AI is not just taking jobs or making jobs, it is rebuilding work itself

Jensen Huang argues that AI is creating jobs through infrastructure, manufacturing, and faster business growth. The bigger story is that AI is reshaping work unevenly, creating opportunity in some areas while putting pressure on office roles, entry-level jobs, and workers without a clear reskilling path.

The bigger shift is not just one CEO being optimistic

The surface story is simple. Nvidia CEO Jensen Huang told a Milken Institute audience that artificial intelligence is creating jobs and could help reindustrialise the United States. His argument is that AI is not just software sitting in the cloud. It needs chips, servers, data centres, power systems, factories, engineers, technicians, builders, operators, and entire supply chains around it. That is a serious point, because the AI boom is already turning into a physical infrastructure boom. But it is not the whole story. Workers are still worried, and they have reason to be. A recent workplace survey found that 41% of U.S. employees said their organisation had adopted AI tools, and employees inside those AI-adopting organisations were more likely to report both hiring growth and layoffs. That is the real issue. AI can create jobs and remove jobs at the same time. The argument is not whether one side is completely right. The argument is who gets the new work, who loses the old work, and whether the transition is honest enough for normal people to plan their lives.

The old way was to talk about jobs as fixed things

For a long time, people talked about jobs as if they were solid objects. You were a designer, accountant, driver, programmer, teacher, assistant, journalist, mechanic, lawyer, or customer service worker. The title described the work, the career ladder, and often the social identity that came with it. AI makes that old way of thinking weaker. A job is not one thing. A job is a bundle of tasks, judgement, relationships, responsibility, experience, communication, and trust. Some tasks are easier to automate than others. Writing a first draft, summarising a call, generating code, sorting support tickets, checking forms, and producing simple images can be changed quickly. Managing people, reading a room, handling conflict, making hard calls, earning customer trust, and taking responsibility when something goes wrong are harder to replace. Huang’s point, as reported, is that automating a task does not always mean removing the whole job. That sounds technical, but the plain-English point is simple. AI may not replace every worker neatly, but it can still hollow out parts of their job and change what employers value.

The worker fear is not imaginary

The problem is that workers are not making this fear up. A Pew Research Center survey of employed U.S. adults found that 52% were worried about the future impact of AI in the workplace, while only 36% said they felt hopeful. About a third said AI would lead to fewer job opportunities for them in the long run, and only a small share believed it would create more opportunities for them personally. That matters because the public conversation often sounds too neat. Executives talk about productivity and new industries. Workers hear that and wonder whether their current role, their next promotion, or their children’s first job will still exist. The fear is not only about mass unemployment. It is about losing the first rung of the ladder. It is about the graduate role that no longer gets posted, the admin role that becomes part-time, the support job that becomes monitoring software, or the creative job that gets pushed into lower-paid editing work.