judicious-decentralization.md 5.8 KB
title: "Judicious Decentralization: Why Blockchain isn't the Answer to Everything" date: 2018-03-27
category: Essay
The hype surrounding blockchain and cryptocurrencies at the end of 2017 and the beginning of 2018 was astonishing. For a cryptography-based database technology, the amount of interest in the tech communities and the wider public has been difficult to understand (at least to me). The tone in many quarters has been that blockchain is going to revolutionize everything from finance to education, science to gaming. What is it that makes this technology seem so powerful to people?
One of the primary appeals of blockchain, in many different use cases, is that it is decentralized, and thus decision making is (potentially) distributed and there is not a single point of failure. These properties should make a whole range of systems more robust. As an example, blockchains for Bitcoin and Ethereum are stored on many computers, all across the planet and because of this, are more resistant to failure or attack than are databases or other data stores which are housed in a single machine (or even on multiple machines, using typical redundancy/backups, which is standard today in large-scale / important apps and services). Furthermore, the consensus that the network demands of its nodes incentivizes cooperation, which is a topic that could have deep social significance as this technology matures.
Decentralization also offers the potential for more social, networked forms of service provision than those which are possible in the existing technical and financial landscape. Use cases of this sort include a blockchain-supported energy grid, comprising micro-providers and micro-consumers, with tiny increments of solar energy being doled out on demand as users wander around a city. Another possibility is a mesh-network of micro-ISPs (Internet Service Providers), where individuals share bandwidth in exchange for micropayments. These sorts of distributed platforms could function as alternatives or competitors to the monolithic infrastructure we see today in sectors like energy, information technology, and finance.
Perhaps the most interesting category of applications that blockchain is predicted to impact concerns high-level, abstract organizational functions, such as decentralized governance and organization (e.g., DAOs, or Decentralized Autonomous Organizations), algorithmic democracy, decentralized lending logic, crowdsourced research labor, and a host of new sharing-economy platforms (cars, bikes, computers, etc). The effects of blockchain on these areas of communal life could be revolutionary.
Despite the theoretical potential for blockchain to disrupt so many areas of life and society, there is today a lot of overwrought rhetoric surrounding the technology. That is to say: though blockchain will probably find a number of important uses in the coming years, it is not going to solve the better part of humanity's problems. We should take a more measured stance regarding its value.
A refrain bouncing gaily around the internet is that we should "decentralize everything". This, incidentally, is an excellent search term if you want to find content produced by some of the true believers in the crypto space (I recommend duckduckgo.com for such searches). Given how important decentralization is in narratives about blockchain, I think it is worth pointing out that decentralization is not inherently superior to centralization as an organizing principle. Some evidence for this comes from biology: higher organisms have a "central" nervous system. The tendency in evolution towards increasing localization of sensory appendages in a head region, from jellyfish through ants, wolves, and humans, is called cephalization. It is not entirely clear that this localization is necessary for higher intelligence, but it is not easy to think of highly intelligent creatures that do not display it.
Another example of centralization being effective comes from human social structures: most businesses have CEOs or presidents, universities have presidents, and most sports teams have captains and coaches. Governments have prime ministers and presidents (and often dictators). Again, these are forms of centralization of control. The people at the head of the organizations function much like the head of an animal: they serve as a hub for incoming information, which they synthesize to generate output that radiates "downward" through the organization. In this way, complex structures and institutions can function as a cohesive unit.
I'll note that decentralization of control clearly does have advantages in some cases. The cooperative behavior of bees, ants, humans, and other animals suggest that a hybrid intelligence, partially centralized and partially decentralized, is often nature's solution to complex problems. More fundamentally, complex systems, such as eukaryotic cells, multi-cellular organisms, ecosystems, and social institutions, tend to exhibit a tension / complementariness between top-down (or centralized) and bottom-up (or decentralized) control. Both appear to be needed, much of the time.
Perhaps it is the case that we have not previously had the capacity or technology to effect proper decentralized organization, and blockchain will answer that problem. Maybe decentralized governance will show centralized leadership to be crude and ineffective, a vestige of our roots. But a more likely outcome, I think, is that we will find that some domains are excellent candidates for decentralized control structures, and others are better managed through centralized operations. For those starting businesses in the crypto / blockchain space, or those investing in said companies, it could be immensely valuable to know which systems fit which profile.