to be tricky. I don't believe we're going-to-toe with big labs on ML. In fact, we're not, because many reasons. What's it going to be used for? Ocean has solved hosting data, presumably for KGs as well as unstructured. The agent ecosystem should be self-funding (might need cash injection, but should get return) for hosting. It feels like a piece of the Hyperon architecture is going to sit there. But what?
Zarqa
That's a carve out though, no?
Ecosystem project, the true diamond in the SNET crown. Just my speculation
Yeah, I'm interested to see how that turns out. Not an expert, but for me the power has always been in the frontal cortex: the orchestration (be that the oracle of Snet, the MoE orchestration, self-improving scaffolds, etc)
Обсуждают сегодня