Collaboration, contracts, and the tyranny of platforms
The following article describes what fundamental problem Blockchains solve by taking a big picture perspective on humanity.
If you appreciate the article follow me on Twitter.
Some of the key insights of the following article;
- The only true problem that humanity has is our inability to collaborate.
- Credible agreements are always over-collateralized.
- The only advantage traditional contracts have over smart contracts lies in the number of assets one can use as collateral.
- Centrally controlled network effects allow platforms to build unexpectedly dangerous incentive structures.
- The incompleteness of contracts explains why we need to decentralize governance through token ownership.
- Global relational contracts could be a way to overcome the limitations of incomplete contracts.
I came up with the thoughts shared here while trying to explore what highly scalable Blockchains allow us to build, as described in my other article “50 million partners“. The broader argumentation laid out here is what the other article is building upon.
Table of Content
- Human Collaboration
- Game Theory
- Contracts vs. Smart Contracts
- Platform Tyranny
- Unbounded Platforms
- Conclusion
Human Collaboration
In order to understand where crypto is headed, I think it is crucial to take a step back to think about what fundamental problems Blockchains solve.
Personally, I think Blockchains solve an economic, or game-theoretic problem.
Let’s begin by taking a very big-picture perspective on our species. Besides distinct characteristics such as our intelligence or consciousness, we often overlook another important property which I would call; the degree of individuality.
Imagine there were hypothetical species that had different levels of individuality. To the left, we have a species with the lowest degree of individuality while on the right we have a species with the highest degree of individuality. Individuals of a species to the left share a common consciousness, with each member sensing and feeling what all others do. In everything that this species does, it takes into account the well-being of each individual. In fact, we can therefore think of it as one creature.
To the right, a world full of psychopaths. Each individual is completely detached from everyone else with absolutely no empathy or shared well-being. Individual entities of that species would literally kill each other for a bar of snickers if there were no consequences to it.
Humanity is certainly rather on the right side of that spectrum. While we do care a lot about friends and family, we’re not really aware of the total suffering around the world, choose to ignore it, or simply only care about it up to a certain point. Being highly detached from each other allows each of us to have unique thoughts, memories, and experiences. I’d argue that this makes us naturally highly innovative as each brain produces new recombination of thoughts. Useful thoughts and ideas even accumulate over time in the sense that we learn to teach them to others efficiently.
Species that are on the left side are probably less innovative as they produce less distinctive thoughts. Their actions are much more synchronized, which leads their experiences and thoughts to be more homogenous. The benefit of such a species is that it is brilliant in maximizing social welfare for any given level of technological advancement. Such a species would not require very strong institutions either, as naturally, it seeks to improve everyone's well-being.
This is different for a species to the right.
While our highly innovative nature is very useful it comes with an inherent inability to naturally collaborate.
Game theory has brilliantly shown that rational individual actions can cause large-scale suffering. Climate change is a good example of that. Each country has little incentive to emit less CO2, which ends up causing huge problems for humanity at large.
As we are naturally equipped to discover and innovate, I’d argue that collaboration, and the consequences of our failure to do so is the only real problem we face. Simple game theory explains why we face climate change, have financial crises (my personal view), and why we can get stuck in dictatorships for centuries. Basic game theory that proves how rational individual actions can make us all worse off is in my view one of the most important discoveries.
Interestingly, the degree of individuality is a helpful mental framework to think about how society could be structured. It plays into the debate around communism vs. capitalism as well as the role of governments. Species to the right create a lot of value by leveraging markets, as they allow each individual to come up with new products that potentially also benefit the species at large. Markets leverage what is naturally given to species to the right; independent thinking.
If we were a species to the left, communism could start to make a lot of sense, as leveraging individual creativity is generally less fruitful while the risk of a dictatorship is reduced as well.
In order to both leverage our natural abilities as well as to collaborate effectively, we have invented tools to overcome collaboration failures. These tools are insanely useful. Democratic governments and Blockchains are technologies that withstand the individual temptations not to collaborate. In fact, they make use of our fundamental nature to create incentives for a sub-set of us (politicians & miners) to act in the interest of others. On top of these fundamental technologies, we have built layers of abstractions such as the legal system or neutral virtual machines that we can use for a vast variety of applications.
TLDR:
- Humans have a natural ability to be innovative.
- Humans have a natural inability to collaborate.
- Both democratic governments and Blockchains are technologies that we can use to solve these issues.
