00:00 Speaker A
They say its new Claude model beats OpenAI and Google on key benchmarks. What are those benchmarks?
00:06 Dan
Yeah, so there there are things that have to do with uh agentic uh operations. So agentic coding, uh uh things along those lines. This is their Opus 4.8 model. This is their their flagship. It’s a step up from 4.7, as you can imagine. It goes, you know, 789, uh I think. And it really they they put it up against the top of the top. So you’re talking GPT 5.5, uh you’re talking Gemini 3.1 Pro. and they say that it knocks those over in all but one category. Uh you know, this is obviously Anthropic’s testing. So, you know, I mean, not that they would lie or fib, but you know, you just got to put that caveat on there that it is their testing. But I think it just shows that as as you just said, they continue to to iterate, innovate, and continue to move forward. Uh and you know, it it does lend to this, you know, continue kind of uh buildup of of products and announcements that they’ve repeatedly had lately. And, you know, you you look at OpenAI and they’re not standing still by any stretch of the imagination. But I think the cadence with which uh uh Anthropic does this uh and the way in which they make these announcements, it shows or at least it makes it feel as though they’re just, you know, beating up on OpenAI. Uh and so obviously investors are very excited about that. Uh and you know, whenever these these companies go public if it’s not this year, then next, I mean, you you got to imagine the demand is going to be huge as long as everything stays at the level it’s at right now and we don’t see some, you know, catastrophic blow up of the AI trade.
01:54 Speaker A
Are we at the point in your opinion where the average consumer can really tell the difference, Dan, between these AI models?
02:04 Dan
You know, I I was talking with someone about this yesterday uh and the way they framed it was at a certain point, it’s Coke and Pepsi, right? It’s and I was, you know, I thought, A, I hate Pepsi. It’s it’s disgusting. And also give me Diet Coke all day. Give me the chemicals or whatever is in it. I don’t care. But it it is going to be like that where it’s like, you know, one is going to be virtually the same as the other, just different flavors, really, different uh palette, if you will. Uh you know, if you look at something like the free version of of Claude right now, it’s it’s limited in, you know, what you can do with regards to the latest news, but you you could still perform similar operations and then you have, you know, Claude Code, then you have uh Code X over at OpenAI. So, you know,
03:00 Dan
I I do think that at at a certain point, you know, it’ll be, well, I guess it’ll be Coke, Pepsi and uh Jolt Cola, we’ll call Google, you know, I guess I’m just trying to think of another cola. RC Cola. delicious. It’ll be along those lines, right? And so, yeah, all three are just going to be continuously ping-ponging back and forth. meta jumping in every once in a while. call them Sprite, I guess. Uh and Dr. Thunder maybe. And that’s really where where I think it’s going to end up. It’s I don’t think it’s ever going to be at the point anymore where it’s like, you know,
03:41 Dan
one of these is running away with it, as it was when OpenAI came on the scene with ChaT and everyone was just like, what do we do now? I think now the the the the talent is in place, the uh the understanding is in place. uh and it’ll it’ll take something big I think to really shake everybody uh and say, okay, well, this is clearly a uh a a failure on one company’s part or the other. I will also point out that you will have investors be overly exuberant when, you know, I don’t know, uh like with Gemini, when Gemini 3 came out, it put Google into the place where it’s in the same conversation. Uh its model was better at that point than than the others. But that it changes every, you know, week or so. So it got them up to speed and now I think you continue to to just see this kind of back and forth between the three.
