Alessio Fanelli: evaluation

Transcript context

…If you're using the cholesterol analogy, then my, then my egg thing can't be that far off. Let's talk about the sort of the state of the art and the, and the GPT 4 class landscape and how that's changed. Cause obviously, you know, sort of the, the two big things or a couple of the big things that have happened. Since we last talked, we're one, you know, Gemini first announcing that a model was coming and then finally it arriving, and then very soon after a sort of a different model arriving from Gemini and and Cloud three. So I guess, you know, I'm not sure exactly where the right place to start with this conversation is, but, you know, maybe very broadly speaking which of these do you think have made a bigger impact? Thank you. Probably the one you can use, right? So, Cloud. Well, I'm sure Gemini is going to be great once they let me in, but so far I haven't been able to. I use, so I have this small podcaster thing that I built for our podcast, which does chapters creation, like named entity recognition, summarization, and all of that. Cloud Tree is, Better than GPT 4. Cloud2 was unusable. So I use GPT 4 for everything. And then when Opus came out, I tried them again side by side and I posted it on, on Twitter as well. Cloud is better. It's very good, you know, it's much better, it seems to me, it's much better than GPT 4 at doing writing that is more, you know, I don't know, it just got good vibes, you know, like the GPT 4 text, you can tell it's like GPT 4, you know, it's like, it always uses certain types of words and phrases and, you know, maybe it's just me because I've now done it for, you know, So, I've read like 75, 80 generations of these things next to each other. Clutter is really good. I know everybody is freaking out on twitter about it, my only experience of this is much better has been on the podcast use case. But I know that, you know, Quran from from News Research is a very big opus pro, pro opus person. So, I think that's also It's great to have people that actually care about other models. You know, I think so far to a lot of people, maybe Entropic has been the sibling in the corner, you know, it's like Cloud releases a new model and then OpenAI releases Sora and like, you know, there are like all these different things, but yeah, the new models are good. It's interesting. My my perception is definitely that just, just observationally, Cloud 3 is certainly the first thing that I've seen where lots of people. They're, no one's debating evals or anything like that. They're talking about the specific use cases that they have, that they used to use chat GPT for every day, you know, day in, day out, that they've now just switched over. And that has, I think, shifted a lot of the sort of like vibe and sentiment in the space too. And I don't necessarily think that it's sort of a A like full you know, sort of full knock. Let's put it this way. I think it's less bad for open AI than it is good for anthropic. I think that because GPT 5 isn't there, people are not quite willing to sort of like, you know get overly critical of, of open AI, except in so far as they're wondering where GPT 5 is. But I do think that it makes, Anthropic look way more credible as a, as a, as a player, as a, you know, as a credible sort of player, you know, as opposed to to, to where they were.…

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