Game Theory
The two most famous categories of games in game theory that show how individual actions can limit our collective well-being are “prisoner’s dilemmas” and “coordination games”. I won’t explain in detail how these work, I recommend that you watch a youtube video instead. In a prisoner’s dilemma, each individual always has incentives not to collaborate with others no matter what they choose to do. Coordination games are different in that regard, as each individual's willingness to collaborate depends on whether others choose to do so. In the most typical coordination game, there are two equilibria, one in which nobody is collaborating and one in which everyone is.
A good example of a prisoner’s dilemma is climate change. Each country never has incentives to reduce CO2 emissions, which ends up causing large problems for all of them. These situations need a persistent modification of the incentives in order to enable collaboration.
Technology platforms are examples of coordination games. The default equilibrium is that nobody is using them. For example, when nobody was using Uber, we were stuck in a game where it was irrational for anyone to start using it, even if the app already existed. The intuition is that if nobody uses the app, then the drivers do not generate any revenue from offering rides, which in return means the app is not useful to consumers. The entrepreneurs behind the app have solved that issue for us. They have taken the risk to move everyone to an equilibrium where we have tremendous gains in trade between riders and drivers.
To summarise; these games contain two or more parties that could exchange value with each other, but fail to do so due to rational individual incentives.
TLDR:
- Individually rational actions can make us worse off.
- Prisoner’s dilemmas require persistent modification to the incentives.
- In coordination games, once we reach a new equilibrium we have incentives to stick to it.
Contracts vs. Smart Contracts
Let’s explore how both the legal system as well as Blockchains solve these issues.
Legal System
Imagine a situation where Marc Zuckerberg is looking for an investment. To get it he’s willing to share the future profits of Facebook with an investor. However, without a legal contract, he eventually has little incentive to share any of the profits once Facebook is a large success. The investor anticipates that of course and won’t invest in the first place. That’s a practical example of a prisoner’s dilemma.
To make a deal work between the two Marc needs to credibly promise that he’ll share the profits of Facebook. The promise is only credible if not holding it makes him lose more than what he can gain. This basic perspective implies that even in the legal system all agreements are overcollateralized. We simply need to extend our perspective on the subject to include assets that are neither liquid nor tradeable. So what is it that Marc is staking?
Well, ultimately he stakes his own freedom.
He voluntarily agrees to lose his freedom by being put in prison in case he defrauds the investor. I think this is a very systematic way of looking at it. The legal system simply allows us to stake a large variety of assets including our reputation, our freedom, future revenue, or illiquid assets.
By staking our own freedom society can make us nearly infinitely miserable very inexpensively (prisons are cheap to run relative to how much it probably sucks to be in them). That means if Marc wants to promise billions of profits to the investor, he can easily overcollateralize that promise to make it credible.
That brings up the question of what laws exactly are? I would say laws are contracts that we were forced into. Within them, we typically stake any sort of tradeable asset or our own freedom. This allows us to change incentives persistently to solve certain prisoner’s dilemmas. Society at large is basically a bunch of overcollateralized agreements.
Staking non-tradeable assets that entrepreneurs naturally possess enables them to receive liquid assets from others without owning any of them themselves. This allows an entrepreneur to fund their business with the liquid assets of others. From that point of view, it is crucial that we can utilize assets that we already possess and that are of significant value to us, in order to attract liquid capital. If an entrepreneur cannot stake his own freedom to credibly promise not to spend seed money on Lamborghinis, he will probably not get funded.
An essential part of what makes it possible to stake all of these assets is that we have created identity systems and also that it is not possible for us to escape our own physical bodies. These properties do not fully exist online yet.
The legal system also has tremendous disadvantages. Contracts are typically expensive, slow to execute, and inherently not global.
Blockchains
Blockchains on the other side enable us to engage in global, inexpensive, agreements that settle instantly.
The disadvantages of Blockchains are that stakeable assets are usually limited to being liquid, tradeable, and on-chain. As such, the core difference to the legal system lies simply in the variety of assets that we can use to form credible promises. Of course, this view might or might not be complete but I certainly find it a very useful perspective.
If we need to possess liquid assets in order to create credible agreements to receive other people’s liquid assets, the value we can create is ultimately limited.
If we look more closely, however, we can observe certain situations in which entrepreneurs can already stake illiquid, non-tradeable assets to create credible, global agreements by using smart contracts. The possibility to do so allows a net amount of liquid assets to flow into financing decentralized applications. This is one of the main reasons why a parallel financial system could be bootstrapped. Let me explain what I mean by walking you through an example.
Imagine Marc Zuckerberg built Facebook on a blockchain. He wants to raise seed money from a global set of investors. He does so by selling ERC20 tokens of a governance contract. Would that work? Actually, it wouldn’t.
