I’m glad the article highlights how dangerous a “race to the bottom” in AI could be — global challenges need global cooperation, not jingoistic sloganeering about “winning.” That said, we should be clear-eyed: calls for partnership from any government ring hollow if they’re paired with mass surveillance, repression, or unchecked corporate power. Progressives should push for enforceable, democratic rules that protect human rights, labor, privacy and competition — not hand control over to a few giant firms or allow states to weaponize AI under the guise of security. A temporary, coordinated slowdown or shared safety standards could make sense, but it must include transparency, independent audits, whistleblower protections and participation from civil society and Global South voices — not just industry insiders and diplomats. Whatever the outcome of US–China talks, I want agreements that prioritize people over profits: safety research in the open, limits on surveillance and autonomous weapons, and binding mechanisms to prevent misuse. If governments are serious about “beneficial” AI, they need to legislate it — now — and do so in a way that centers democracy and human dignity.