Nvidia Chief Executive Jensen Huang (Jensen Huang) pushed back against forecasts that artificial intelligence will eliminate large numbers of white-collar jobs.

Business Insider reported on July 28 that Huang drew a line at Y Combinator Startup School, saying AI eliminates not jobs but “tasks.”

Huang said all jobs will change, but that does not mean jobs will disappear. “All jobs will change, and many new jobs will be created,” he said. “The narrative that AI destroys jobs has been completely reversed,” he added.

Huang’s main argument is that people should distinguish between a job and a task. Using customer service as an example, he explained that steps such as answering calls, taking complaints and responding based on database information can be automated. But the disappearance of individual tasks does not mean a person’s entire job goes away. “A person’s job has a purpose, and under that purpose there are many tasks,” he said.

That view intersects with customer service, software development and the broader professional fields where the impact of AI adoption is concentrating. Uber cut 10 percent of its customer service staff last week and cited AI technology as the reason. Huang said task automation and shrinking employment should not be viewed on the same plane.

He explained the medical field with the same logic. Huang said automation has already advanced in tasks such as reading radiology scans, but radiologist jobs are still increasing. With a large backlog of patients, he said AI tools can reduce the burden of reading scans, allowing hospitals and doctors to take more patients. “There is an enormous patient waiting backlog,” he said. “Now doctors and hospitals can take far more patients,” he added.

Huang said software development is undergoing similar changes. He said developers are shifting from writing code themselves to directing and reviewing AI coding tools such as Claude Code and Codex. Huang said programming itself is closer to a task that can be automated, but argued that more software workers may be needed because ideas and demand are piling up. “The backlog of ideas and ambitions is too big,” he said. “If we can automate programming tasks, we can hire more software engineers to do more work,” he added.

Legal technology startup Harvey was also cited as an example. Huang said some expected the tool to reduce paralegal jobs, but argued that demand for the role is rising rapidly. The U.S. Bureau of Labor Statistics forecasts that the occupation will see little or limited change from 2024 to 2034.

Huang’s remarks contrast with the earlier position of Anthropic CEO Dario Amodei (Dario Amodei), who warned that AI could eliminate half of entry-level white-collar jobs. Amodei later moderated his tone toward the view that AI will sharply raise productivity, but he still maintains that job losses could be an inherent result of AI.

Huang said higher productivity does not prevent employment from expanding. “Productivity growth is a classic case of growth,” he said. “As growth increases, more employment follows,” he added. AI will clearly change how people work, but he said the unit of change should be seen not as entire jobs but as a reshuffling of tasks.