Something caught me off guard mid-task today. Was reading through Newton Protocol's ($NEWT ) token design — the NEWT Foundation blog post laying out how the economy is supposed to work — and I stopped on the developer royalty mechanism.
@NewtonProtocol built a revenue share into the protocol for AI agent developers. Publish a model to the Newton Model Registry, get picked up by an operator, and you earn a royalty cut of the NEWT fees that flow through it. That's a real economic incentive loop. #Newt
Hold up — then I went to check what's actually on Newton Explorer right now. Operator attestations. Policy evaluations. That's it. The Model Registry that would let developers list agents, collect royalties, build a market... still roadmap. Which means the "open economy for AI agents" has its incentive structure fully designed, but only one of its four intended participants is actually active.
I spent a few minutes trying to figure out if I was missing something. I wasn't. The developer layer, the user layer paying NEWT gas to instruct agents, the royalty flows — all of that is sitting behind a component that hasn't launched yet. Meanwhile 78.5% of total supply is still locked per recent tokenomics data, so the token inflation pressure is going to keep building well before the economy those tokens are supposed to power comes online.
The architecture of the open economy is genuinely thoughtful. The sequencing is the question I can't quite resolve.
When does an economy stop being a design and start being a market?
@NewtonProtocol built a revenue share into the protocol for AI agent developers. Publish a model to the Newton Model Registry, get picked up by an operator, and you earn a royalty cut of the NEWT fees that flow through it. That's a real economic incentive loop. #Newt
Hold up — then I went to check what's actually on Newton Explorer right now. Operator attestations. Policy evaluations. That's it. The Model Registry that would let developers list agents, collect royalties, build a market... still roadmap. Which means the "open economy for AI agents" has its incentive structure fully designed, but only one of its four intended participants is actually active.
I spent a few minutes trying to figure out if I was missing something. I wasn't. The developer layer, the user layer paying NEWT gas to instruct agents, the royalty flows — all of that is sitting behind a component that hasn't launched yet. Meanwhile 78.5% of total supply is still locked per recent tokenomics data, so the token inflation pressure is going to keep building well before the economy those tokens are supposed to power comes online.
The architecture of the open economy is genuinely thoughtful. The sequencing is the question I can't quite resolve.
When does an economy stop being a design and start being a market?