GPT-4, every mention

mentionsepisodesper episode tap a year for its mentions

Verbatim, from the transcripts: the passages where GPT-4 comes up

loading…

▶ 17:14 Sam Parr You said that the GLP ones were going to be bigger than GPT four.

▶ 32:31 Sarah Guo I think like a lot of your listeners will have seen the GPT for demo, where it's like a voice that may or may not sound like ScarJo talking like in real time.

▶ 32:20 Brett Adcock You just gotta like, you know, find this stuff, whether it's Wikipedia papers, uh, like Google searches, uh, you can use GPT-IV.

▶ 41:02 Shaan Puri If we make GPT five better than GPT four and we improve a bunch of capabilities, the model gets smarter.

▶ 47:35 Shaan Puri So, so chat GPT now, GPT four, it's, it can do it. 3 times in the scene

▶ 1:11:00 Martin Shkreli This is an, is a, is an attraction of GPT-IV. 2 times in the scene

▶ 18:12 Dharmesh Shah So GPT, uh, three and now four is, is a reasoning engine.
▶ 23:56 Dharmesh Shah It wasn't, I mean, sure things are moving at a fast pace, but that's not, I don't think there's a causal effect that all of a sudden the number of bunkers has gone up by 800% simply because GPT-IV was launched.
▶ 25:31 Dharmesh Shah We're now doing multimodal with GPT four. 3 times in the scene
▶ 42:51 Dharmesh Shah Lane chain helps you kind of chain those things together and makes it easier for you to kind of work with either an individual large language model, like GPT-IV,
▶ 1:12:07 Dharmesh Shah Front engineering is almost exactly the same thing, except you're talking to a large language model, something like a GPT-IV to kind of get it to produce.

Turn any episode into a week of clips.