There is no single definition of artificial general information. Humanity CEO Dario Amodei previously predicted that “powerful AI” could arrive by 2026.
Humanity CEO Dario Amodei seems convinced that we are rapidly approaching a new technological threshold: the creation of AI systems that are “better” than humans.
He is more than willing to pin the term “Agi” or artificial general information dangling over it.
“For me, AGI has never been a well-defined term,” Amodei said in a recent appearance on CNBC’s “Squawk Box.” “I’ve always thought of it as a marketing term. But the way I think about it is that at some point we’ll end up with an AI system that’s better than just about any human at almost every task.”
AGI is top of mind for various technology leaders, and many predict this technology is rapidly approaching.
The problem is that there is no widely agreed definition of AGI, and far fewer concrete timelines for reaching it.
In general, many describe it as a form of AI that can meet or exceed human capabilities. In Amodei’s case, he said he preferred to think of future advanced AI systems as a “land of data center geniuses.”
“It’s kind of an evocative phrase for all the power and all the positive things, and you know all the potential negative things,” Amodei said. “That’s something I think we’re very likely to get over the next two to three years.”
Amodei has previously written essays on the subject of AGI, which he prefers to call “strong AI.” In it, he laid out a set of parameters for the technology, including being “smarter than a Nobel Prize winner in most relevant fields.” In the same work, he predicted that it could arrive by 2026.
After a decade in the AI field, Amodei said he is more confident than ever about such a timeline.
“Throughout the 10 years I’ve worked in this field, I’ve always said: powerful or human-level systems,” Amodei said.
“I think there’s still uncertainty. I think it’s important to be humble, but I think the uncertainty has decreased significantly for me over the last six months,” he said. .
Regarding recent concerns about the practicality of continuous scaling of AI models, Amodei said he is not worried. He said there have only been “five or six” times throughout his career in AI. There, he said, technology seemed to hit a wall and “something a little different” was always invented to enable continued progress.
“Scaling AI, it feels like this river,” Amodei said. “You know it hits a rock every once in a while, but it always finds a way to turn.”
