Hugo Nguyen on why Proof-of-Stake is a dead-end and can’t protect value.

Steel man Proof-of-Stake thesis: 

Access to large amounts of electricity (something tangible in the real world) is a lot more susceptible to regulatory capture than an intangible cryptocurrency, and that that is therefore more prone to inegality

 If you look at cost to attack of ETH’s PoS its actually much higher than BTC. For example: PoS only needs 1/3rd of the validators to remain honest, instead of 51%. It also slashes attackers instead of just invalidating their proposed blocks, making attacks far more costly and unsustainable. 

The response by Hugo Nguyen:

This is wrong in several ways.

PoS needs 2/3 of active stake to stay honest, not 1/3. The Writings of Leslie Lamport has proven this mathematically 40 years ago in his seminal paper, and absolutely nothing has changed since then. Math doesn’t change. Period.

 Your definition of “attack cost” is missing a time element. Most people make this mistake when thinking of PoS. They are able to think of only snapshot analysis (what’s going on in a given moment) not zooming out to a long time frame. All PoS models require unstaking which means your “cost” is temporary and only has meaning within that time window. Beyond that, it’s meaningless. This is the basis of “costless simulation” aka “long range attacks” which all PoS models suffer from. Also, even within the small staking window, buying coins on the open market (another naive assumption) is not the only way to acquire stake, which makes this “cost” even more of a make-believe bullshit metric. As an example, an insider/hacker with access to Coinbase or Binance staking accounts can cause havoc.

Instant Bitcoin & Cryptocurrency Exchange

Another common mistake is to assume slashing can be fair. Slashing assumes you know who’s truly at fault when in reality a simple event like network partitioning can make nodes seem like they “misbehave”, even though from their PoV they all behave honestly. PoW makes no judgment analysis. Its currency / metric is purely energy (more accurately, number of hashes spent, not “watts” like Lowery said. PoW approximates energy spent, not actual energy spent).

Bottom line is PoW anchors the coin in something real and to overtake the network you have no choice but brute forcing: spend energy and sustain it. PoS is entirely imaginary and has no anchor. In various failure modes, it will devolve into trusting a few people to resolve chain splits aka fiat/central banking. PoS is the modern version of Perpetual Motion, where people believe something of tremendous value (a robust global monetary network that even enemies can rely on in adversarial environments) can be created out of thin air. It is 100% delusional.