The term “decentralization” has seen a massive increase in usage of late, becoming one of the more common tech buzzwords of 2018. Its meteoric rise parallels a growing frustration with the injustices that derive from its seemingly evil twin, “centralization”. For others, decentralization merely appears to be a new concept popularized by the advent of cryptocurrencies. In reality, proponents of centralization and decentralization have been waging war against each other for centuries.
Even within the cryptocurrency community – founded upon the principles of decentralization – platforms labeled as centralized exist. What constitutes being centralized or decentralized is open to debate, and everyone seems to have specific ideas as to how to make the distinction.
One argument that is difficult to refute, however, is that centralized systems share several inherent weaknesses. Vitalik Buterin summed them up nicely in his essay on decentralization, namely low fault tolerance, low attack resistance, and low collusion resistance.
Low Fault Tolerance
Most people generally understand what is meant by a single point of failure. If a system retains one, it is said to have a very low fault tolerance: one mistake or breakage will bring down the entire system. Consider a village with 20 homes all connected to a single power generator. If the power generator fails, the whole village will be without power. Now consider the same village, but with a small power generator in each house. If one house’s generator fails, the village system as a whole still functions.
Taking this a step further, imagine a low fault-tolerant system that partially resides within a larger interconnected network of systems. Systems with this characteristic allow interdependent cascading failures to occur. Sounds bad, right? It is. On September 28, 2003, a blackout in Italy resulted in a widespread panic when a multiple system collapse ensued (think rail networks, health care systems, financial services). Worse yet, the partial degradation of the nationwide communication system further impaired the power grid management system.
In recent years, network traffic and system interconnectivity have increased dramatically. Moreover, this phenomenon will soon grow exponentially, particularly as the Internet of Things (IoT) gets underway. Consequently, preventing cascading interdependent failures is becoming paramount. A responsible plan for doing so requires a concerted effort to reduce the number of centralized systems that exist and to improve to improve their fault tolerance.
Low Attack Resistance
Similar to low fault tolerance, the low attack resistance of centralized systems results from a single point of failure; however, in this case, the cause of any crash that occurs is not accidental.
Why have empires throughout history sought to expand (or decentralize) across continents and around the globe? Most would say for resources, or perhaps prestige. However, an equally important reason is to increase their ability to defend the empire. That is, they raise their attack resistance by moving away from a centralized system. For an aggressor to defeat a decentralized empire, most or all of its bases must be attacked and overcome. More often than not, this is prohibitively expensive.
A cool example of a centralized system (provided by none other than VB) with a low attack resistance was Star Wars’ first Death Star, an immensely powerful planet-destroying superweapon. The only problem (for the Galactic Empire, at least) was that its thermal exhaust port presented a relatively easy means to destroy the entire thing.
Low Collusion Resistance
Some (politicians, in particular) argue that through a stronger leader or set of rules, centralized systems offer a form of security against unrest and adversaries. That’s all well and good if you have a benevolent, democratically inclined leadership. However, it also represents a significant risk in itself, especially if the ruling order turns tyrannical and lower-ranked members within the system suffer as a result. The negative consequences that flow from such systems takes many forms, such as financial inequality (which plagues most, if not all, countries), lack of freedoms, etc.
It doesn’t even take a malevolent actor for collusion to occur. After all, collusion can be somewhat subjective. A government working secretly with private industry can be seen as either collusion or cooperation, depending on whether one supports or opposes the government. Alternatively, how such behavior is perceived may also depend on how much one stands to benefit from the action. A more accurate way of defining collusion could be “any coordinated activity by an entity seen by the majority as unjust or immoral, that creates or strengthens control over somebody or something.”
Centralized systems present favorable conditions for collusion since a few powerful individuals with aligned interests can often take control. At the outset, most may have good intentions, but once they get a taste of power, it can be difficult to limit it. Also, it’s not uncommon for leaders to engage in morally wrong activities when they are under the delusion that they are doing if for the greater good.
So, What’s Next?
You could be forgiven for admitting you’re sick of hearing the words decentralization and blockchain. In the eyes of the general public, they appear to have little meaning beyond what some remember as a passing fad. Unfortunately, the average user doesn’t think about the merits of decentralized systems. Indeed, there is no centralized vs. decentralized “winner-takes-all” battle. Paras Chopra wrote a nice article explaining why this is the case. Here’s a key quote from it:
“Decentralization-centralization is a continuum and not a binary choice. This means that in a free-market, systems migrate towards a point that creates maximum value for customers of that system. A system that starts with a fully decentralized vision (bitcoin, Youtube) migrates towards centralization while a system that starts with a fully centralized vision (USSR) migrates towards decentralization.”
The winners of the next race will be the teams that can adequately identify the strengths of decentralization, and turn them into superior easy-to-use products that create maximum value for customers. This focus is in-line with the points made by Todd Hoff in this article. He lists many points that he feels are likely to prompt a clear divergence to decentralization.
- Complete deterioration of trust such that avoiding the centralization of power becomes a necessity.
- Radically cheaper cost basis.
- It becomes fashionable.
- The decentralization community manages to create superior applications as convenient and reliable as centralized providers.
I like to think of them as the starting points for product development. Simply being decentralized isn’t enough; however, if your product or platform doesn’t originate from or produce one of the effects above, it’s probably not going to make it in the real world.