Elon Musk offered a blunt reaction to the AI safety debate early Sunday. He wrote that he hopes AI is nice to us, echoing a stark warning from investor Naval Ravikant.
The post landed just a minute after Musk commented “Interesting exchange” on a thread previously posted by Anthropic CEO Dario Amodei, who was defending his AI risk warnings against claims that his messaging leans too negative.
Why Amodei Defends His AI Risk Warnings
The thread began with investor Gavin Baker. The Silicon Valley fund manager argued that rumors about Amodei spread easily because they matched the CEO’s public tone.
Amodei rejected the idea that he mostly talks about danger. His two major essays split evenly between risks and benefits, the Anthropic chief said. Moreover, short interview clips skew negative because they earn clicks, he argued.
Machines of Loving Grace, his case for AI’s upside, exists because the industry painted too bleak a picture, the CEO said.
Amodei also pointed to a deeper cause. Ordinary people distrust companies, governments, and the tech industry, according to the Anthropic chief. Consequently, AI inherits a suspicion that has hardened over decades.
That view shapes his policy work. In June, Amodei demanded mandatory frontier model testing instead of voluntary pledges. Days later, the Anthropic chief used his G7 unity appeal to warn against a fractured global approach.
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.
Musk Weighs In as the Cure Promise Returns
Musk called the exchange interesting, thereby drawing attention to the complexity of the debate. In addition to Amodei’s regulatory focus, the issue involves a philosophical side that investor Naval Ravikant neatly captured: you cannot create God and put him on a leash.
Musk directly echoed this sentiment with his own response, writing, “I hope AI is nice to us”, a laconic acknowledgment that if human control over a superintelligence fails, society will ultimately have to rely on the benevolence of its own creation.
Musk has praised Anthropic before. In July, the Tesla CEO blamed OpenAI for closing up while singling out Amodei for a different path.
Amodei also attacked a familiar Silicon Valley framing. Regulation does not automatically concentrate power, he said. California’s SB 53, an AI safety law backed by Anthropic, exempts companies with less than $500 million in revenue.
Similar tensions surfaced in June, when Vitalik Buterin criticized rising AI nationalism in Washington.
Amodei then named two approaches he supports. One is the Trump administration’s reported plan to test frontier models before release. The other is a watchdog modeled on FINRA, the US broker-dealer self-regulatory organization, proposed by Google DeepMind chief Demis Hassabis.
Finally, Amodei returned to his boldest promise. AI applied to biology could cure most human diseases within five to 10 years, he wrote. Anthropic and Pfizer already treat AI in healthcare as core infrastructure.
The Anthropic CEO tied that urgency to his father, who died of Hepatitis C shortly before curative antivirals reached patients.
Yet Amodei dismissed calls for a glossy marketing push. Saying AI will cure cancer reads as a cliché now, he argued. Actually, curing it is the only fix.
Therefore, Anthropic’s next months in biology may test whether that pledge survives the lab.
Bitrue