Interesting debate broke out on Twitter between advocates and opponents of proof of stake consensus algorithms.
It all started as Zcash founder @zooko asked for good resources to learn more about PoS. He got a quick reply from Nic Carter, a Bitcoin maximalist and Catle Island Ventures team member.
you have to look externally for these critiques. those with a vested interest in designing these systems are, unsurprisingly, not very introspective. which is why there is a dearth of good criticism: it's hard to understand the complexity of PoS and few outsiders do.
— nic carter (@nic__carter) September 28, 2018
Carter goes on to say that Ethereum’s FAQ intentionally ignores important questions raised by PoS skeptics like PoS having lower churn in the validator set (mining being costly, staking being costless). As Carter views it, this “FAQ optimizes for “PoS works because [economic formulae] and [additional complexity]” and fails at its central goal, which should be convincing outsiders. It preaches to insiders. PoS is a political and social issue, not just a software problem to be solved. The problem that no PoS proponent has ever convincingly answered is as follows: why should a costless-to-exercise perpetual anti-dilution right be granted to the originators of an economic system?”
Then it gets interesting as another high-level defender of PoS jumps in – Ari Paul David, CEO of BlockTower Capital, another crypto investment fund.
1/ Wanting ongoing dilution is a weird talking point in favor of Bitcoin. Any BTC holders face little dilution after the second halving forever forward. The control distinction is maybe meaningful, maybe not.
— Ari Paul (@AriDavidPaul) September 28, 2018
He goes further to explain how game theory works in PoS trying to drag bitcoin’s PoW into the debate as a system that has same issues as PoS does. That triggered another well-known crypto name to join the discussion, on the side of PoW and against PoS, Eric Lombrozo.
Miners suffer serious opportunity costs during the attack (they need to expend tremendous resources and receive no direct compensation) as well as the costs associated with destroying the value of the coin.
— Eric Lombrozo (@eric_lombrozo) September 29, 2018
Strongest point in this discussion came from Nic Carter again, as he explained how bitcoin miners lose their strength in the PoW system as ASICs power depreciates. On the other hand, staking is costless and who ever gets to originate the system can retain power perpetually. As Hugo Nguyen put it concisely:
“Irretrievable sunk cost” is the key in PoW.
The “stake” used in PoS creates a temporary illusion of security/immutability. It’s neither sunk nor irretrievable, and that’s a fundamental flaw you can never get around.
Lombrozo went on to further cement the PoS as perpetual motion machine causing another prolonged debate with Ari Paul.
There are several PoS systems at scale today. How do you see them failing in practical terms, and why haven’t they failed yet?
— Ari Paul (@AriDavidPaul) September 29, 2018
When asked by Lombrozo to show him these systems, Ari bailed out by saying the topic is complex and requires more time and he is doing a post about it. Then he mentioned Ripple as an example of PoS system that, admittedly has vulnerabilities, but works on a large scale.
Ari then continued in somewhat philosophical fashion, claiming that after all, people are behind everything and you need to trust people. Eric Wall hat a different opinion on it:
You don't need blind trust. You simply need to validate that the transaction you have received is in the chain with 10^27 hashes (with atleast
5949437371609*6 more after it). All you need to trust is the cost of SHA256. Consensus rules are *supplementary* security ontop of that.
— Eric Wall (@ercwl) September 29, 2018
The discussion is still ongoing and will probably stay as one of the hot topics in the industry for time to come. Both camps have solid arguments although PoW is a more trustless and proven concept that PoS is currently.