The thing is in the seed stage, his promise is not overcollateralized and thus not credible. In most cases, he could simply sell the tokens to investors, get liquid assets from them in return, and then copy-paste Facebook’s codebase to start from scratch. He would essentially defraud the investors by simply ignoring the agreement. In the legal system, he’d implicitly stake his own freedom as judges would likely either order the confiscation of his assets or directly send him to prison. I think that explains why funding whitepapers in 2017 was indeed a bad idea and why seed-stage deals still happen within the legal system.
When does that change? In my opinion, it changes as soon as network effects start to kick in due to a community of early adopters. There are two aspects here. In case Marc starts from scratch, early users will most likely not join the copy-pasted Facebook version anymore since it is common for early users in crypto to invest in the projects they like, which means they got defrauded by him. Losing these early users is one part of the collateral. The other is the convexity of the opportunity. Facebook has network effects, which means sticking to the old contract could lead to exponential gains for Marc. Starting from scratch could mean losing the opportunity to a competitor. The steeper the upside is, the more likely he is to stick to the agreement.
I’d therefore argue that both network effects and the presence of a significant community allow a net amount of liquid assets to flow into financing a project. While it is not possible to either stake our own freedom or a large variety of other assets in smart contracts, network effects and communities prevent entrepreneurs from defrauding token holders. We often intuitively think of it that way but breaking it down can sharpen our understanding.
In the long run, I expect that we’ll be able to use a larger variety of assets as stakes in smart contracts. Especially interesting in that regard are Blockchain-based identity and reputation systems. They will possibly allow us to stake our own reputation. The stake itself is likely to become more valuable the more services require a high reputation. This could lead to a self-reinforcing dynamic in which we can use something of greater and greater value as collateral for global, credible agreements.
TLDR:
- Credible promises are always overcollateralized.
- The legal system allows us to stake more assets than Blockchains do.
- Blockchains enable global, inexpensive, and instantly settled agreements.
- Network effects and a community of early users are useful collateral for Blockchain-based agreements.
Incomplete Contracts
If we want to deepen our understanding of how we can collaborate together, we need to consider further limitations of contracts. Nobel Price Winner Oliver Hardt has made groundbreaking work on that subject. I can highly recommend watching some of his lectures. One of his key insights is that contracts are inherently incomplete. That means reality is simply infinitely complex and as a result cannot accurately, or completely be described by contracts. This is a much bigger problem than people realize and plays into many aspects of how we design organizations generally and also DAOs.
Hardt points out the importance of ownership. Owning something essentially grants someone so-called residual rights over it. Residual rights give you the ability to do whatever you want with that thing unless a specific contract prevents you from doing so. Ownership is nothing else than the assignment of residual rights. Think of a house that you own, as long as you fulfill obligations that you might have with either the government or a mortgage provider you can literally do with it what you want. The infinite amount of possible things you can do limits how binding contracts are. Contracts simply cannot cover all possibilities. Owning the residual rights of something gives you all these choices and thus, also protects you from others having them.
This observation has important consequences when we try to form credible agreements. Incomplete contracts are the primary reason why VCs want voting rights when they invest in a company. In fact, that problem explains huge parts of how governance is structured generally. If contracts were complete, investors could simply write a comprehensive contract that prevents an entrepreneur from doing anything other than properly working on the business. In that case, the entrepreneur could keep all voting rights as the extensive contract with the investor would simply prevent him from defrauding the investor in any imaginable way. The reality is, however, that neither the entrepreneur nor the investor know in advance in which way the entrepreneur could eventually cheat. That is the reason why the investor rightfully expects to own a part of the residual rights of the company, namely, shares.
It is also the reason why it is inherently hard to separate voting rights from financial participation in a company. Voting rights will in the end always play into the division of value. An entrepreneur could for example vote to pay out the entire company’s profits as his salary.
“Owning” a board seat goes in the exact same direction as it gives members the rights, within the boundaries of what boards are allowed to do, to decide on everything that is not specified in any contract.
Relational contracts are an alternative solution to the problem of incompleteness but are of a less precise nature. The idea is that a third party subjectively considers the intentions and the context behind a contract and then decides upon anything that is not covered directly by it. That means relational contracts aim to establish trust between participants in an ongoing relationship not based on pre-defined exact rules but rather based on a set of norms and guiding principles that take into account the economic and sociological context of the relationship.
In the next part, I combine what I have described so far and show how insanely important it is that large platforms are not allowed to be centrally controlled. Or at least that society at large should aim to control large platforms through either credible relational agreements or through the decentralization of residual rights.
TLDR:
- Contracts cannot reflect reality. They are in many cases incomplete.
