming out in terms of how AI is impacting workers currently and perhaps in the near future."
