Anthropic CEO Dario Amodei addressed criticism this week about how he talks about artificial intelligence in public. The comments came after investor Gavin Baker argued that Amodei's warnings have hurt the AI industry.
Baker made his case on the All-In podcast and later on X. He said Amodei's focus on AI risks has helped spark backlash against the technology, including opposition to data centers.
Baker also said Amodei has "lost the argument" on AI regulation. He suggested Amodei should act as a more positive voice for the industry he leads.
Amodei responded in a series of posts. He disagreed that his messaging has been overly negative.
He said his writing has been "about equally balanced" between the risks and benefits of AI. He pointed to his essay "Machines of Loving Grace" as an example.
Amodei said he wrote that essay because he felt the AI industry wasn't painting an inspiring enough picture of the technology's potential.
He did agree that the public holds a negative view of AI. He called this "a big problem."
A Crisis of Trust
Amodei said the backlash isn't mainly caused by AI leaders warning about risks. He offered a different explanation.
"I think it is fundamentally a crisis of trust," Amodei said. He said ordinary people don't trust companies, governments, or the tech industry in general.
He said people often suspect these groups are working against their interests. Amodei framed the AI backlash as part of a much older pattern.
He said the most accurate criticism of AI companies, including Anthropic, is that they haven't delivered on their big promises yet. He said that responsibility falls on the companies themselves.
As an example, he said promising that AI will cure cancer is a cliche. He said actually curing cancer is what would change public opinion.
Amodei's Response on Regulation
Amodei also responded to Baker's claim that AI regulation leads to concentrated corporate power. He said this framing presents a false choice.
He said this view is common in Silicon Valley but is an oversimplified way of looking at the issue. He said many people outside that world see regulation as something that limits corporate power and helps ordinary people.
Amodei said Anthropic writes its policy proposals carefully. He said the goal is to slow down large AI companies while giving smaller competitors more room to compete.
He explained that AI tends to concentrate power by nature. He said open weight models don't fully fix this, since power then shifts toward whoever has the most computing resources and chips.
Amodei said the right regulations could do three things at once. They could address AI safety risks, limit the power of large AI firms, and still leave room for open weight models.