- Owning residual rights is often required to make agreements work.
- Relational contracts could be a valuable alternative to the traditional way of forming agreements.
Platform Tyranny
We all have the intuition that large web2 platforms screw us over. The idea of “cutting out the middlemen” is what has been driving crypto for a long time, especially in the early days. I have recently begun to understand that the intuition behind this can be substantiated and that the problem in some cases can be much larger than is commonly understood.
Changing the Game
We said before that platforms can be thought of as coordination games. The adoption of a large global platform would be the equivalent of an equilibrium in which we have managed to coordinate towards using it. Once we reach that equilibrium we all have incentives to stick to it. It is obvious that large platforms, therefore, become monopolies and can extract profits from the value our mutual interactions create. This is not per se bad, as profits are needed to stimulate entrepreneurs to build these platforms in the first place.
However, I think it goes way further than that. My assumption/idea is that major global platforms with literally billions of users are in very unique positions to further strengthen their network effects. They can leverage both profits and their existing network effects to further strengthen them. For example, Facebook seems to be in a unique position to launch a global payment-, reputation-, or identity system. Imagine they’d launch their own payment system to allow for borderless and instant transactions globally. It is trivial to see that we would be using that system for a vast amount of different transactions either work-related or for our private lives. As trust is often needed in a large variety of transactions it could make a lot of sense for Facebook to implement a reputation system to further stimulate the usage of the platform. At first, that system would simply allow some users to credibly signal to have a track record of not defrauding others in specific transactions. That reputation could become more and more valuable the more additional services are built on top of the platform, for example offering under collateralized loans to people with a good reputation. Over time more and more incomes and wealth would depend on that reputation. I think such a scenario is not very far-fetched.
When Facebook rolls out a new feature we generally get two choices, adopt the feature or leave the platform. As everyone else is on the platform, we are likely to adopt the feature. The more we subsequently begin to use the feature the more essential it can become to our daily lives. Facebook could step-by-step roll out new features, that are perfectly rational to adopt for us while continuously strengthening their network effects.
Punishment
Things start to get really bad when Facebook manages to find a way to “punish” us for leaving the platform that goes beyond us not receiving the benefits of using the platform. That might sound highly unlikely. We can always just opt out, right?
No, not necessarily.
There is a famous example of how the Grameen Bank in Bangladesh successfully developed incentives structures around micro-lending that lead to a very high repayment rate on these loans even though no legal contracts were involved.
The bank allowed people to self-select in groups of 5 (maybe it was 10) to apply for a loan. If one individual in the group would not pay their loan back, the entire group subsequently would lose the ability to get further loans. To minimize that risk, people would naturally self-organize to form groups that highly trust each other. The bank rightfully assumed that if a given member would not pay back the loan they’d lose something of great value to them, namely, the beneficial relationship with their peers. Credible agreements without any legal contracts, brilliant.
An implication of such a structure is that the more value each group member draws from the ability to get loans, the stronger they will sanction other members privately for not paying them back. To put it differently, the more valuable the service is to the group members the more powerful the service provider becomes.
Dangerous Implications
If Facebook strengthens their network effects more and more, they could use that exact same method to enforce policies on us that eventually would make us worse off compared to before the company existed. Imagine your entire family had a reputation score on Facebook that is of very high value to them. Facebook could initiate a policy that lowers all of your family’s reputation scores in case you stop using the platform. The main idea here is that changing the rules gradually to strengthen their network effects can be perfectly rational for us to comply with but can lead to bad outcomes. We are essentially trapped in a coordination game where the equilibrium is modified continuously but none of us has ever incentives to jump to another one.
The more we bring this to its logical conclusion the clearer it gets that this structure begins to resemble something incredibly bad; a dictatorship. In such a structure each individual has only incentives to repel the tyranny when everyone else does so as well, a coordination game. The problem is when you repel alone, you face serious punishments. Facebook initially does not have the ability to punish us. Over time, however, leaving the platform can be punished by lowering the reputation scores of our peers and thus negatively affecting our relationships with them.
Technically, if we’d anticipate all of that, it would be rational to leave the platform before all of that happens. The only question is if we really anticipate it? Don’t get me wrong I do not at all expect this to happen nor do I think the people at Facebook are planning this somehow. I’m just saying, incentive wise this could be the eventual conclusion of a few simple assumptions.
Of course, we’d also try to regulate the sh** out of them before that occurs. However, the inherently global nature of these platforms and the local nature of our governments is something to be concerned about in that regard.
The conclusion of this part is that a centralized control structure over large platforms is much more dangerous than we think. Large coordination games controlled by a few people are not the way we should organize ourselves in the 21st century.
