Goodmorning Manu, thanks for taking the time to gather your thoughts & contribute to this conversation! I think you raised some great questions & edge cases that I think should rightfully be explored further.
I believe if multiple Named Neurons vote on behalf of a Neuron, the “fee” deducted from the Named Neurons rewards should be split amongst them - to my knowledge, the followee system adheres to the consensus of the Named Neurons followed.
In the instance a Named Neuron doesn’t utilize the voting power at all: I don’t think they should be rewarded.
While thinking about this however, it’s making me start to believe the Named Neuron system as a whole needs an overhaul. What happens in the instance you follow 3 Named Neurons, 2 vote, and 1 does not? Does the Neuron vote or abstain?
This is also a good point - in the instance a Named Neuron follows DFINITY, for example, on technical topics, while an individual chooses to follow the Named Neuron on those topics, I believe the rewards should go directly to the executing voting party. This avoids scenarios in which true voting parties would be “snubbed” of their rewards due to receiving a fraction of a fraction of voting rewards.
And to add to these details, how this incentive would be distributed to the followed neuron :
would it be added to its own stake, making it not liquid ?
Or would it be transferred to any chosen address, making it potentially liquid, allowing more flexibility ?
It sounds like a meaningless detail, but by this I am thinking of promoting optionality, decentralization and democracy. Here is why : in the second case, people could spontaneously create co-managed neurons and share the fees during the whole time of co-governing, and even create democratic neurons with a potential turnover within their members due to elections etc, without having to sacrifice any fees.
But in the first case, (if it has to go to the neuron’s staking), people could not co-manage named neurons with an incentive. We would only incentivize individuals neurons. But from my experience, a co-organized neuron is a very powerful tool for decentralization. Codegov and Synapse are very good example of organizations successfully joining forces to allow more decentralization.
I envisioned it accruing as maturity for the Named Neuron, although adding it directly to the stake is also an interesting concept that could make more sense than accruing as a potentially liquid reward.
This would contribute to the Named Neurons voting power; while passively rewarding them through personal rewards. In the instance the Named Neuron wants to “claim” this maturity, they have to disburse the neuron - yet they simultaneously lose the Named Neuron.
Ok, so it would be designed for named neurons managed by individuals rather than by groups, except if transitivity allow to reward C when A follows B and B follows C as you wisely justified it.
You could be right yes. Anyway, partial or complete, transitivity is clearly necessary. I think that B should be rewarded if it votes before C, but not rewarded if it just followed C. It would create the good environnement for decentralized co-managed neurons.
Even if you are right, I still think that is better than following someone I don’t really trust.
In sum, your current proposal just makes the following story: You get my voting power, and I should pay you for that.
I have a better suggestion: The NNS should set up an Automatic Voting button so that I can vote by myself with sufficiently low cost and without any need for following someone I don’t really trust. For instance: An Automatic Voting “Rejection” Button.
If the names neuron got people to follow then they deserve their money. For example there are senators that write bills and senators that just vote. They all make same amount of money.
With multiple neurons it should be split among neurons that voted in the direction your vote went.
Dfinity, ICA, Synapse, CodeGov, and perhaps ICDevs all cast their votes by following multiple voting members. Technically, only one of up to 15 Followee neurons that are configured for these neurons actually triggers the entire liquid democracy aligned to their follower maps.
I’ve never observed it happen (which would be quite difficult) with regard to Synapse and GodeGov, but it is possible for one of the individual Followee neurons to actually follow another neuron that is not known to be aligned with the known neuron. Timed just right, the vote that triggers liquid democracy might not be one of the Followees for the known neuron.
There is also a significant fraction of follower neurons that follow more than one of the known neurons. The amount of liquid democracy that is triggered by a known neuron is different depending on which known neuron votes first and whether the votes of the known neurons are Yes or No (yes requires >50% while No requires >=50% in order for a neuron with Followers to vote).
Interesting piece of trivia…CodeGov and Synapse.vote (NEW) are both controlled using the Manage Neuron proposal topic and they have no known private key. It’s not possible for these neurons to ever disburse ICP because that capability is intentionally excluded from control parameters of the Manage Neuron proposal topic. However, they can spawn neurons from maturity and assign a new principle ID as part of that process.
I like the idea of this proposal, which would go a long way to incentivizing people to get involved in NNS governance.
However, NNS has 13 governance topics, and the vast majority of neurons only vote on a limited number of topics (Governance and SNS), with most of the others following DFINITY. this is actually quite risky, and the IC chain is still heavily reliant on DFINITY for proper functioning.
I think how to incentivize named neurons to vote qualitatively for all topics is something that must be considered before implementing a reward system
You are wrong. If you only follow DFINITY, you’ll lose some voting rewards.
The named neurons get those followers’s voting power, which is already a big compensation.
If a named neuron thinks that this compensation is not enough, it can just delist itself.
I don’t like the idea of taking part of the rewards from the followees. Maybe everyone should be able to configure the topics they actively vote on and what commissions they charge for delegating their vote to them, just like in many other protocols? We live in different countries and it is difficult to give an ideal % of commission that would be satisfactory for all of us.
Looking for a one-size-fits-all solution seems too centralized.
I would like to hear from someone with financial expertise as to whether or not this could require licensing in some countries in the future. I can see a situation arise here where the named neuron could be seen as a service provider, automating your voting and in return receiving compensation for the service. Some jurisdictions may see this as business or possibly even banking practice which may require licenses to then be maintained. Additional cost to neurons for any licenses that had to be maintained would then increase ICP sell pressure would it not?