Was wrapping up a datanet submission on @OpenLedger when I noticed the $OPEN 7-day print — up 14.3% on the week ending May 23rd, with $13.43M in 24h volume. Not massive, but notable given how flat activity had been. openLedger had been one of those projects I kept meaning to actually use rather than just track.
The thing that stayed with me isn't the price move. It's how differently the platform feels depending on which entry point you use. ModelFactory in default mode is genuinely frictionless — upload data, point at a base model, fine-tune, done. It moves fast. The vibecoding pitch makes sense here; you're not writing infra, you're configuring intent.
But then you go one layer deeper — actual attribution verification, checking that your datanet contribution is correctly linked on-chain — and the UX drops off a cliff. It's still doable, it's just clearly built for a different person than the one ModelFactory is designed for. The attribution engine is the whole thesis, but the smooth path hides it.
I came away unsure who this is actually optimized for right now. Data contributors who care about Proof of Attribution need to dig. Developers who want fast AI deployment find it immediately. Those two groups aren't always the same person.