This month, Ethereum completed what has been dubbed “The Merge” — transitioning from a Proof of Work consensus mechanism to Proof of Stake. It is an impressive technical accomplishment. However, what is lost in the hype is the significance of the problem it is attempting to solve.
All public blockchains are effectively trying to create a digital ledger that accepts and settles transactions from anyone at any time. To achieve this, the ledger must be replicated and validated by all network participants to ensure a consistent view of token ownership. The blockchain must also have a mechanism to determine the time ordering of transactions — critical to preventing the “double spend problem,” where the same digital token is spent with two different parties simultaneously.
The TL;DR: there is no universally “right” consensus mechanism. It is context dependent.
Proof of Work
Proof of Work entails the authentication of transactions by a collection of transaction signers called miners. Miners produce signatures by a process called mining — they produce proofs of work on successive blocks of transactions. Miners are anonymous, making it incredibly difficult to bribe them to produce invalid transactions or censor certain participants. Mining is permissionless — participants can enter or exit at any time. The key feature: there is a real-world cost to publish new transactions. That cost is energy. The real-world cost is important because it precludes a handful of privileged entities from arbitrarily creating currency units at no cost.
Proof of Stake
Proof of Stake replaces energy expenditure with economic stake. Validators lock up tokens as collateral and are chosen to validate transactions based on their stake. It is significantly more energy-efficient than Proof of Work. However, it introduces a different set of tradeoffs — particularly around centralization risk and the ability of large stakeholders to accumulate disproportionate influence.
My view: In the context of Bitcoin — whose narrative is that of a scarce digital bearer asset — Proof of Work is more logically consistent. It anchors the protocol to a real-world cost, which supports the scarcity properties of the underlying code. For Ethereum, whose stated goal is to become a world computer facilitating a diverse ecosystem of applications, Proof of Stake may make sense. Different goals warrant different architectures.