High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / commitment

Published · transcript-backed

David Luan: commitment

22 Mar 2024 Latent Space Why Google failed to make GPT-3 + why Multimodal Agents are the path to AGI — with David Luan of Adept

“Like I will die on this hill, but I mean, like again, this whole time, like we've been on this podcast, we're just going to continually saying these models are basically behavioral cloners.”

— David Luan

Source trail

Everything needed to verify it.

Speaker
David Luan
Attribution
Verified speaker
Claim type
commitment
Recorded
22 Mar 2024
Publisher
Latent Space

Transcript context

…And if I told you, hey, fire your entire warehouse workforce and, you know, put robots in there, isn't that a strategy? Oh yeah. Yeah. Sorry. I'm not questioning whether they're doing smart things. I genuinely don't know what they're doing as much, but I think there's two things. One, I'm so excited for someone to train a foundation model of robots. It's just, I think it's just going to work. Like I will die on this hill, but I mean, like again, this whole time, like we've been on this podcast, we're just going to continually saying these models are basically behavioral cloners. Right. So let's go behavioral clone all this like robot behavior. Right. And then you figure out everything else you have to do in order to teach you how to solve a new problem. That's going to work. I'm super stoked for that. I think unlike what we're doing with helping humans with knowledge work, it just sounds like a more zero sum job replacement play. Right. And I'm personally less excited about that. We had a Ken June from InBoo on the podcast. We asked her why people should go work there and not at Adept.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.

Search evidence