OpenAI o3, every mention

mentionsepisodesper episode tap a year for its mentions

Verbatim, from the transcripts: the passages where OpenAI o3 comes up

loading…

▶ 10:11 Francois Chaubard This starts from, like, literally Tagula-Rasa weights, and it can outperform at that time, if we go back, you know, we had, um, O-three, if you remember back, way back when. 2 times in the scene

▶ 18:51 François Chollet In particular, the OpenAI O-one and then O-three models, which by the way, they were demonstrated by OpenAI on Arc, because it was the one unsaturated reasoning benchmark that was really showing that this model was different. 2 times in the scene

▶ 10:32 Calvin French-Owen For, uh, I think it was O three, like one of the reasoning models.

▶ 37:47 Garry Tan This is like actually a really powerful moment because literally you have a hundred, you know, I think of, uh, O-three as basically about a 130 IQ, maybe O-three pro can be even smarter than that. 2 times in the scene

▶ 3:31 François Chollet OpenAI previewed its, ah, all three model, and they used a version of it that was, ah, fine-tuned specifically on Arc, and that showed human-level performance on that benchmark versus time.

▶ 4:59 Garry Tan What surprises you the most with the latest models like O-three and what emergent behaviors or use cases are standing out to you right now? 2 times in the scene

▶ 27:47 Diana Hu And you tried it with, uh, O-three versus Gemini 2.5, and you found discrepancies. 3 times in the scene

▶ 11:45 Bob McGrew The thing that's been really exciting, of course, is, you know, going from, ah, the LLM scaling of pre-training, where you're just bringing bigger and bigger corpuses and trying to predict the next token, and shifting gears, and using…
▶ 13:56 Garry Tan It's, uh, you know, reasoners are here, and then, uh, I'm hearing a ton about innovators, so taking a thing like, uh, O-three or, you know, maybe when O-three pro comes out, that'll be a real moment where you can, uh, hook that up to a bio… 2 times in the scene

▶ 17:54 Jared Friedman Sam was just telling us that like O-two and O-three are not far behind.
▶ 32:11 Diana Hu Because they're getting, they might just get deprecated from the improvements of O-one, O-two, O-three.

Turn any episode into a week of clips.