I am withdrawing this proposal.
Thank you to everyone who engaged here, and to the many people I spoke to outside the forum over the last two weeks. The quality of the responses was higher than I’d hoped for, and several of them changed my thinking.
The strongest thread running through the feedback is about discretion. A fund that generates yield needs someone deciding where the capital goes, and someone deciding who receives the proceeds. Both are judgment calls, and therefore both introduce risk: manager discretion, counterparty and contract risk, and the political risk that comes with any body allocating a growing pool of capital. Several of you made the point that validator rewards today are algorithmic, and that replacing them with an allocation, however well governed, turns security from something the protocol guarantees into something a human process decides. These are risks the protocol should not internalize.
We also want to make sure that buy back revenue from intents and other sources doesn’t ever come back to circulation - that was my underlying assumption with a proposal but the proposed mechanism doesn’t guarantee that. Making sure it’s that so is important here.
The second, and related, concern is around “public goods,” which now has a lot of baggage and is perceived as easily corrupted. Any definition wide enough to be interesting is wide enough to be captured, and the discussion here surfaced good arguments that the category tends to expand to fit the funding available. If we revisit a similar idea in the future, the scope of public goods should be tightly specified and boring: validators, node operators, MPC, i.e. the infrastructure the network can’t run without. In order to minimize the risks that come with human decision-making, there shouldn’t be room for judgment about what does and doesn’t qualify.
As for where attention should go, I think the feedback is right that the binding constraint today is demand, not what we do with the treasury. The highest leverage work in front of us is to grow assets and volume on Intents, adoption and NEAR staking for AI. Those generate real external revenue and utilization of NEAR tokens. Until that revenue is large and durable, questions about how to invest it are premature, and the structure could likely cost more than it earns.
To be clear, I think this discussion went the way it should. I put forward an idea, people with real skin in the game pushed back with numbers and specifics, and the outcome is different from what I proposed. That is what this discussion process is for, and it sends a better signal about NEAR’s governance than the proposal passing would have been. Thank you especially to those who deeply engaged with this conversation.
Now the floor is open. If you have a view on how protocol revenue should be handled, whether that is burn, native staking with permanently locked principal, direct algorithmic distribution to validators, or something not raised here, please post it. Several of the alternatives in this thread deserve their own threads, and I’d love to see them debated on their own terms and carry the conversation beyond the replies to mine.