,
27 tweets,
5 min read
Read on Twitter
Thanks for organizing this event. I'm transcribing here the first part of this morning's presentation, dealing with governance since it's in a form suitable for a long tweet storm, so here goes ...
In 2014, before the Tezos paper came out, almost no one was talking about governance of cryptocurrencies. Today, there isn’t a single project that doesn’t discuss its own plans for governance. I am proud that Tezos led the way here. But what is governance, and why do we need it?
Governance is a procedure used to manage shared resources. When resources are private, and not shared, there is no need for governance. For example, if you own an egg, you can eat it now, eat it later, let it grow into a chicken, give it away. It’s your egg.
If 2 people own an egg, their desire can conflict. In order to allow them to make decisions about one egg, their needs to be rules. One rule could be that the older person decides, another rule could be that they flip a coin, another that they have so split the egg.
That is governance in a nutshell : it takes many. potentially conflicting, inputs and gives a single output. Governance exists in many institutions, inside companies, inside families, inside clubs, inside sports team, etc.
When we think about cryptocurrencies, we think mostly about private property. If I own a coin, I can keep it or spend it. f I keep it in a multi-sig address, there are governance rules, but very clear and simple : if m out of n agree to a transaction, then it can happen.
We do not immediately think about governance because it doesn’t look like cryptocurrencies need it. If everyone decides what they want to do with their own coins, why is there a need? It turns out that, in these networks, some resources are shared by the entire community.
One example of a shared resource is the network itself! Who runs it? Who operates it? In Bitcoin, miners maintain the network. They are rewarded for doing so by creating new coins in every block. This is a form of governance, where inflation is used to incentivize honest behavior
Another shared resource is the protocol itself, the rules by which the cryptocurrency operates. If those rules never change, then there is no need for governance. However, if the rules evolve, then there needs to be a procedure for deciding how they should evolve.
In the world of open source software development, a traditional solution for governance is to fork. If I want to use a piece of software a certain way, and you want to use it another way, there is no need for conflict.
Fork-based governance works well for software if there is no need for us to run the same version. However, cryptocurrencies need consensus, and they require everyone to run the same version. When the software for a cryptocurrency is forked, two incompatible assets are created.
Since the software is open source, each of us can modify and tweak our own version. We do not need to agree to run the same version. This is called a fork, and I refer to the procedure as fork-based governance.
This is a problem, because money derives its value from a network effect. Forking reduces the network effect. Typically, when cryptocurrency forks, one fork will capture a much larger portion of the mindshare. So maybe there is no problem after all? Maybe the best fork will win?
Unfortunately, there is no guarantee that the “best” fork wins. Because one fork typically "wins", we know that a form of governance is happening, but what does it reflect? Does it reflect market judgment? Does it reflect a choice by the miners or the block validators?
In reality, this is a coordination game. Market participants have to identify which branch will attract the most interest from the other participants. This is very circular, everyone is watching what everyone else does and tries to do the same thing.
In the end, the winning branch is the one people expect will win. It is not primarily based on merit, but rather on the ease of coordinating around that solution.
Attempting to address this underlying problem of coordination in the absence of formal rules is one of the main motivations for Tezos and the on-chain governance mechanism.
By having a clear mechanism from the start, the aspiration is to ensure that the network can adopt better technology as it becomes available *and* that network effects are preserved. That coordination can be maintained around one chain.
However, it shouldn’t be *too* easy to implement major changes to the protocol. The system for upgrading Tezos - or any blockchain protocol- needs to be designed so that it is possible to implement improvements but difficult to implement changes which harm the network.
In building this sort of mechanism, on should encourage coordination around the solution that keeps the network united together. The way Tezos does this is through a multi-stage voting process taking roughly 3 months from proposal to activation.
Compare this with current fork-based governance in other projects that have extremely long roadmaps with sometimes a year or 1.5 years between upgrades. This allows the network to upgrade more incrementally, more often.
How do we know that the formal rules will stand the test of time? The good thing is that not only does Tezos allow you to amend the protocol, but it also allows you to amend the rules governing changes to the protocol.
The current implementation of Tezos governance is very simple and uses conservative parameters. The purpose was to bootstrap the network with a system that allows decision-making to represent a large enough share of the network.
I would like to see refinements in the governance process that harden the network against social attacks and allow for greater accountability. One such mechanism is incorporating prediction markets to rank proposals, followed by a vote.
In my view, prediction markets and voting are very complementary mechanisms.Voting is useful to make sure that there is widespread approval of the proposal, that most of the participants are willing to see it’s pass. This is important for stability.
Prediction markets are good for getting *informed* inputs into what some of the best proposals may be. The two mechanisms complement each other beautifully.
Beyond prediction markets, constitutional rules, imposed programmatically on proposal open up interesting possibilities. Meta governance is something I'm definitely interested in watching closely! ∎
