Please follow us on Truth Social, X, YouTube, Minds, Telegram, Rumble, GETTR, Gab, Instagram
The 3 most powerful men in Silicon Valley – Dario Amodei (Anthropic), Sam Altman (OpenAI), and Elon Musk – seem to agree on one thing today: slow down AI. Strange, given that the former head of AI policy at the White House calls all of their various rationalizations for industry-wide slowdowns 'bulls**t'.
In his lengthy essay Dario Amodei calls on the leading companies in the AI space to "pace the frontier" by "stopping the race to build ever more powerful models as fast as possible." This can be done in three steps, Amodei claims: 1) allow third-party safety evaluators to have employee-level access to the AI systems of a company; 2) work towards coordinated safety standards for the leading AI labs in the West; and 3) work towards eventually also coordinating with governments from around the world, including those from so-called "authoritarian" regimes.
Sam Altman followed within hours of the essay's publication on X. He stated that pacing advanced capabilities has been "a primary topic at OpenAI for some time now" and that the company is "opening up our systems to independent evaluators, just as Anthropic is doing with Dario's plan for independent safety evaluation here." Elon Musk followed minutes later, commenting "Dario is right" – remarkable given the CEO has previously claimed that Anthropic "hates Western Civilization" and is "basically a robot cult."
This is all coming as Anthropic readies for a potential IPO that could be worth trillions of dollars. OpenAI, meanwhile, will not go public this year, Sam Altman said, because an IPO would be "ill-advised right now." Two companies that have burned through capital for years, racing each other to build ever more powerful models of AI, are now saying that they should all slow down. And that they should all lock in their current market positions.
White House AI ex-czar David Sacks issued a withering takedown of the trio's stodgy attempt to portend virtue. "Stop pretending you need to suspend antitrust law in order to create a cartel. Stop pretending METR is independent of Anthropic when they are literally intertwined with Anthropic's investors and staff. And stop pretending you need these same people to police your competitors who are not even at the frontier of AI."
Sacks cuts to the heart of the matter. He notes that the required evaluators that Amodei is pushing for are deeply embroiled in Anthropic's funding network, and that the "safety" standards which Altman is suddenly advocating for will be the biggest burden to the small startups trying to reach the frontier, not to OpenAI and Anthropic who are already there and will be writing the rules for everyone else.
Note that the two companies – OpenAI and Anthropic – are also now at risk of enormous product-liability damage if they ever happen to allow a single cyberattack to occur via one of their models. So it is simply good business for them to slow down, in order to try to prevent as many uncontrolled outputs from occurring as possible.
Treasury Secretary Scott Bessent put the real stakes plainly at a Breitbart event last week: "Beating China — there is no day after tomorrow if China wins at this." That framing is conspicuously absent from Amodei's essay, which calls for coordinating with authoritarian governments on AI development rules. China does not coordinate. China steals, copies, and races. Any voluntary slowdown by American companies is a competitive gift to Beijing, not a safety measure.
Facebook famously called for regulation as well after it had established market dominance.
This is NOT philanthropy. The market consolidation in the AI safety space is real — and it will be dressed up in language that sounds like it could save humanity. David Sacks is right to call it out. The question is whether Washington is paying attention.
