I met China’s top AI experts. They are also coming out

Just beyond a A week ago, I went to the Major artificial intelligence Meeting in Zhongguancun, Beijing’s most advanced district.

It was full of interesting sessions about everything from self-revolution—the idea that models can change their code and progress indefinitely—to humanoid robots. And it featured several computer legends, including Whitfield Diffie, the co-inventor of private keys, and Andrew Barto, who won the Turing Award and Rich Sutton for his pioneering work in reinforcement learning.

But I came away with one thing more than anything else: the US and China need to put their fierce AI competition aside.

Frontier AI’s threats to cybersecurity and systemic risks are too serious to ignore, and the increasingly advanced models could cause chaos unless national AI authorities work together. “AI is a global technology with global benefits, global harms, and a constant tendency for new possibilities to multiply,” Stephen Caspera computer scientist at MIT who spoke at the conference via video, told me later.

Until now, the US has largely viewed China’s AI advances as a threat to its economy and national security. Washington has imposed tough restrictions on chips and chip-making equipment to hamper the country’s development of powerful AI. Recently, the US government he ordered Anthropic preventing aliens from accessing its most powerful species, Mythos and Fable 5, continued national security concerns. In response, Anthropic blocked access to everyone. One company that was particularly worrisome, WIRED previously revealed, was a The South Korean telecom giant is said to be cooperating with China.

But the conference, organized by the Beijing Academy of Artificial Intelligence, promoted the idea that the US and China will lose out if AI is developed too quickly and carelessly. As AI becomes more powerful, more sophisticated, and more integrated into everyday life, the risks that it can be used to launch cyberattacks or fail in dangerous ways will only grow. Because the world’s two leading AI powers are the most sophisticated, cooperation between them seems inevitable.

Casper pointed research showing that the benefits of international cooperation on AI risks outweigh any national security risks that come from working together. He compared the situation today to how the US and the Soviet Union were forced to work together on nuclear threats, even though they only wanted to sell each other.

“One thing that almost everyone in AI can agree on right now is that AI does not need a Chernobyl moment,” Casper said.

One full-day session featured cyber challenges created by advanced AI. This includes new vulnerabilities in the Is it made by AInew attack methods supported by operational tools, and automated methods social status attack.

After another session, I spoke with Lin Yun, a professor at Shanghai Jia Tong University it works very well on AI and computer security. Yun told me that he expects hackers to gain an advantage in the short term, but new techniques, including the use of AI, should pay off in the long run.

Yun said that even if international cooperation is difficult due to competition, it should be a priority. “If different countries understand the threat in similar ways, it will be easier to implement shared security policies and technology,” he told me. “The key is to find areas where sharing can reduce system risk without revealing too much of the work.”

Perhaps the most important question for both countries is how to balance flexibility and risk. Heavy duty open models have become the focus of research and innovation, with Chinese models becoming popular in the US. But as these models advance, it will become increasingly difficult to ensure that they do not help hackers identify security vulnerabilities and cannot be used as cyber tools.