What Every DAO Is Getting Wrong Today: Part 1/4

Arbitrum DAO is suffering from a governance process which is costing us efficiency and accountability. Most DAOs are having this same problem. They suffer from the well known pattern of being decentralized and funding too many disjointed efforts causing them to over-correct into centralization. Then, the calls for accountability - which started during the decentralized funding without strategy era - end up driving the further centralization of the DAO by the entities built to steward "progressive decentralization".

Basically, we are struggling to put the O in the DAO.

In Arbitrum's case, the proposed solution of an OpCo has incredibly high operational expense of $30 million in less than 3 years and does not at all attempt to move the DAO towards an onchain decentralized model which can assure builders they won't get rugged by the platform the way Zynga got rugged by facebook.

So the question is "Why would the Arbitrum delegates make the exact same mistake every other DAO has made for the past 6 years and expect different results?

The reason is because they, like most other DAOs, don't take the time to really understand the problem. You can see here from a recent survey, that accountability is what is seen as the biggest issue.

This would explain why people think that an OpCo could solve the problem. The problem is accountability, right? If the OpCo was accountable for holding everyone else accountable, then we would be good, right?

I hate to be a bummer, but the incentives are for the OpCo to hire more and do everything themselves where they can control the outcomes best. When people say, "how are you earning that $30 million?", they are going to think they need to be Michael Jordan.

In reality, we need the OpCo to be Phil Jackson. A purebred enabler of talent willing to step back from the limelight. As the entity setting up OpCo, I'd like to see Entropy Labs working as Phil Jackson, but their current trajectory is to be like Mike. They shouldn't be graded on how many things they get done alone, but how many great teammates (other service providers) they develop. By leading by example, they could build the OpCo that really would drive onchain decentralization that sustainably supports a web 3 economy long into the future.

This won't happen by optimizing for DAO revenue.

I'm here to show you how we are making three critical mistakes (Parts 1-3) and will finish with some suggestions on how we can fix these issues and put the O back into DAOs - without centralizing (Part 4)!

These are the three critical mistakes every DAO is making today:

Optimizing for revenue because we don't understand how autonomy vs mandates affects the system as a whole - we forgot why decentralization and autonomy are critical.
Not defining our product as the collaborative goal of the network with a positive feedback loop between our core product success and the price of the governance token
Settling for a governance process which only emphasizes decision making while completely forgetting about sense making before and action taking after the decision.

Why Decentralization and Autonomy Are Useful

We Already Have Revenue Optimizing Organizational Structures

DAOs with a tradable governance token should never optimize for revenue because it creates a fundamental tradeoff state between the contractors and subcontractors. I'll illustrate it below.

We already have centralized corporations that are great at driving revenue anyway! We don't need DAOs to be good at revenue. In fact, the corporations we have today are so good at revenue they often become parasitic or cancerous to society. Decentralized and autonomous organizing offers an answer to this problem.

How We Became So Centralized

Before the industrial revolution, most production was decentralized. Each town had its own grocer, blacksmith, and banker. Sure, the Medici family was already expanding banking centralization during the enlightenment era, but in general, decentralization was the standard before mass production.

Centralization came when we realized the economies of scale benefits provided by mass production. Powered by telegrams and telephone communication, steam engines, cars and railways, we suddenly found ourselves able to create assembly lines of mass production. Mass production utilizes the specialization of individual workers in a collaborative system to create increasing returns. Another way of saying increasing returns is positive-sum value.

However, capitalism rewarded the capital put into these mass production facilities. The mass production companies kept all of the increasing returns - deemed profits to the owner(s) of the company. Capitalism has driven immense innovation. Similar to centralization and decentralization, I believe capitalism to be a wildly useful social technology when used in the right ways.

A side effect of this centralization was that the person who owned the production facility became the go to lender for the "last mile" suppliers. These are the farmers who grew the corn or the loggers who cut down the wood. Over 150 years a pattern of the centralized mass producer getting yield from banks which then charge the last mile providers interest for the loan. The worst part of this is that the average payment terms from the distributors are now > than net 120 days. This means the big company gets paid by the bank to store their money while the little guy supplier is basically offering a free loan to the big company while waiting to be paid! I've done a deep dive on the $55 trillion supply chain industry.

The Physics Problem Bitcoin Solved

As the internet developed, we naturally wanted to communicate over longer and longer distances. An unintended consequence is there is an easy way and a hard way to solve the fundamental problem underlying internet communication - the speed of light.

Scarcity creates value. Information asymmetry is a false scarcity in that it is created rather than the natural state. If I have insider information, I can trade that before anyone else - when it is scarce. Therefore, as the value of information transmitted through the internet became more and more valuable, centralized entities not only bought the best access, but they closed off the system to others.

Trading on information depends on the speed one receives and sends information combined with the ordering of the transactions. Let's take out the possibility of manipulation by an exchange like the NYSE and discuss how the simple speed of light problem can create value that is captured by centralized entities.

Example 1 - High Frequency Trading

Before bitcoin, the way transactions were ordered for value exchange was generally done by an exchange. This caused the real estate closest to the NYSE to become incredibly expensive. Not because of the scarcity of the land for any other reason than its signals would get to the stock exchange faster allowing the ability to front run trades from further away. Read all about it in Michael Lewis' Flash Boys.

This model of ordering transactions first come, first serve created a false scarcity of land value.

