Blockchain and the Promise of the Perpetual Consensus Machine
by Avtar Sehra
Note: This post was initially part of a talk given in late 2015, and was later posted on Linkedin as a short overview. I have reposted here as I am moving most of my activity onto Medium. Over the coming months I will publish an updated version of this with a more tangible/detailed discussion on entropy and PoW/PoS consensus mechanisms.
There is a lot of talk about Proof of Stake (PoS) consensus mechanisms being the Holy Grail to drive “efficiency and scale” on open public blockchain networks. In some ways there are similarities with the debates of an open or closed web of the 90s.
In the early/mid 90s many companies, convinced of the evils and dangers of an open Web, tried to provide closed versions. This probably sounds absurd now, but at the time many did not believe an open public infrastructure could work at scale for real commercial use cases. So ensued use cases that pushed the creation of “walled gardens” or other forms of closed communication networks. Such networks did not succeed and we now have an open public web — where you can manage security and privacy using end user applications/protocols. This is now the way companies that range from start-ups to global banks deliver trusted online services.
In the present day there is a similar debate being held for blockchain networks, but in a different way. Many people/firms still think that open public blockchain(s) can’t work so some companies are creating their own blockchain technology protocols for a closed group. While the open public networks are scaling globally and have a market capitalisation in excess of ~USD1o0billion, no private network has managed to effectively scale and provide real world production benefit. While this can change, the principles of network effects are increasingly against companies being able to scale a private blockchain network to compete with the open public networks. So it is likely that history will repeat itself and public networks will be the ones that reach market dominance — potentially slower than the web did due to the highly regulated nature of transacting money/assets.
However, there is an even more fierce debate taking place that is much more subtle than that of open or closed networks. Most people who understand the nature of open and closed systems are perpetuating that the “real issue” is not about a chain being public or private but about the type of consensus mechanism that underpins the synchronisation of the network, the debate is now around Proof of Work (PoW) vs Proof of Stake (Pos). This will likely be the point of contention over the coming years. Possibly in the future people will discuss the absurdity of this PoW/PoS debate.
There is no doubt about it PoW wastes a lot of energy, and in its current form may not be sustainable, but is PoS a real alternative?
An open public blockchain is one that can survive challenges in an open hostile environment where (assuming) all parties are self interested and focused on maximising their own value. In this scenario the question becomes how can you incentivise all parties to work for the greater good, while fulfilling their self-interest. This leads us to the real innovation of the blockchain, the incorporation of a native digital store of value, a token, which can be used as an incentive to compete to reach consensus as quickly as possible. Here trust is created from mistrust through work. This order from chaos (or increasing complexity) can effectively be seen as a reduction in entropy of the system. This can also be considered as fundamental to the “structural integrity” of a blockchain. Work must be done to bring order (the unlikely scenario of trust) to chaos (a more likely scenario of mistrust), and to incentivise parties to do so they expend energy (through PoW) and are compensated through an intrinsic store of value. It’s a self-contained system that is simple and beautiful in its implementation, and requires no more rules and controls than are necessary. It relies on simple rules of nature from math, economics and also (what very few people talk about) physics!
In this mental model it can be seen that expenditure of energy (PoW) is not just an interchangeable component of an open public blockchain, but is one of the most critical components that enables a blockchain to exist. Without PoW consensus there would need to be a much more complex set of mechanisms in place to ensure integrity of the system and support increasing complexity.
Consensus mechanisms for open distributed systems are a difficult computer science problem, and just because the concept of blockchains exists now this problem won’t be magically solved through something like PoS. It can be seen intuitively that PoS mechanisms have a tendency towards a centralised system, where breaking the controlling equilibrium becomes increasingly unlikely. This is a system that becomes completely pointless, as you might as well drive further efficiency and eliminate the need for duplicating cost of storage and computation of the nodes and revert to models of trusted parties i.e. remove the blockchain altogether!
As noted above PoW (as it currently stands) may not be sustainable in terms of energy usage, however it doesn’t mean that it is the wrong approach, and it certainly does not mean that PoS is the answer. While there may be viability in some form of hybrid PoW/PoS solution, these are still early days in the world of blockchain so we are likely to see some interesting developments in this critical area of consensus mechanisms.
People eventually realised that the Internet is not just a giant intranet, they may also realise that a blockchain is not just a giant perpetual consensus machine.