When Anthropic CEO Dario Amodei warned that AI could wipe out half of all entry-level white-collar jobs within five years, he wasn’t hedging. He told Axios it could spike unemployment to 10-20% and gut hiring across tech, finance, law, and consulting. Amazon Web Services chief Matt Garman has heard the prediction. He isn’t buying it.Speaking on Casey Newton’s Platformer podcast, released Tuesday, June 23, Garman drew a hard line between two words. “Wipe out and change are different,” he said. His argument is that white-collar jobs will be reshaped by AI, not erased by it, and he runs one of the best vantage points in the business to make that call. AWS is a roughly $150 billion-a-year operation, Anthropic’s primary training partner, and the platform where millions of companies run their AI experiments. From that seat, the doomsday framing looks, to him, like bad math.Why the AWS boss thinks Anthropic CEO Dario Amodei’s “wipe out” prediction misses the pointGarman runs one of the best listening posts in the AI economy. AWS is a roughly $150 billion-a-year business, the primary training partner for Anthropic, and the platform where millions of companies run their AI experiments. From that seat, he sees the panic as overblown. His reasoning is partly economic. “If you believe half of jobs get wiped out, then the whole economy collapses on itself and everything goes away, and then you’re not going to have AI, and you have to go back to those other jobs at some point. The math doesn’t work out,” he told Newton.Amodei’s original prediction, delivered to Axios in May 2025, was blunter. The Anthropic CEO warned AI could spike unemployment to 10-20% and eliminate half of entry-level roles across tech, finance, law, and consulting, urging the industry to stop “sugar-coating” what’s coming. He has since doubled down, arguing in a January 2026 essay that enduring job loss may be an “intrinsic property” of how AI replicates human cognition.
The Excel argument: how old technology reshaped work instead of killing it
Garman reached for a familiar analogy to make his case. Spreadsheet software like Microsoft Excel eliminated the jobs of people who once calculated by hand, he noted, but those workers learned to use computers and moved on. New tools spawn new roles. “The key thing is not to look at a still picture of the world and say that job’s not going to exist,” he said. “New jobs will be created.”It’s an argument with company. Nvidia’s Jensen Huang and Microsoft co-founder Bill Gates have made similar cases, while Google DeepMind’s Demis Hassabis talks up a future of “radical abundance.” The data, so far, is muddy. A Stanford study last August found AI hitting entry-level workers hardest, particularly young software engineers. But economists like Apollo’s Torsten Slok pin recent graduate struggles on broader economic forces, noting the unemployment gap for young grads opened months before ChatGPT launched.
Why Amazon is hiring 11,000 fresh grads this year
Garman’s clearest evidence is a hiring number. Amazon plans to bring on 11,000 interns and new college graduates in 2026. His logic is practical, almost blunt. Entry-level workers are the cheapest to hire, haven’t picked up bad habits, and arrive eager to learn new tools. “They come in with an energy and excitement, a new view on things,” he said. “If you just have the exact same people you’ve had for the last 15 years, you don’t get that energy and excitement and new ideas.“
The catch: your job in two years won’t look like today’s
The optimism carries a condition. Garman tells Amazon employees their roles will look “vastly different” in two years, and survival depends on a willingness to learn. What he now screens for isn’t a specific skill set but the ability to pick up new ones. “Do you have the willingness to dive in and learn new things?”The tension in his position is hard to miss. Amazon has cut roughly 30,000 corporate jobs since October 2025, and CEO Andy Jassy has said AI will “reduce our total corporate workforce” in the years ahead. AWS itself sells agents that recruit, code, and process insurance claims. Garman is talking up human labor even as his company ships the tools that may displace it. His answer is that Amazon has automated work for 25 years, always finding higher-value tasks for people to move into. Whether that pattern holds at AI’s speed is the open question.