High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / uncertainty

Published · transcript-backed

Shawn Wang: uncertainty

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

“I was kind of excited at Adept experiments or Adept workflows, I don't know what the official name for it is.”

— Shawn Wang

Source trail

Everything needed to verify it.

Speaker
Shawn Wang
Attribution
Verified speaker
Claim type
uncertainty
Recorded
22 Mar 2024
Publisher
Latent Space

Transcript context

…st a totally different pre-training corpus. And so a depth spent a lot of time building that. And so the public for use and stuff aren't trained on our actual corpus, it's trained on some other stuff. But you take a lot of that data and then you make it really fast and make it really good at things like dense OCR on screens. And then now you have the right like raw putty to go make a good agent. So that's kind of like some of the modeling side, we've kind of only announced some of that stuff. We haven't really announced much of the agent's work, but that if you put those together with the correct product form factor, and I think the product form factor also really matters. I think we're seeing, and you guys probably see this a little bit more than I do, but we're seeing like a little bit of a pushback against the tyranny of chatbots as form factor. And I think that the reason why the form factor matters is the form factor changes what data you collect in the human feedback loop. And so I think we've spent a lot of time doing full vertical integration of all these bits in order to get to where we are. Yeah. I'll plug Amelia Wattenberger’s talk at our conference, where she gave a little bit of the thinking behind like what else exists other than chatbots that if you could delegate to reliable agents, you could do. I was kind of excited at Adept experiments or Adept workflows, I don't know what the official name for it is. I was like, okay, like this is something I can use, but it seems like it's just an experiment for now. It's not your product. So you basically just use experiments as like a way to go push various ideas on the design side to some people and just be like, yeah, we'll play with it. Actually the experiments code base underpins the actual product, but it's just the code base itself is kind of like a skeleton for us to go deploy arbitrary cards on the side.…

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

Search evidence