NVIDIA CEO Jensen Huang argues that artificial intelligence will automate certain work tasks, but not entire jobs. He believes this distinction will determine whether the future job market faces mass unemployment or a new wave of growth and opportunity.
Speaking at the Y Combinator Startup School in San Francisco last weekend, Huang directly addressed the impact of AI on the labor market. "Many work tasks will be automated," he said in a video of his talk released by Y Combinator on Monday. "Every job will change, and a large number of new jobs will appear."
However, Huang made it clear that work tasks are not the same as jobs. Therefore, he does not agree with the predictions of widespread white-collar job losses made by some prominent AI industry leaders. Instead, Huang believes AI will create more jobs, not fewer. "The idea that AI will destroy jobs is the opposite of what's happening," he said. "AI will automate some tasks, but it does not necessarily eliminate jobs."
To support his point, Huang explained that a person's job has a purpose, and that purpose includes many tasks. He noted that even if some of these tasks are automated, the purpose remains. He cited the profession of radiology, which "godfather of AI" Geoffrey Hinton once predicted would soon become obsolete. In 2016, Hinton said hospitals should stop training radiologists because deep learning would surpass them within five years. Hinton later revised his view, admitting his initial statement was too broad.
Contrary to Hinton's original prediction, the field of radiology has not shrunk. According to a study in the Journal of the American College of Radiology (JACR), the number of practicing radiologists grew by about 12% between 2010 and 2022. Another JACR study projects the field's workforce will increase by another 25.7% to 40.3% by 2055, depending on whether residency positions continue to expand.
"The reason is an extremely serious patient backlog," Huang explained. "Now doctors and hospitals can treat more patients. To treat more patients, you need more nurses and radiologists." Although AI is increasingly helping radiologists analyze scan images, experts in the field say reading images is only part of their job. Radiologists also consult with other doctors and monitor patients, while interventional radiologists perform actual treatments. If AI takes over the task of reading images, radiologists can shift their focus to other work.
Huang says the same applies to software engineers. Even if AI agents can automate the task of writing code, companies will still hire more developers to achieve increasingly ambitious goals. "There are a huge number of ideas, ambitions, and visions waiting to be realized," Huang stated. "If we can automate the task of programming, companies can hire more software engineers to explore more business, and we can set even higher goals."
Admittedly, Huang has previously acknowledged that rising automation will inevitably cause some pain. In January, he admitted that, as in any period of major technological change, some jobs will disappear entirely. He also said that workers who are good at using AI will outcompete those who are not. Nevertheless, Huang's overall judgment is that companies will use the productivity gains from AI to expand their business scale, rather than to conduct mass layoffs. "This is a classic example of productivity driving growth," Huang said. "Increased growth leads to more job opportunities."
Comments