Anthropic CEO Dario Amodei said in a recent interview that he feels "calm" about the current state of competition between Anthropic and OpenAI, despite the brewing cold war between the world's two largest artificial intelligence companies. However, he couldn't help but take a swipe at OpenAI CEO Sam Altman.
In an interview that aired on Wednesday, Amodei took an apparently pointed swipe at Altman, saying: "At the end of the day, if you and the other party don't share the same ideas and don't trust each other, why bother arguing with them? The way to resolve differences is to go our separate ways. We do ours and they do theirs. I'm totally fine with us sticking to our own path and the other side going theirs."
Ultimately, Amodei said, someone will judge who was right and who was wrong.
"We'll see who wins in the market and we'll see who wins in public opinion," he said. "I think those are more persuasive than any argument about who left and why."
The story of Amodei, his sister Daniela, and nine former OpenAI employees who left OpenAI in 2020 to found Anthropic has now become a "legend" in Silicon Valley. Anthropic, once a weak player, is now widely considered to have surpassed OpenAI in the race for generative artificial intelligence.
Amodei's departure sparked renewed concern after Ronan Farrow published an exposé in The New Yorker exploring whether Altman could be trusted. The report repeatedly cited notes Amodei took while interacting with Altman during his tenure at OpenAI. A few days after the report was published, Altman's home was attacked. He partly blamed the incident on the New Yorker report (without naming him directly), and later pointed the finger at Anthropic's remarks about OpenAI.
“I don’t think the doomsday talk is helpful. I don’t think the way some of the other labs are talking about us is helpful,” Altman said on a podcast that aired in April, adding, “I don’t think the way Anthropic is talking about OpenAI is helpful either.”
Amodei blames the most famous incident of his rivalry with Ultraman — when the two refused to join other industry leaders in unifying their stance — on the “extremely disorganized” nature of the Indian Artificial Intelligence Summit. The Anthropic CEO said other summits attended by world leaders tend to be a mess.
"Look, I don't know what to tell you, okay? It's like Narendra Modi suddenly appeared on stage and asked everyone to hold hands," Amodei said of the Indian prime minister, who was standing next to Ultraman and Amodei.
When asked what it would take to convince the world that AI companies will cooperate on major AI safety issues, Amodei said the key is not whether everyone gets along, but who sets the agenda.
“I think what needs to happen is that the trustworthy players need to come together and force the untrustworthy players to adopt the same standards,” he said. “With my vast experience, I’ve seen that there are people who won’t take the initiative to do the right thing, but if the majority of the industry is doing the right thing, then I think it’s going to be hard for the remaining players to make a difference.”
Amodei made it clear that the notion of "mutual distrust" in the field of artificial intelligence was not true, citing his relationship with Google DeepMind CEO Demis Hassabis as an example.
"I've known him for 15 years. We've worked together on many issues," Amodei said. "We buy computing resources from Google. We often exchange ideas on security. So my view is that on the one hand, some players are more trustworthy than others; on the other hand, I think there are some players outside of Anthropic that I trust and are trustworthy."