Won’t speak to the to-token-or-not-to-token question.
- Fun experiment, run it!
- Will this face all the same engagement issues as other DAO plays? If validators are human, they’re going to be a bottleneck. If not, they’re effectively evals, and then who builds those evals (it’s a pretty specific and highly valuable skill that typically does not come for free).
- Either way seems like thing you’d game if you knew the reward structure.
- This is 100% happening outside of Web3 (honestly prob before even the current AI boom, feels like Delphia maybe tried some version of this a while ago, probably more in-house versions), so a good question to ask is how Web3 primitives could do this better. More broadly, better understand / do some background on how these evo approaches compare in this domain with end-to-end learning methods or analytic Black Scholes style equation modeling.