Even the ‘Godfather of Artificial Intelligence’ has no idea what AGI is

Are you confused about general artificial intelligence, or AGI? That’s the kind of thing OpenAI is desperate to create, in a way that “benefits all of humanity.” Maybe they should be taken seriously, as they’ve just raised $6.6 billion to get closer to that goal.

But if you’re still wondering what AGI is, you’re not alone.

During a wide-ranging discussion Thursday at the Responsible Governance of Artificial Intelligence Summit, Fei-Fei Li, a world-renowned researcher often referred to as the “godmother of artificial intelligence,” said she didn’t know what AGI was either. Elsewhere, Lee talked about her role in the birth of modern artificial intelligence, how society needs to protect itself from advanced AI models, and why she believes her new startup, World Labs, will change everything.

But when asked what she thought of the “artificial intelligence singularity,” Li was just as confused as the rest of us.

“I come from an academic artificial intelligence background and I’m trained in more rigorous and evidence-based methods, so I really don’t know what all those words mean,” Lee told a packed auditorium in San Francisco, next to a large window overlooking the Golden Gate. “Honestly, I don’t even know what AGI means. As people say, you’ll know it when you see it, but I don’t think I’ve seen it. The truth is that I don’t spend much time thinking about those words because I think there are much more important things to do…”

If anyone should know what AGI is, it’s probably Fei-Fei Li. In 2006, she created ImageNet, the first large dataset for training and testing artificial intelligence, which was instrumental in driving the current boom in artificial intelligence. From 2017 to 2018, she was the Chief Scientist for Artificial Intelligence and Machine Learning at Google Cloud. Today, Li runs the Stanford Human-Centered AI Institute (HAI), and her startup, World Labs, develops “grand world models.” (That term is almost as confusing as AGI, if you ask me.)

OpenAI CEO Sam Altman attempted to define AGI in an interview with The New Yorker last year. Altman described AGI as “the equivalent of an average human being you could hire as a colleague.”

Meanwhile, OpenAI’s charter defines AGI as “highly autonomous systems that outperform humans in most economically valuable tasks.”

Apparently, those definitions weren’t good enough for a $157 billion company to work on. So OpenAI has created five levels that it uses internally to assess its progress towards AGI. The first level is chatbots (like ChatGPT), then logicians (supposedly OpenAI o1 was at this level), agents (that is supposedly coming next), innovators (artificial intelligence that can help with inventions), and the last level, organizational (artificial intelligence that can work for the whole organization).

Still confused? So am I, and apparently so is Lee. Also, this all sounds like a lot more than the average human colleague could do.

Earlier in the conversation, Lee said she’s been fascinated by the idea of ​​intelligence since she was young. This led her to study artificial intelligence long before it became profitable. In the early 2000s, she says she and a few others were quietly laying the groundwork for the field.

“In 2012, my ImageNet combined with AlexNet and GPUs — many people call it the birth of modern artificial intelligence. It was driven by three key ingredients: big data, neural networks and modern GPU computing. And when that moment came, I don’t think life was ever the same for the entire field of artificial intelligence, nor for our world.”

When asked about California’s controversial AI bill, SB 1047, Lee was careful to say that she didn’t want to reignite the controversy that Governor Newsom had just vetoed ad acta last week.

California’s governor recently hired Lee, along with other artificial intelligence experts, to form a task force to help the state develop rules for implementing artificial intelligence. Li said she takes an evidence-based approach in that role and will do her best to advocate for academic research and funding.

Although Lee advises California on AI regulation, she also runs her own startup, World Labs, in San Francisco.