Dario Amodei Challenges Jensen Huang’s Vision of Global A.I. Integration

The national security risks of selling A.I. chips to China far outweigh the benefits of spreading U.S. technology worldwide, according to Dario Amodei. The Anthropic founder and CEO is pushing back against recent policies that frame such sales as a way to integrate the technology of leading U.S. companies, such as Nvidia, into global ecosystems.

“Are we going to sell nuclear weapons to North Korea because that produces some profit for Boeing?” asked Amodei while speaking at the World Economic Forum in Davos, Switzerland, today (Jan. 20). “That analogy should make clear how I see this trade-off—that I just don’t think it makes sense.”

In recent months, restrictions on A.I. chip exports have been eased under the Trump administration, in part thanks to Nvidia CEO Jensen Huang’s lobbying. This pullback gives A.I. leaders less time to understand the technology’s development, societal impacts and existential risks of new technologies, according to Amodei. “The reason we can’t [slow down] is because we have geopolitical adversaries building the same technology at a similar pace,” he explained.

Demis Hassabis, CEO of Google DeepMind, echoed the need for a more nuanced approach to A.I.’s geopolitical challenges while speaking alongside Amodei on a Davos panel. When it comes to establishing safety standards, international cooperation between nations like the U.S. and China is “vitally needed,” said Hassabis.

Hassabis added that his concerns extend beyond governments to academia. He said he has been “constantly surprised” by how few economists and professors are seriously examining A.I.’s effects on issues like job displacement and wealth distribution.

Getting A.I.’s societal deployment right, Hassabis argued, will require the technology’s evolution to slow. Achieving that slowdown, however, “would require some coordination.”

The early days of A.I.’s labor impacts

Both Amodei and Hassabis said they are already seeing A.I.’s influence on the labor market within their own companies. Hassabis pointed to a “slowdown” in hiring at Google DeepMind, especially for entry-level roles such as interns.

Amodei, meanwhile, has long warned that A.I. could trigger major labor disruption. Last year, he said the technology could wipe out 50 percent of all entry-level white-collar jobs within five years. He said this transformation is already taking place within Anthropic, and his company is thinking internally about how to manage the shift “in a sensible way.”

Both leaders believe the job market will ultimately adapt, including through the creation of new A.I.-enabled roles. Still, Hassabis stressed that work is about more than income. Questions around how meaning and purpose are tied to jobs are among those “that keep me up at night,” he said, adding that the financial effects of labor disruption are easier to solve “than what happens to the human condition, and humanity as a whole.”

While the Anthropic and Google DeepMind CEOs largely agree on the geopolitical, societal, and labor implications of A.I., they diverge on timing. Amodei believes A.I. could reach the capabilities of a Nobel laureate within just a few years. Hassabis, by contrast, puts the odds of human-level A.I. at 50 percent by the end of the decade.

Even so, both agree that neither timeline leaves much room for companies, policymakers, or governments to develop a coherent response to the technology’s growing influence. “There isn’t a lot of time before this comes,” said Hassabis.