Resist Recentralization!

By Dominic Fox, Principal Software Engineer

The more you tighten your grip, Tarkin, the more star systems will slip through your fingers.

Princess Leia Organa, Star Wars Episode 4: A New Hope

It will always be tempting for enterprise industry players to try to reshape the world of decentralized computing and distributed consensus in the image of their own priorities: central command and control, with everything legible and accountable from a single standpoint. I’ve seen these dynamics playing out in sales pitch after sales pitch, conference presentation after conference presentation. The tech isn’t terrible, and the incentives are understandable, but isn’t it all missing the point? Shouldn’t we stop and think about what we’re actually trying to build?

The origin story of Ethereum is widely known: Vitalik Buterin, a keen World of Warcraft™ player, quit the game in despair after Blizzard software “nerfed” the powerful spell around which his warlock character’s build revolved, at a stroke decimating the accrued benefit of countless hours of dedicated gameplay. “On that day,” Buterin wrote, “I realized what horrors centralized services can bring.”

On one reading, Ethereum was born of a desire for irrevocable ownership: what’s mine is forever mine until I choose to sign it away. On another, perhaps closer to Buterin’s intent, it was born of a desire for decentralized governance: no single authority can unilaterally make decisions affecting the good of others. Transactions on the Ethereum blockchain can still be revoked, via the mechanism of a hard fork, but a majority of miners must agree to reverse recent history and proceed anew in a different direction. If the entire Warcraft community had come to the conclusion that the “Siphon Life” spell unbalanced the game, Buterin would perhaps have been content to abide by their decision.

Given this background, I was amused to hear a speaker at a recent blockchain event promoting the benefits of “hybrid” public blockchains and permissioned networks in terms of game companies being able to exercise centralized control over the movement of assets such as weapons and costume items. The public blockchain would manifest players’ ownership of tokens representing such assets, but transactions would be funnelled through an off-chain rollup controlled by the game company, which would retain the prerogative of reclaiming tokens at its discretion. If a player went on a foul-mouthed, bigoted rant on their Twitch stream, the game company would be able to enforce their terms of service and cut them off. This is certainly a very understandable thing for game companies to want to do – every public media platform carries some gatekeeping responsibilities, and those that fail in those responsibilities invariably become wretched hives of scum and villainy. But in grafting those capabilities to a technology platform specifically designed to be censorship resistant, to disperse authority and disintermediate the grand old institutions, you aren’t “getting the best of both worlds”: you’re building a chimaera which is profoundly confused about its reason for existing at all.

Why place game assets on a public blockchain in the first place? One answer is that you anticipate a secondary market arising for such assets. If they come to be regarded as valuable due to their collectability or immediate utility within the game world, then players will be attracted to your game by the possibility of acquiring wealth through their prowess or willingness to spend hours grinding. The purported advantages of using NFT contracts to manage ownership of virtual bling, besides the prospect of syphoning quick cash from speculators, center on the confidence holders of these assets place in their security and durability. As soon as you fuse these contracts with a mechanism which enables tokens to be snatched away at whim by centralized fiat, you have removed the entire basis for that confidence. Does anything remain besides the public network’s questionable utility as a data distribution mechanism?

NFTs have enjoyed a brief moment in the sun as get-rich-quick vehicles, and the lure of easy money has energized a lot of motivated reasoning about their appropriateness for other use cases. At bottom, though, an NFT represents the consensus of a network about ownership of some bundle of rights. That’s an interesting thing to be able to represent, but magical thinking about blockchains converts this into the instantaneous global manifestation of an incontrovertible fact: as soon as you record on the blockchain that player dom_fox420 holds the fabled Axe of Nopolt, every game in the world will supposedly be able to make use of that information. You’ll be able to melt the axe down for raw materials to use in manufacturing spaceship armor plating, and so on. As enticing as this vision of a universal imaginarium may be, someone has to design the rules that determine how the salient properties of the fabled Axe of Nopolt are to be represented and interpreted across all the different contexts in which it might appear. Making the data available for ingestion is the easy part, and a public blockchain is not an especially efficient or convenient mechanism for doing so.

It’s no small task to design a lingua franca for my spaceship building game and your swords-and-sorcery game which will enable the former to interpret as potential spaceship-armor-plating raw material what the latter presents as a 1.3kg lump of mithril infused with chaos magic. The advantage of being on the Ethereum blockchain isn’t that these two semantic horizons are magically reconciled, but that the basic operation of exchanging one token for another can be reliably executed regardless of what kinds of things they tokenize, and without either game’s developer controlling the mechanism of exchange. But this is only true of the layer 1 network: for this kind of free interchange to be possible, the token must slip out of the walled garden provided by the centrally intermediated rollup channel.

At the same event, it became increasingly apparent to me that The Enterprise was Striking Back, in the form of centralized integration platforms which offered to turn every blockchain in the world into a service endpoint for a glorified BPEL engine. It’s not hard to see the appeal of such systems. Managing assets on multiple blockchains to realize extended value chains means taking command of a vast and heterogeneous space. This is what the Enterprise toolchain does well (indeed, it’s the historic mission of that toolchain). All the pieces are there to integrate everything with everything else – now you just need to pay the consultants to configure them. In the process, however, both the socio-technical raison d’etre and the commercial potential of blockchain systems is whittled down to a nub.

It’s worth spelling out again what the socio-technical promise of blockchain technologies is meant to be: what draws enthusiasts into the space, what kind of future they are hoping to carve out there. Isn’t the core premise that each “node” in the system is sovereign over its own identity and history, and that standing within the network is established by rule-driven consensus among sovereign nodes rather than determined by fiat of institutional gatekeepers and referees? What gives a “blockchain asset” value is the consensus of peers over the rights it betokens: it isn’t just a row in a database table owned by an entity that gets to decide unilaterally what that data does and doesn’t mean, but the concrete and deterministically verifiable outcome of a process of reaching agreement.

The more we compromise on this vision, the closer we get to a worst-of-both-worlds scenario in which an emerging technology, with all of its unsmoothed edges and cognitive unfamiliarity, is blindly pressed into service as a risky and inefficient means of doing business as usual, purely for the sake of capitalizing on the buzz around its novelty. Enterprise software is not meant to be exciting: it seldom comes out of midlife crises involving sports cars and high living with its dignity intact. The cutting edge of the blockchain space, on the other hand, remains an enormously exciting place – even more so as the hype and magical thinking that have so often surrounded it start to drift away, and the underlying technical challenges and opportunities become more clear. We have an opportunity to invent the future: let’s not concrete over the green shoots to build yet another asset carpark.