Dario Amodei, CEO of Anthropic, said that the negative view of artificial intelligence fundamentally reflects a crisis of trust between the public, companies, governments, and the technology sector, rather than being a direct result of the warnings issued by industry leaders. He made the remarks in response to investor Gavin Baker, who argued that Amodei’s warnings about the risks of artificial intelligence had helped fuel a backlash in the United States, particularly toward data centers.
Baker had criticized Anthropic’s support for certain forms of regulation on the All-In program and on X, including a California bill imposing transparency requirements on large AI companies. He said that Amodei, given his leadership of a company likely to become one of the world’s most important, should present a more positive image of his sector, arguing that he had lost the debate over AI regulation.
Amodei Rejects Blaming Warnings for Skepticism
Amodei responded that his messages had not been disproportionately negative, noting that they were distributed roughly evenly between the risks and benefits of artificial intelligence. He added that he wrote his essay “Machines of Loving Grace” because he did not believe the industry was presenting an inspiring enough picture of the technology’s ability to change the world for the better.
While acknowledging that the overall view of artificial intelligence is negative and that this represents a major problem, Amodei rejected the claim that this negativity is primarily caused by warnings from industry leaders. According to his explanation, ordinary people do not trust companies, governments, or the technology sector, and tend to believe that these entities may develop a new way to harm them. Amodei sees the backlash against artificial intelligence as the latest manifestation of a crisis of trust that has developed over decades.
What Does Amodei See as a More Accurate Criticism?
Amodei said that the more accurate criticism of AI companies, including Anthropic, is that they have not yet delivered on their major promises to benefit the world. He added that responsibility lies with the companies themselves, and that broad promises, such as saying that artificial intelligence will cure cancer, have become closer to clichés than sources of inspiration. In his view, what may change people’s attitudes is delivering tangible results, not improving marketing or media messaging.
Regulation Between Limiting Risks and Concentrating Power
Amodei also rejected portraying the regulatory debate as a binary choice between deploying artificial intelligence without restrictions or confining the technology to a small number of companies because of government rules. He said that the common idea in Silicon Valley equating regulation with regulatory capture and the concentration of power oversimplifies reality, since many people outside that circle see regulation as limiting corporate influence and protecting individuals.
He explained that Anthropic is carefully developing its policy proposals with the aim of slowing leading AI companies while giving smaller companies a relative advantage. Amodei believes that artificial intelligence is a technology that, by virtue of its structure, tends to concentrate power, while open-weight models only partially help address that problem because they shift some of the concentration toward entities with the greatest computing and chip capacity.
According to Amodei, appropriate “rules of the road” can address cybersecurity, biological, and alignment risks, and institutionally limit the power of leading AI companies while keeping the field open to open-weight models and addressing their particular risks. This view places the essence of the disagreement in how to build trust and deliver practical benefits, not merely in choosing a more optimistic or pessimistic tone.