The discussion centers on Xi Jinping's first in-person WAIC appearance and his most extensive AI statement, which emphasized 'secure and orderly development' and signaled China's intent to lead on AI safety and governance, particularly for the Global South. The launch of the World AI Cooperation Organization (WAICO) with 29 founding members offers concrete deliverables like AI training slots, though skepticism about follow-through remains. The hosts analyze Moonshot's Kimi K3, the fourth frontier-class Chinese model released in a month, highlighting the closing gap with U.S. labs despite export controls. They explore the U.S. push to restrict Chinese open-weight models, noting the irony that many American companies rely on them, and the challenge of finding effective policy tools. The conversation also covers the rise of AI agents, which are upending data protection regimes and super-app models, and the complex meaning of 'ānquán kěkòng' (secure and controllable). Finally, they preview the upcoming U.S.-China AI dialogue, emphasizing the need to break a negative securitization feedback loop and address China's exclusion from programs like Mythos, all while the rapid pace of recursive self-improvement in AI makes this a critical moment for cooperation.