The United States and China are set to open a bilateral dialogue on superintelligence. Scott Bessent, the U.S. Treasury Secretary, confirmed this at the gala marking the 60th anniversary of the National Committee on U.S.-China Relations, sending a message of détente to Beijing: “reduce risks, not decouple.”

The dialogue on artificial intelligence will have three components:

🔹 Exchange of best practices for evaluating advanced models before launch.

🔹 Ensuring that human judgment remains at the center.

🔹 A direct channel for notifying each other of serious incidents involving their models.

Bessent framed it by citing Henry Kissinger’s concern about superintelligence. And that wasn’t the only thing he announced:

🔹 A Trade Board that will seek more favorable tariff terms for $30 billion worth of non-sensitive goods in each direction, with an agricultural working group.

🔹 China commits to importing at least 25 million tons of U.S. soybeans and 10 million tons of U.S. coal over two years.

🔹 An Investment Board as a forum for companies from both countries to discuss market access.

All of this is underpinned by the personal relationship between Trump and Xi: Beijing in May, Washington in September, Shenzhen next week, and a fourth meeting in Miami. According to Bessent, no two leaders from the countries had ever met so many times in a single year.

In my view, the most significant development is the incident channel. The fact that the two powers leading the AI race have agreed to notify each other when a model fails is a step that seemed unthinkable until recently. The technological competition continues, but it is beginning to have some basic shared rules—and that also affects the entire ecosystem combining AI and blockchain.

Full analysis at territorioblockchain.com

#IA #Superinteligencia #China #USA