Vanar

Dario Amodei was asked if open source could eventually replace Anthropic. His answer breaks the entire industry narrative. He said " I don't think open source works the same way in AI that it has worked in other areas. Primarily because with open source, you can see the source code of the model. Here, we can't see inside the model. It's often called open weights instead of open source to kind of distinguish that. But a lot of the benefits, which is that many people can work on it, that it's kind of additive, it doesn't quite work in the same way. I've actually always seen it as a red herring. When I see a new model come out, I don't care whether it's open source or not I ask, is it a good model ? Is it better than us at the things that we're doing ? That's the only thing that I care about. Because ultimately, you have to host it on the cloud. The people who host it on a cloud do inference. These are big models. They're hard to do inference on. One of his two reasons was that these models are just hard and expensive to run, so open weights don't really set you free. That was true when he said it. It isn't now.Inference on those same models got cheap fast, with some providers cutting costs more than 20x, and the coding gap that used to separate them from Claude has basically closed on the benchmarks. Nobody outside the lab really knows why one model is good and another isn't. That's the actual moat, and it's a strange one, because it protects Anthropic by protecting no one has the knowledge just doesn't exist in a form you can copy.