Hospitals, law firms, and technology companies are previewing how AI might reshape work by automating tasks without eliminating basic jobs.
That’s the central message emphasized by Nvidia CEO Jensen Huang on his recent No Priors podcast.
In a wide-ranging interview, he argued that fears of mass job destruction often confuse the “mission” of a job with the broader “purpose” of the role. In his view, AI changes the way tasks are performed, but the purpose remains the same. This means that this technology is unlikely to destroy jobs and may even increase the demand for people who are accountable for their work outcomes.
Huang’s framework is straightforward. Most jobs include repeatable tasks that can be compressed through technology and broad objectives that remain human-driven. He highlighted radiology as a real-life example.
A few years ago, AI pioneer Jeffrey Hinton predicted that AI would eliminate many radiology jobs and advised students to avoid radiology jobs. The opposite happened. As AI automates many radiology departments, taskin fact, more radiologists are employed today than when Hinton predicted in 2016.
Below are some killer statistics shared in this 2025 blog post that explain why radiologists are still in great demand. In 2025, America’s diagnostic radiology training programs will offer a record 1,208 positions, an increase of 4% from 2024, and the field’s vacancy rate is at an all-time high. Additionally, by 2025, radiology will be the second highest-paying medical specialty in the country, with an average income of $520,000, more than 48% higher than the average salary for radiologists in 2015 (the year before Hinton’s prediction).
How did this happen? Huang insisted that the purpose of the job was not to “read scans.” These are tasks automated by AI. The real purpose of a radiologist is to diagnose disease, guide treatment, and support those efforts with research. When AI allows clinicians to evaluate more images with greater confidence, hospitals can serve more patients, generate more revenue, and justify hiring more experts.
Is typing really work?
The same logic applies across the economy, he said.
“I spend most of my day typing,” Huang said, noting that typing is a task, not an object of work. Tools that automate writing do not eliminate the need for executives. In many cases, it expands the amount of work that leaders and other employees can take on, he said.
“The fact that someone can automate a lot of my typing with AI is something I really appreciate and it helps me a lot,” he said. “That doesn’t mean I’ve actually gotten any busier. I’ve gotten busier in a lot of ways because I’ve been able to get more work done.”
coding, law, restaurants
This “task and purpose” framework is becoming increasingly prominent in knowledge work, with AI tools accelerating and automating tasks such as writing, summarizing, and generating code.
Huang pointed to software engineering as an example where AI can reduce time spent on the core task (writing code) while increasing demand for the purpose of the job: solving problems and identifying new problems worth solving.
He said Nvidia is actively hiring AI coding tools such as Cursor, even though they’re pervasive among the company’s engineering teams because the increased productivity allows the company to pursue more ideas. This could increase revenue and provide more funds to hire new staff.
Another example he gave was law. Although a lawyer’s job is to read and draft contracts, the goal of a lawyer is to protect the client and resolve disputes. While AI can accelerate document-heavy tasks, the real value of this role lies in judgment, strategy, and accountability, which requires experienced and trusted human lawyers.
This also applies to waiters working in restaurants. Their job is to take food orders, but the goal is to make sure guests have a good time, Huang said.
“Whether AI is taking orders or delivering food, their work is still helping us have great experiences,” the CEO added. “They will restructure their jobs accordingly.”
Huang’s point is not that AI won’t disrupt roles, but that it will. But he argues that early evidence suggests a redesign of jobs rather than a total collapse.
For employees, the implications are real. In other words, if your role is primarily defined by repeatable tasks, AI poses a direct threat. If AI is rooted in outcomes such as diagnostics, customer experience, problem solving, and dispute resolution, it could become a lever rather than a replacement, changing what people spend their time doing while keeping the purpose of their work intact.
Sign up for BI’s Tech Memo newsletter here. Please contact us by email. abarr@businessinsider.com.
