I was caught red-handed.
It’s all there in this tweet from Vitalik Buterin, which shows me on stage at Deconomy, arms outstretched and shilling BSV with all of my God-given might. Except that isn’t what happened at all. Because what was happening on-stage was a discussion of collective action problems in bitcoin.
For those unfamiliar with the term, collective action problems are a type of market failure and the most significant impediment to on-chain scalability. One example is the “free-rider” problem, which exists anytime there is a difference between the work a network pays for and the work it needs participants to do. Free-rider pressures are why such a massive percent of Ethereum traffic flows through Infura — and why no-one seems to have a strategy to deal with the problem.
Markets with collective action problems can develop failure modes — equilibria where everyone behaves selfishly even though they would all be better off if they cooperated, such as by jointly-funding a high-throughput network. But cooperation is hard to sustain in blockchains, first because they are open systems, and second because miners and stakers that act altruistically systematically lose market share to those that do not. And because the amount of unpaid work that any network needs to fund expands with scale, collective action problems get worse with scale too.
So simply taking offense at the mention of BSV is silly. We all understand that Vitalik has an ongoing feud with Craig Wright. But like Craig or hate him, the BSV position is important because it articulates an economic solution: the classic proof-of-work insistence that miners will sort it all out. Whether this is actually going to happen is anyone’s guess — these problems are called market failures for a reason — but the BSV approach at least tells us how the network is supposed to work: miners will coordinate with each other to punish free-riders. Far from shilling BSV, one of my points at Deconomy was that this has serious consequences: for miners to be able to punish defectors by denying them transaction flow requires a certain amount of collusion in the network. Will BSV hit the sweet spot? Is there even a sweet spot?
Were Vitalik attuned to the nuance of what was happening on stage, he would have noticed that what I was offering was a criticism of BSV, while the other panelists — whom he also smeared as shills — in fact never mentioned the name of the project. But why shouldn’t we discuss the differences between the major bitcoin forks based on their economic assumptions about who pays for growth? I suppose the panel could have talked more about Bitcoin Cash’s approach, but their team has yet to make their own assumptions explicit. The only other project with a declared approach to these market failures is Bitcoin Core, which considers them unsolvable and — as such — falls back to insisting that the blocksize must never grow large enough that incentivization problems spiral out of control.
If Vitalik wants to insert himself in a public debate about the economics of scaling, a good starting point would be for the Ethereum team to make its assumptions explicit. Are they expecting external markets to fund the network activities they can’t handle? And have they thought about the failure modes associated with this model? If the solution to an underfunded network is getting companies like Infura to subsidize access, why do Ethereum’s own protocol developers consider Infura’s dominance of the Ethereum network a threat? Should we worry about the tendencies towards monopolization in ancillary markets? Or the risk that once companies start monetizing transaction flows they will undermine the independence of the network? Is throwing the problem to markets outside the network really better than just throwing it at miners and letting them sort it out? Convince us.
Instead of addressing these issues, most developers simply leave these issues unexamined. The danger here is that blindness to the underlying incentive problems leads developers into proposing technical “solutions” that feel constructive even if they fail to assist resolution of the actual problem. Getting further into that discussion would take far more space that is suitable here.
At least as far as the accusations of shilling goes, perhaps I can make my position clear with one final point. My personal conviction is that neither proof-of-work nor proof-of-stake can solve these problems at scale without sacrificing some qualities of network openness. This is why I am working on Saito, an application platform that fixes these underlying incentive issues on the most fundamental level. And I mention this merely so there is no confusion that when I shill, I shill Saito.
In any event, I’m grateful to Vitalik for drawing attention to the discussion and the economic problems raised on stage. I hope the publicity will drive people to watch the event and think about the actual issues that were raised regarding the collective action problems latent in the POW and POS mechanisms. Because while there may be multiple solutions to these problems — the issues can’t be fixed until they are understood.
Extracts from this piece first appeared in a Decrypt Media story on 11 Apr 2019.