Dario Amodei says he’s not trying to be a “prophet of doom” on AI and jobs, but the Anthropic CEO warns that mass displacement may be part of what makes AI work, rather than just a temporary growing pain for the technology.
In a new policy essay, Amodei said that despite efforts to soften the blow, there is a “good chance” that AI could cause “significant and lasting job losses”, which he said “may be an inherent property of the technology and the way it broadly reproduces human cognition.”
This discussion reframes one of the AI industry’s most uncomfortable questions. If AI systems are designed to perform many of the cognitive tasks that humans do, job losses may not simply be the result of bad corporate behavior or short-term adjustments, as some executives claim. Amodei suggests they may be a structural consequence of successful AI development.
In other words, it’s a feature, not a bug.
Mr. Amodei has sounded the alarm on this point before. He has previously warned that AI could eliminate half of entry-level, white-collar jobs and increase unemployment to 10% to 20% within five years, and called on companies and policymakers to stop “facing” the risks. Rather than predicting a specific jobs apocalypse, his latest essay details what governments should do in the event of an intolerable displacement.
His answer is twofold. It’s about slowing the damage and sharing the benefits.
Amodei called for better “measuring and tracking” of the impact of AI on the labor market, including expanded government statistics. He also supported “employment incentives” such as wage insurance for workers forced to take lower-wage jobs because of technology, retention tax incentives, workforce training grants, and improvements to job-matching infrastructure.
If AI permanently reduces the demand for human labor, he writes, governments may need to go further, including “long-term income support” such as universal basic income, funded by taxes on related businesses and higher capital gains taxes. He also advocated a universal capital account as another way to spread the wealth created by AI.
The essay comes amid a broader shift in tone, as some of the industry’s most prominent AI leaders, including Amodei and OpenAI CEO Sam Altman, have recently emphasized productivity gains and new economic opportunities rather than warnings of job losses.
Business Insider recently reported that executives who once emphasized the disruptive effects of AI are now spending more time discussing how workers and society can benefit from the technology’s benefits in the lead-up to the company’s long-awaited IPO.
Amodei’s latest policy memo says Anthropic aims to help Rather than focusing solely on cost savings, enterprise customers discover new revenue and “get the most out of their existing workforce.” But if the benefits of AI are truly as huge as he claims, society needs a plan for workers who may not automatically share in the benefits, he said.