High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / belief

Published · transcript-backed

Shawn Wang: belief

22 Aug 2024 Latent Space Is finetuning GPT4o worth it? — with Alistair Pullen, Cosine (Genie)

“I think one of the decisions before you as a CEO is how much you have like the house model be like the one true thing, and then how much you spend time working on customer models.”

— Shawn Wang

Source trail

Everything needed to verify it.

Speaker
Shawn Wang
Attribution
Verified speaker
Claim type
belief
Recorded
22 Aug 2024
Publisher
Latent Space

Transcript context

…we're gonna, we're gonna get really, we're gonna be the best in the world at doing that. Um, and continue being the best in the world at doing that. And throwing it as many models as we can. Um, seeing what the performance is like and seeing what things improve performance in what places. Um, and also making the data set larger is like one of the biggest things we're gonna be working on. I think one of the decisions before you as a CEO is how much you have like the house model be like the one true thing, and then how much you spend time working on customer models. That's the thing that really gets me so excited, genuinely. Like, we have a version of Genie. That we named after one of our employees. It's called the John. We have a version of Genie that is fine tuned on our code base. So we basically, it's the base, base Genie. And then we run the same data pipeline that we run on, like, all the stuff that we did to generate the main data set on our repo. And then all of a sudden you have, like, something that is both very good at software engineering, but is also extremely good at your repo. And that is phenomenal to use. Like, it's really cool. More…

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

Search evidence