Example 2 - Google Spanner System

When writing a google doc with someone who is on another continent, the system still knows who typed what letter first. Amazing! This is done using atomic clocks at data centers around the world, constantly timing how long it takes messages to travel from one to another, then a proprietary system is used to adjust each person's input into an ordering flow.

This proprietary system creates a false scarcity of data and code because the incentive is for the centralized entity to "close the door behind them". Google spanner is not open source.

Both of these systems reinforce the artificial scarcity created by the incentive for centralized entities to protect the scarcity that creates value.

The Benefits of Decentralization

Decentralization provides the economies of scale benefits to many centralized organizations who use onchain transparent and auditable processing rules to organize in positive sum ways. The alignment of an open and permissionless set of participants is done by changing the game theory of participants at two levels ensuring all participants that the platform will reliably output results as it is intended. It ensures that no single actor can change the rules to create a false scarcity which only benefits them.

Optimizing Individual Strategy

In every situation where an individual has autonomy, they are given two fundamental options.

Play the game as intended creating an economies of scale benefit for all participants including themselves. In game theory this choice is called "collaborate".
Play the game in an unexpected way in an attempt to capture more value than is their rightful portion of the increasing returns generated by the other collaborators.

In bitcoin, the miners are the participants who could run the software out of the goodness of their heart, but not enough run it without being rewarded. The individual miner strategy is to collaborate because it is in their self-interest to be eligible for the block reward.

Optimizing Against Well Financed and Collective Action

This is what makes bitcoin resilient to nation state level actors. The system uses a small amount of inflation of the asset to reward miners for collaborating. This encourages more miners to participate. As more miners participate, it means a wider set of people would need to be convinced (or hacked) to change the code on their miners to change the monetary supply. People that want a consistent monetary supply asset buy Bitcoin pushing the price up. The positive feedback loop is between the asset price and the goal of the Network.

This means to attack at large, one would have to purchase enough hardware to run half the network before being able to attack effectively without the individual strategy optimizations making it not worth it. However, now we have a new idea called crypto-economic security. It is derived from statistics and the law of large numbers. By the time a well finance actor could attack the system, there would be no way to capture the value before the system as a whole lost value.

The optimal strategy for individuals and well finance actors and collectives is the same - always collaborate!

So why can't we do this for DAOs?

Mandates vs Autonomy

The crypto-economic security derives from a law of large numbers. Bitcoin and Ethereum are able to achieve these large numbers by incentivizing participants. They have a strict set of rules to follow that turn inputs into reliable outputs.

DAOs can't incentivize the needed large numbers because when we pay for people to respond, unlike the very specific rules of hardware miners, we get more noise as we pay more. Its exactly the opposite of the higher signal bitcoin receives.

Now before I divulge the solution for this problem, you must understand why autonomy is needed. Take a look at my sophisticated diagram with a worker, a factory, and a town.

Without autonomy, the worker is going to be paid by the factory which serves the town. Maybe the town people pay the factory through taxes. Maybe direct. We all agree that the worker is being paid by the factory and his mandate is to do his factory job. He does not have autonomy. There is one entity telling him what he's worth.

Centralization means that the decision making autonomy is centralized in the factory, not the worker.

Now we know that this little dude would be SOOO much more productive if he was inspired, working for himself, in his lane, moisturized, and thriving. Why do we know this?

Centralization is good for aligning individual actors through mandates to collaborate in a way that the factory gets to keep the increasing returns from the collaboration.

Decentralization is good for letting the little dude go and price his skills on the market and optimize his skills to be the best at what he does. Capitalism is a great tool for driving this.

Autonomy is a good tool for the little guy to want to go optimize his skills. If he stays employed with the factory instead of running his own business he will be motivated to be as good as the average DMV worker. Unless the centralized entity is transparent about his pay an others. If he feels that his pay is legitimate, he will work to the extent of his capability.

DAOs are a new mechanism trying to bring the economies of scale benefits to many centralized entities. The only way to do this is to maximize for the centralized entities - the workers - to optimize as if they were selling to a market. Now how do they allow the autonomy to the worker to sell to the market and optimize 100% while still ensuring they are in collaborate mode and not defect mode?

This is the problem DAOs are trying to solve.

DAOs Are Not Entities

Not even that cute Cayman foundation or the well structured OpCo which must listen to token holders. Those are both centralized entities that are part of the network. DAOs are ecosystem. They include a bunch of centralized entities which are good at optimizing for revenue.

DAOs are ecosystems like a city.

All the people living there (users) and the businesses they support (protocol and dapp builders) need the infrastructure like the roads and electricity (open source dev tools and blockchain code). We need service providers to pick up sanitation and provide public services like parks (security subsidy fund) or even specialized SPs to drive subsidies for businesses (development incentives) in a food desert or low cost capital for real estate developers (liquidity incentives).

Yes, there is a place for a city hall (foundation or OpCo), however, if city hall does all the work itself it will just bloat. Because it hires by mandate, it can coordinate collaboration in a centralized and gatekept way. Even when trying to decentralize and hire more service providers - only the friends of City Hall will feel like they have a chance to be hired.

You aren't in a DAO, you are participating in decentralized and autonomous organizing. The difference of good centralized governance and good decentralized governance is that decentralized must be so good people choose to opt into the rules.

Should the city optimize for revenue? No. It should optimize for more users moving to the city. Not by paying for their first month rent and watching them move out of the city after the program ends, but by supporting the things that make a uniquely enjoyable experience living in the city.