Q Let's move, uh, uh, you know, out and a layer down. Like what does having access to an American open source model mean, or just more and more powerful, like, uh, open source AI models mean for the application ecosystem?
A I mean, I know like a lot of people and some enterprises are really iffy about like using like, The best open source model. They're like worried. It's like, there's nothing wrong with them today. There's nothing in them today. Right. You know, there's the worry that one day they check. I mean, you don't, but you can just vibes it out. Like they're like competing with each other to just released as fast as possible. Right. Like, like deep seek and moonshot and all these other, you know, Alibaba, et cetera. Like they're competing to release as fast as they can with each other. The Alibaba teams in Singapore are like, I don't think that they're like putting Trojan horses in these models. Right. And like, There's some interesting papers that Anthropic did on like, you know, trying to embed some stuff in models and ended up like being detectable pretty easily. Again, like, I don't know how to, you know, I'm not, I'm not too much into that space of interpretability and like evals, but I just don't think that they are, right? It's just a vibes thing. But some people are worried that they could be, or they're just like iffy, like, oh, I don't want to use a Chinese model. It's like, well, fine, but now you're going to go use a service. That is backed by a Chinese model, which is fine. Like, you know, like, uh, but they, you know, they're fine with that. They just don't want to directly …
AI assessment note: “it mostly just really interesting because it continues to move the commodity bar up”