AI Regulation, Robotics Bans, and the OpenRouter Deal

The Trump administration has banned Chinese humanoid and quadruped robots by adding them to the FCC's "covered list," citing vulnerabilities that could allow attackers to manipulate both data and physical operations. This kicked off a wide-ranging discussion among four venture investors about where the line between national security and regulatory capture falls, whether AI researchers' latest calls to slow development are sincere or performative, and why Stripe might pay roughly $10 billion for a company doing $140 million in annualized revenue.

The Chinese Robotics Ban: National Security or Negotiation Tactic?

The FCC's rationale focuses on network vulnerabilities in advanced robotic systems, but the panelists saw multiple layers beneath the surface. David Weisberg framed the ban primarily as a negotiation posture ahead of Xi Jinping's September visit to the White House, arguing the Trump administration is "anchoring high" as part of a comprehensive package with China. Jason Calacanis took a harder line, arguing the danger is real even if the probability is low. He described watching combat robotics in Tokyo where four-foot-tall Unitree robots fought each other, controlled by humans with gamepads, and noted that putting an LLM inside such hardware with open-ended instructions changes the risk calculus entirely. "You