- There’s no single definition for artificial general intelligence.
- Dario Amodei, the CEO of Anthropic, previously predicted that “powerful AI” could arrive by 2026.
- Amodei said AGI is more of a “marketing term,” and he has a different way of describing the next AI milestone.
Anthropic CEO Dario Amodei is confident that we’re rapidly approaching a new technology threshold — the creation of AI systems that are “better” than humans at most things.
He’s just not willing to pin the buzzy term “AGI,” or artificial general intelligence, on it.
“AGI has never been a well-defined term, for me,” Amodei said on a recent appearance on CNBC’s “Squawk Box.” “I’ve always thought of it as a marketing term. But, you know, the way I think about it is, at some point, we’re gonna get to AI systems that are better than almost all humans at almost all tasks.”
AGI is top of mind for various tech leaders, with many predicting that the technology is rapidly approaching viability.
The issue is there is no widely agreed-upon definition of AGI, much less a concrete timeline for reaching it. However, in general, many describe it as a form of AI that can meet or exceed human capabilities.
In Amodei’s case, he said he prefers to think about future advanced AI systems as a “country of geniuses in a data center.”
“It’s a sort of evocative phrase for all the power and all the positive things, and, you know, all of the potential negative things,” Amodei said. “That’s the thing I think we’re quite likely to get in the next two or three years.”
Amodei has previously written an essay on the subject of AGI, which he prefers to call “powerful AI.” In it, he laid out a series of parameters for the technology, including being “smarter than a Nobel Prize winner across most relevant fields.” In the same piece, he predicted it could arrive by 2026.
After a decade in the AI field, Amodei said he’s more confident than ever about such a timeline.
“Throughout the 10 years that I’ve been working in this field, I’ve always said, ‘You know, I don’t know for sure, this is the direction it seems like it’s going in, we could be getting to very powerful or human level systems,'” Amodei said.
“I still think there’s uncertainty, I think it’s important to be humble, but over the last six months, I would say that uncertainty, for me, has decreased a great deal,” he said.
As for recent concerns about the practicality of continued scaling of AI models, Amodei said he isn’t worried. He said there have only been “five or six” times throughout his career in AI where the tech appeared to be hitting a wall and that “something slightly different” was always invented that allowed for continued progress.
“The scaling of AI, it feels like this river,” Amodei said. “That, you know, every once in a while it runs into a stone, but it’ll — it always finds a way to go around.”
