Nearly nine in ten recent and upcoming graduates worry AI could replace entry-level roles, according to a recent Monster survey.

But AI won’t eliminate skilled workers. It will change what they can accomplish. In industry after industry, the advantage will belong to people who can wield this powerful technology with sound judgment.

This isn’t the first time a transformative technology has inspired predictions of mass job displacement. When spreadsheets arrived in the 1980s, people worried accountants would become obsolete. Search engines were supposed to replace researchers and librarians. 

Instead, those technologies automated routine work — and increased the value of workers who could analyze information, solve problems, and advise people. 

MIT economist David Autor argued in a recent National Bureau of Economic Research paper that AI’s greatest economic opportunity may lie in extending expertise rather than replacing it. In effect, AI can enable more people to perform higher-value work. 

History suggests that technology reshapes jobs more often than it replaces workers altogether.

AI promises a similar shift on a much larger scale. It can summarize reports, organize information, generate code, and draft first versions of documents in seconds. But those outputs still require people to ask the right questions, recognize what matters, identify mistakes, and make decisions when the stakes are high.

That is why employers continue to prize communication, critical thinking, and problem-solving above almost everything else, according to survey research from the National Association of Colleges and Employers. 

Those skills remain difficult to automate — and will become even more valuable as AI becomes more capable.

Healthcare offers a clear example.

AI can sift through vast amounts of data from continuous glucose monitors, heart monitors, and diagnostic tests far faster than any human being. It might detect that a patient’s overnight glucose levels have begun drifting upward, flag an irregular heart rhythm that deserves closer attention, or identify a subtle pattern in diagnostic results that would otherwise go unnoticed.

But AI can’t decide whether a patient should change medications, whether a clinician should alter a treatment plan, or how to explain those findings to someone sitting in an exam room.

AI has helped my colleagues at Abbott develop medical technologies that deliver more personalized health insights — from helping people understand how a meal may affect glucose levels to identifying meaningful patterns hidden within diagnostic data. 

And with our distribution network for medical devices expanding, AI has our helped IT, business and distribution teams improve inventory placement and identify opportunities to reduce transportation costs. It is the people on these teams who apply advanced analytics and mathematical modeling so we can effectively position products across the country and the world.

Each of these breakthroughs depends on people. Engineers have to build the algorithms. Scientists have to validate them. Clinicians have to determine how they fit into patient care. Product developers have to earn the trust of the people who use them.

Employers are already beginning to reward that combination of human judgment and AI fluency. Microsoft’s latest Work Trend Index found that seven in 10 managers would prefer to hire a less experienced candidate with AI competencies than a more experienced one without them.

That’s encouraging news for new graduates. AI fluency is becoming a marketable skill — one that can help distinguish candidates entering an increasingly competitive job market.

Every generation has had to master the defining tools of its era. AI is this generation’s power tool. It won’t eliminate skilled workers. In skilled hands, it will allow them to accomplish far more.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.