The False Promise of a U.S.-China AI Detente

The push for a dialogue between Washington and Beijing on artificial intelligence has been touted as a necessary step in preventing a catastrophic arms race. However, beneath this veneer of cooperation lies a complex web of interests, with both sides seeking to gain an upper hand without making meaningful concessions.

The timing couldn’t be worse. Treasury Secretary Scott Bessent’s meeting with Chinese Vice Premier He Lifeng is set against the backdrop of President Donald Trump dismissing public warnings about runaway AI as a “hoax.” Meanwhile, the state-owned China Daily has accused America’s leading AI companies of using calls for a slowdown to “blunt China’s AI advance.”

These statements reveal entrenched positions rather than genuine cooperation. The reality is that neither the United States nor China is ready to slow down in their pursuit of AI dominance. Neither trusts the other, and rapidly advancing AI systems won’t wait for diplomacy to catch up.

The Problem with Parallel Progress

Recent attempts at forging a U.S.-China partnership on AI research and development have been plagued by a fatal flaw: they assume both sides can agree on a common goal without addressing underlying power dynamics. This approach is reminiscent of past diplomatic efforts, where negotiations were used as a means to advance interests while pretending to make concessions.

Failed talks are a familiar pattern in U.S.-China relations. Technical exchanges on AI safety, for instance, seem like progress but ultimately mask deeper divisions between Washington and Beijing. By focusing on testing models for dangerous capabilities and behaviors, both sides can claim progress without making meaningful concessions on the technology itself.

The Limits of Crisis Communication

Proponents of a U.S.-China AI dialogue recommend establishing a crisis-communication protocol to share information about emerging AI incidents in real-time. This would prevent crises from spiraling out of control. However, such a channel is challenging to operationalize, especially given China’s record of suspending crisis-communication channels during times of high tension.

Moreover, even if the protocol prevents immediate crises, it doesn’t address underlying issues driving the U.S.-China competition on AI. It’s a Band-Aid solution that fails to confront deeper power dynamics at play.

The True Stakes of the AI Arms Race

As diplomatic posturing and technical one-upmanship continue, it’s essential to remember the true stakes of the AI arms race. We’re not just discussing potential catastrophic incidents or loss of control over powerful systems; we’re also debating the fundamental nature of human existence in a world where technology increasingly shapes our destinies.

The current push for a U.S.-China dialogue on AI serves as a stark reminder that leaders are still grappling with implications of this new reality. Rather than pretending to cooperate while advancing their own interests, they should work towards a more nuanced understanding of the challenges ahead.

Ultimately, the failure of these talks will come from an inability to confront deeper realities driving the U.S.-China competition on AI. As we move forward, it’s essential that we prioritize a more honest and realistic approach – one that recognizes inherent complexities and tensions between two superpowers in a rapidly changing world.

The stakes continue to escalate, making clear that the current path of diplomatic efforts is unlikely to yield meaningful results. It’s time for our leaders to stop pretending and start confronting the true nature of the AI arms race.