Crypto-economic Networks as Technological Enablers of Scalable Tribe-like Collaboration
“Homo sapiens rules the world because it is the only animal that can believe in things that exist purely in its own imagination, such as gods, states, money and human rights.”
Human beings are social animals which have for most of our history (up until about 12000 years) evolved in tribes in no more than 150 people. The rapid proliferation of our species was among other things enabled by language through which we were able to create stories. Religion, nation-state, and money are all examples of stories that enabled us to create larger societies which eventually conquered smaller, tribe-like communities of people that didn’t evolve sufficiently over time. Besides that, female humans are choosy mates (as opposed to most other primates) by which given an unrestricted choice, a small percentage of males receive attraction of the majority of women (Pareto distribution). The inverse is not the case, the distribution of female sexual market value is normally distributed. This phenomenon is studied in research on ancestor’s genes and is also manifested in the online dating market. So essentially, this is the nature’s way to filter out quality which enabled us to accelerate the mastery of our environment.
While living in tribes, self-governance was possible as a way to establish strong reputational networks among people that knew each other. This way people policed themselves, without any need for a central entity. As humans gathered in larger and larger groups through story-telling, we were able to collaborate with people that we didn’t necessarily know. Christians fighting side by side on the Crusades didn’t need to know each other, they just needed to believe in the same story, the story of the Bible.
The problem with self-governing communities is that they don’t scale. The challenge of larger groups of people is designing the right kind of social structure to establish order and maximize societal well-being. Whereas the market-based economy enables efficient allocation of resources, it is often the case that the market incentives are not aligned with societal values. An obvious example of this is the mechanism of privatized profits and socialized losses during the financial crisis. The financial decision-makers didn’t suffer the consequences of their actions nearly as much as the society at large did. This was a clear example of a lack of skin in the game.
“Show me the incentive and I will show you the outcome.”
Up until recently (2009), there was no way to coordinate people across the world that didn’t know each other without any central entity coordinating the interactions and making sure that people behave as expected by the designers of the system.
Bitcoin was the first example of that. It consists of a series of incentives (incentive structure) written in a 9-page white paper which is embedded in code that everyone has access to. It’s a peer-to-peer way to move money from one person to another without any central entity that anyone needs to trust for the value to be exchanged in a reliable way. This way, a crypto-economic network was bootstrapped into existence.
The internet enabled a free flow of information and Bitcoin enabled a free flow of money. Crypto-economic networks more broadly enable free flow of trust.
There are many ways to look at a crypto-economic network. The one I find that compresses the concept the most is defining it as a trustable incentivizing machine (more on this).
“Software is simply the encoding of human thought, and as such has an almost unbounded design space.”
Computers in the network provide tribal-like trust between people such that whenever they do something, they know that they will be compensated for it as expected. This is similar to tribal reputational networks, except there is no need for people to know and trust each other, they transfer the trust to the network mechanics. Cryptography as a collection of techniques for secure communication enables this. The human activity tracked on the network can be either transferring money, providing a service, borrowing or investing into a physical asset or any other activity that includes some kind of transaction that indicates an exchange of value.
The computers enable individuals to trust the network because there is an economic reward for providing computational resources to secure the transactions that are executed on the network.
It’s embedded into software by the network creators and executed automatically across different computers across the world.
In essence, they can make people do stuff
Given all of this, what are some actual examples that exist today which enable scalable collaboration between people that were not possible before? In this writing, I will focus on collaboration sharing either information or data. The reasoning behind the choice is that it’s where there is often a lack of incentive to collaborate. By providing these two examples we can grasp the idea of where this could lead us in the future.
It’s old news that data is the new oil. Google, Facebook, Uber, and other tech giants often open source parts of their technology stack (e.g. Tensorflow, Prophet, Pyro) which is definitely not the case with their data. The reason behind this is that there is a lot of potential value in the data that these companies accumulate. The more data they have, the better the user experience their products can (potentially) provide. By having a large user base which often grows even larger because of (data) network effects, they build data monopolies. This depends on the specific use-case, but the essence is the same: companies are not incentivized to share their data. Is there a way to change this?
Ocean Protocol built a crypto-economic network in the form of an open data economy. Participants are incentivized to share the data that they accumulate with others. They are financially compensated based on the estimated value generated from the shared data. This has the potential to break the information silos that exist today. This also lowers the ability of smaller companies to compete with larger ones for data science talent because of the lack of data. In data science, data often matters more than algorithms. An example of a shared data economy is a consortium of car manufacturers which is pooling their data in order to train a self-driving car model. Another example is Ocean Protocol partnering with Roche Diagnostics to “improve care for heart disease patients through safe and secure data sharing.” This way data owners are able to set pricing for their data and see how it can be used and by whom. Data scientists are able to run algorithms on this data whilst its kept stored encrypted, on-premise. Ocean creates the connecting substrate between problem solvers (data scientists, AI researchers) and problem owners (NGOs, enterprises, governments) by which an AI commons can be created. With that, the scalable tribal-like collaboration is enabled by a trustable network that’s driven by economic incentives for collaboration.
The investment industry is a typical example where there is little incentive for collaboration between different actors on the market. Success (ROI) depends on some kind of information asymmetry between the investor and the market (being non-consensus right). This can be done either through fundamental or technical analysis of the financial markets (among other things). Sharing one’s money-making investment strategy means losing the informational edge on the market, a zero-sum game (in order for one to gain, the other needs to lose). Once again, is there a way to change this?
Numerai built a crypto-economic network in the form of an AI hedge fund which changed the incentive structure of sharing financial information. They leverage the mechanism of auctions and staking to incentivize data scientists to contribute their predictions using the provided data. They are compensated for their contribution to the overall performance of the hedge fund’s meta-model. The data scientists are provided with abstracted data which means that they don’t know what that data represents. This removes human biases and overfitting (predicting well on the data available but not generalizing further on unseen and/or new data). It also disables people to take the data without giving back their contributions. A cryptographic token makes it economically irrational to submit models that overfit. It’s not about telling people what they should do. It’s about incentivizing them for the desired network behavior to emerge.
They transformed the financial market’s underlying incentive structure from competition to “an invisible collaboration to build the meta-model”. They leverage the principle of ensembling (combing many diverse models), which enables “lower error rates in machine learning, higher returns on individual trades, lower portfolio volatility, and higher portfolio exposure”.
Crypto-economic networks enable a fundamental shift in the way we organize our society, without any central entity establishing trust and incentives for the interactions between individuals. Meanwhile, as with any technology, the more powerful it is, the more we need to be careful in its implementation. Yet the only way to learn about how to leverage it is through tinkering.
It’s easy to take the current organization of our society and compare it to some idealized alternative. It’s tempting to sprinkle the new technology on every business use-case that we have today. Is there a need for a decentralized Uber? Or maybe decentralized Facebook? There are a lot of conditions that need to be satisfied for this to happen as building decentralized applications is substantially more complex than building centralized ones. There are clear trade-offs (e.g. the value of decentralization versus scalability) that need to be taken into account when building these applications, at each of the layers. Also, there are applications that will be built that we can’t even conceive of today, similar to what happened with the Web 2.0 technology. Although it’s possible to decentralize every centralized software product, it doesn’t mean that we should.
And finally, a thought that we can all regularly revisit to remind ourselves of the long-term potential of crypto-economic networks, especially during all the crypto bubbles:
Please note that many of the ideas presented in this essay are borrowed from others while exploring the crypto community rabbit hole.