This post is exploring some ideas related to the Network Nervous System in order to improve the governance of the Internet Computer.
Let’s start with a few problems the Internet Computer community is experiencing right now and then we’ll dive into a few possible solutions (they are by no means the only solutions) that could potentially solve these problems.
The Problem
The NNS follower structure is too centralized
Right now almost all neurons are following the Internet Computer Association or DFINITY Foundation neurons. There are a few completely understandable reasons for this:
1. You have to vote to get rewards, so you want to make sure you vote on every proposal. However, you aren’t notified when a proposal is posted, and there may even be some proposals that don’t show up on the NNS frontend (can someone verify if this is true?). So to maximize your rewards, you need to follow someone that votes on everything, meaning the ICA or Dfinity foundation.
2. The default options (and when I say default, I mean the only named options to choose from) when choosing followers are ICA or the Dfinity foundation, so the power of defaults incentivizes new ICP stakers to follow these neurons. In fact, you have to enter a neuron ID to follow someone else, meaning you would have to know who you wanted to follow and their neuron ID if you wanted to follow them.
People lack the necessary expertise to vote intelligently on proposals
Non-technical folks will almost never have the expertise needed to review a pull request to see if the proposed change makes sense for the Internet Computer Protocol. Many others fit in this same group as well. If people don’t have the technical expertise, they need to rely on trusted experts to vote appropriately.
It is too hard to find expert neurons
Like I mentioned above, to follow an expert neuron you would need to (1) identify the expert through social media or referral, (2) ask them for their neuron ID that they use to vote, (3) trust that they will vote on ALL proposals relevant to their expertise even though they aren’t notified of proposals going through the NNS, and (4) trust that they will vote in a way that aligns with your Internet Computer beliefs and viewpoints. This is a hard thing.
Internet Computer experts are not incentivized to get followers or vote on all proposals
It would be easier to find expert neurons if experts went around advocating their ideas and vision for the Internet Computer, sending their neuron ID to those interested in following them, and generally promoting themselves as an expert for a particular category of NNS proposals. BUT, there is no incentive to do so, other than altruism. You earn the same amount of rewards if you have 0 followers or 1k followers.
Summary of Problems
Because of all of these problems, I don’t see the centralized voting and follower structure of the NNS changing anytime soon. People are incentivized to maximize their rewards (ensuring they vote on all proposals), experts are not incentivized to get followers, and it is too easy to simply follow the ICA or Dfinity foundation.
But wait! Isn’t the goal of the NNS decentralized governance of the Internet Computer? Now that we have a functional NNS governing the Internet Computer Protocol and a community interested in its success, can’t we do something about it? I think so, so let’s discuss a few ideas that might incentivize decentralization of followers in the NNS.
One Possible Solution (two parts)
There are two possible solutions (out of many) that might work together to solve these problems.
1. Provide a small reward to neurons based on the voting power that is following them
2. Create a new app on the IC that facilitates the exploration, discovery, and following of expert neurons
Rewarding Followees
If a neuron received a small reward for the amount of voting power in neurons that were following it, I think we would immediately see decentralization of followers. Nobody would want all of those rewards to go straight to the Dfinity foundation or the ICA, and experts in particular domains would be incentivized to gain a following.
This would be very similar to a representative democracy. Anyone that stakes ICP could follow (or “vote” for) an expert in each proposal category that would then represent them and their interests. They would be trusted to vote on every proposal in that category as well as to do some due diligence on every proposal to ensure that the proposal submitted to the NNS was representative of the will of their followers. If they did a good job, they could gain even more followers. If they did a bad job, they could lose them.
Adding a small reward for neurons that have a lot of voting power in neurons following them could potentially have negative consequences, such as people trying to buy your follow or incentivize who you follow in some way based on ulterior motives, or popular neurons receiving followers due to popularity and not expertise. However, I generally think this incentivizes decentralization of followers while still protecting the community governance of the NNS.
App to Facilitate Neuron Exploration and Following
The second piece builds on the first one. Assuming we create a new reward based on the number of followers your neuron has, I think we’ll need a way to allow for the discovery and following of expert neurons. So I propose a new application to be built on the Internet Computer that facilitates this.
This app will be a campaigning ground for expert neurons wanting to grow their following. Experts can pick their category, create a profile showing their credentials, and create a statement illustrating their vision for the IC and why ICP stakers should follow them. Then the app can list the expert’s voting history on relevant proposals and include a button that allows someone to follow that expert neuron.
The app enables followers to make an educated decision about which expert neurons to follow and provides accountability for the expert neurons because everyone will be able to see their voting history, percentage of proposals voted on in their category, etc.
Conclusion
I believe that adding a small reward for neuron followers and building an app (as described above) will (1) dramatically increase the decentralization of the voting/following in the NNS, (2) enable the community to make educated decisions to pick experts to represent them, and (3) facilitate greater scrutiny and feedback on NNS proposals as experts have incentive to review NNS proposals more carefully for their followers.
