OpenNashPodcasts

Person

Fei-Fei Li

1 episodes · 8 quoted claims

Fei-Fei Li's Stanford lab announced an open dataset and benchmark called Behavior for robotic learning in simulated environments.

“my Stanford lab just announced an open dataset and benchmark called Behavior, which is for benchmarking robotic learning in simulated environments.”

Latent Space: The AI Engineer Podcast · 0:05:00

Li says she worked with the first Trump administration on a bill to scope a national AI compute cloud and data repository.

“We work with the first Trump administration on this bill called National AI Resource Resource, NAYER bill, which is scoping out a national AI compute cloud as well as data repository.”

Latent Space: The AI Engineer Podcast · 0:07:22

Li says her lab's image captioning work was being done at the same time by Google, independently.

“We thought we were the first people doing it. It turned out that Google at that time was also simultaneously doing it.”

Latent Space: The AI Engineer Podcast · 0:15:30

New York Times reporter John Markoff, covering Google's captioning work, learned of the Stanford work and wrote about both.

“it was John Markoff from New York Times, was breaking the Google story. But he by accident heard about us.”

Latent Space: The AI Engineer Podcast · 0:16:02

Li contrasts video generators, which lack exact camera control, with Marble, which lets users place a camera precisely in its 3D world.

“you cannot say pan 63 degrees to the north, right? You just don't have that control. Whereas in marble, you have precise control in terms of placing a camera.”

Latent Space: The AI Engineer Podcast · 0:33:54

Li credits psychologist Howard Gardner with describing multiple intelligences, including spatial intelligence, though she places it in the 1960s.

“There is a psychologist, I think his name is Howard Gardner, in the 1960s, actually literally called multiple intelligence to describe human intelligence.”

Latent Space: The AI Engineer Podcast · 0:42:36

Li says nature took about 540 million years to optimize perception and spatial intelligence.

“in nature, you know, it took 540 million years to optimize perception and spatial intelligence and language.”

Latent Space: The AI Engineer Podcast · 0:48:43

Li gives half a million years as a generous estimate for the development of language, far shorter than for perception.

“The most generous estimation of language development is probably half a million years.”

Latent Space: The AI Engineer Podcast · 0:48:43