Unbounded Platforms
If our democratic system is rightfully restricting the growth potential of such platforms, how would the world look like if our governments would instead encourage it as the ownership rights of the platform would either be decentralized or a global relational contract would prevent the platform from implementing harmful policies?
That is of course already happening in DeFi. When using decentralized platforms we can be sure to not face these risks, we can all be vastly better off coordinating towards them. As a result, these platforms will tend to get massively larger.
I think factors such as competition can limit the ability of platforms to screw us over. We do not have to decentralize every platform. But I do think the large global web2 platforms pose a substantial risk to all of us in case they remain centrally governed. Besides that, so far I only focused on the most extreme outcomes. Even before someone can create a dictatorship they typically can extract much more rent than we initially anticipate when we start to use the platform. If a decentralized version can promise a fairer distribution of value then us coordinating towards that platform is likely to be stimulated and the created value as a result likely to be vastly higher.
I liked this recent tweet of Ryan Adams related to that:
Imagine if our governments would want Facebook to have its own currency, reputation system, etc. If they would encourage the platform or would even start to use it.
Global Governance
While the distribution of residual governance rights is one way of aligning incentives between the owners and stakeholders of a platform, another would be to allow the platforms only to exist in the context of a global governance system. A protocol-level on-chain governance structure could enable that. Such a system would obviously need to be enormously robust and reflect the preferences of humanity at large, essentially a global democracy. While that is incredibly hard to build, of course, organizational-wise, it could have benefits that entrepreneurs could keep large portions of ownership rights of their platforms and would thus, have more incentives to build them. We’d reduce neutrality to the lowest level of the structure if you will (similar to how governments do it). If we were to build such a system I think we’d be in a unique position to rebuild how agreements are structured generally by making more and more use of relational contracts. A protocol-level governance system would essentially act as a judge for large platforms to engage in such relational contracts with various stakeholders to establish trust between them.
TLDR:
- Global, centralized platforms are possibly more dangerous than we think.
- Decentralization of global platforms has to occur through either the wide distribution of ownership rights or through the use of relational contracts.
Conclusion
In this article, I argued that the most fundamental problem of our species is our inability to collaborate. Putting assets at stake can fundamentally solve that problem by establishing trust in agreements. While the legal system allows us to stake more assets, Blockchains, for the first time, allow us to engage in global agreements.
Large global platforms can leverage network effects to abuse our inability to collaborate internationally. If neither competition nor governmental interventions restrict the power of such platforms, then I believe some of them might be well-positioned to further add to their abilities to negatively influence society. Platforms that credibly promise to act in society’s interest can and will become vastly larger than the ones that fail to do so. Blockchains are the core technology to allow for that and thus for unprecedented value creation.
Credible agreements are inherently restricted by the incompleteness of contracts. With traditional agreements, the distribution of residual rights is often the only way to establish trust in agreements when contracts fail to adequately describe the complexity of reality.
Relational contracts can be used to overcome these limitations. On-chain governance systems on the protocol level could allow large Blockchain-based platforms to engage in such agreements. This would have the advantage that entrepreneurs could keep large parts of the platform’s voting rights while credibly limiting their ability to act against society’s interests. This is, however, very hard to implement but could be a long-term goal worth pursuing.
I hope I could provide valuable insights and thoughts on the nature of contracts, human collaboration, and the power dynamics of platforms.
As I said in the beginning I had these thoughts while focusing on another subject, namely what highly scalable Blockchains will allow us to build. In this article, I propose a possible answer to that question.
If you appreciate the article follow me on Twitter.
Selected References
- Wood, P.J., 2011. Climate change and game theory. Annals of the New York Academy of Sciences, 1219(1), pp.153–170.
- Wikipedia: Prisoner’s Dilemma
- Wikipedia: Coordination Game
- Argenziano, R., 2005. Network markets and coordination games. Yale University.
- Cathy Barrera, ICOs Leverage Network Effects
- Wang, F. and De Filippi, P., 2020. Self-sovereign identity in a globalized world: Credentials-based identity systems as a driver for economic inclusion. Frontiers in Blockchain, 2, p.28.
- Hart, O., 2017. Incomplete contracts and control. American Economic Review, 107(7), pp.1731–52. Vancouver
- Wikipedia: Relational Contracts
- Chris Dixon, Why Decentralization Matters
- Khavul, S., 2010. Microfinance: creating opportunities for the poor?. Academy of management perspectives, 24(3), pp.58–72.
- Fred Ehrsam, Blockchain Governance: Programming Our Future
