Roon, an OpenAI researcher turned de facto company spokesman on Twitter, pushed back hard against calls for unilateral pauses in AI development. He argues that sidelining any single lab would barely dent overall progress—talent and compute would just shift elsewhere, and safety-focused teams would lose influence and funding. Geoffrey Irving of the UK’s AI Security Institute counters that “magic” arguments about scaling AI have fallen apart; real-world advances came without mystical steps, so we need concrete, coordinated plans for a global slowdown rather than hand-waving. The piece lays out five practical moves labs can make to support a collective pause. First, publicly pledge to slow down only if every competitor does the same, even at legal or financial risk. Second, lobby governments for a formal agreement. Third, use public outreach to build popular backing—maybe even slip slogans into chatbot outputs. Fourth, assemble an internal task force to draft the technical and verification details. Finally, include “merge-and-assist” clauses to share compute and talent with compliant peers. Then came “Pacing The Frontier,” an open letter signed by over a thousand employees across OpenAI, Anthropic, Meta and other frontier labs. It urges the U.S. government to spearhead international efforts to develop tools and governance for deliberately pacing AI breakthroughs. Notable signers include chief scientists from OpenAI and Anthropic, several DeepMind co-founders, and the originators of Claude. Dario Amodei joined just before publication; Sam Altman hasn’t signed but OpenAI’s corporate account retweeted the letter, and Altman recently hinted in an interview that “we may have to pace the rate of AI development.” No public support has come from Demis Hassabis, Elon Musk or Mark Zuckerberg yet. Some suspect OpenAI and Anthropic quietly orchestrated the letter to look grassroots, keeping CEOs off-the-record until release. Whatever the inside story, this marks a shift: major players are treating paced development as plausible policy rather than radical protest. It gives pause advocates a shared reference point for policy debates and signals to everyone inside these labs that safety-first isn’t just fringe rhetoric.