Dario Amodei says AI's trust problem will be solved by results, not marketing
He argued AI structurally concentrates power, defended designing rules to slow frontier labs while exempting smaller rivals, and said the public's distrust reflects a broader crisis, not his warnings.
- Ideas & essays
- Government & policy
- Notable
In a two-part thread on X, Dario Amodei responded to a critic who had argued that his public messaging on AI was disproportionately negative and that regulation would concentrate power in a few hands. He rejected the second point as a “false choice”, arguing that fair institutional processes can decentralise power rather than hoard it — that at their best, institutions “vest power in ideas rather than people.”
On regulation, Amodei said Anthropic deliberately designs its proposals to slow frontier developers while advantaging smaller competitors. He noted that California’s SB 53 exempts companies below $500m in revenue, and that the testing regimes Anthropic has advocated at the US CAISI and the White House — and the Pacing the Frontier letter — apply stricter scrutiny to frontier models than to those catching up, which he argued differentially disadvantages the leading labs. His underlying claim was that AI structurally concentrates power because of the scaling laws, and that open weights only partly offset this by shifting concentration toward whoever holds the most compute and chips.
On public perception, Amodei denied that his own record had been one-sided, saying he had written one major essay about each side — Machines of Loving Grace making the optimistic case and ‘The Adolescence of Technology’ setting out the risks. He argued the public’s negative view of AI is fundamentally a “crisis of trust” in companies and institutions that predates the technology, and that a marketing campaign would not repair it:
The thing that will work is actually curing cancer. I think by far the most accurate criticism of AI companies including Anthropic is that we haven’t yet delivered on our big promises to benefit the world. That is totally on us.
He disclosed that he had lost his father to Hepatitis C a few years before the arrival of antivirals that cure most cases, and said Anthropic was accelerating its work in biology and medicine, promising “early glimmers in the coming months.” The thread also registered his support for the Trump administration’s reported move toward pre-deployment testing of frontier and near-frontier models, and for Demis Hassabis’s proposal for a FINRA-style oversight body.