‘Blockchain’ is a semantic wasteland

semantic

#1


via medium

Blockchain this, blockchain that. It is a concept so momentous that it managed to shed even its article. Proponents don’t speak of the blockchain or a blockchain. Instead, they reverently preach Blockchain: the solution to all enterprise needs, in particular supply chain management. The brute, unassailable, self-evident concept has disrupted not only the rules of commerce but of grammar, too. Question it, and be exposed as a hopeless rube and a luddite. Use it sincerely, and be lumped in with the hypemen and techno-utopians.

It is impossible to avoid. TV ads for IBM blare out promises about revolutionizing tomato-tracking with Blockchain. The UK finance minister recently claimed that Blockchain might be a solution to Britain’s Brexit woes. At a recent conference held by dubious-blockchain Ripple, former US President Bill Clinton said of blockchain: “the permutations and possibilities are staggeringly great.”

Blockchain is satire-resistant, because it is such an obvious target, and hence not funny any more. It is simultaneously the glittering future and the dismal present — almost all current touted use-cases are obvious nonsense. Blockchain proponents are almost totally immune to ridicule. Nothing arrests their indefatigable enthusiasm: there are press releases to be written, conferences to attend, and corporate R&D dollars to waste.

The ICO mania of 2015-17 that is now unwinding was premised, in large part, on the ability of blockchains to disrupt firms with markets. (In simpler terms, the idea was to uberize every conceivable service, and replace the intermediary with a magic database of who is doing favors for whom.) Some of those pitches invoked trillion (yes, that’s trillion with a T) dollar total addressable markets.

Both today’s soulless corporate Blockchain, and the opportunistic, ICO-based blockchains have been much scorned and ridiculed. Yet the term persists, an empty semantic husk, kept alive by a thousand press releases, conveying as little meaning as possible. Well past the point of no return, the term is used to refer simultaneously to projects, structures, and databases that have virtually nothing in common. As a consequence, attempts to define it are hopeless, general failures. In this post, I’ll try to explain the origin of ‘blockchain’, and what we should do about it.

Where the hell did ‘blockchain’ come from, anyway?

Most histories of ‘blockchain’ will mention that Satoshi Nakamoto created the first one. Except… he/she/it/they didn’t. Nakamoto never referred to bitcoin as a blockchain, calling it instead a ‘chain of hash-based proof-of-work,’ a ‘chain of blocks,’ and even ‘timechain’ (in an early comment within the original codebase). Imagine. We were so close to living in a world of enterprise timechains and strawberries-on-the-timechain.

Satoshi was careful to emphasize that the chain was a set of proofs of work, each linked to the hash of its parent. (See for yourself!) The proof of work is absolutely essential to the concept. It is proof that anyone proposing a block has, well, worked for it. It enables the system to achieve sybil-resistance and to come to convergence (the longest chain [under the same ruleset] is the correct history, by definition) without any single arbitrator.

The data structure, the inclusion of hashes of previous blocks, ensures that the past is preserved and the database is consistent. Replicating the database to every node in the network ensures that it can’t be shut down or altered unilaterally. The whole system was built with an adversarial context in mind. Hostile governments had shut down all previous attempts at e-cash. They certainly would have shut down Bitcoin if they could have. Thus, it was built for purpose. To clarify: Satoshi may have created the first popular, widely-used linked list structure — not the first of its kind— but the innovation was in the merging of the computational hardness of adding new entries to the chain with the linked list.

Does this sound to you like what any of the enterprise blockchains are trying to accomplish? Of course not. There is no shadowy organization dedicated to forging strawberry provenance records (that I’m aware of) that might seek to interfere with IBM’s supply chain blockchain. Thus, it does not need to be built to the same standard as Bitcoin. The kinds of records preserved on enterprise blockchains do not need the kinds of protections that Nakamoto consensus ensures. They do not need or want open validator sets. Some trusted organization could just vouch for the database, or a consortium of interested parties could share records between them. For more on the failures of private blockchains, I recommend this take from a reformed private blockchain consultant.

Why is the term so popular?

From what I can tell, people witnessed the success of Bitcoin, which relies in part on an ersatz, expensive database, and wanted to generalize it to other uses. Even early Bitcoin developer Hal Finney mused about disaggregating the data structure from the monetary system.

Hal Finney, Cryptography Mailing list, Jan. 24, 2009

However, to the best of our knowledge, these systems only really work if the reward is internal — that is, if well-behaved validation is rewarded with the “native” token. If Bitcoin miners were paid in USD, they wouldn’t necessarily have any incentive to mine on top of the longest chain. Recognizing the value of their hardware depends on the continuing existence and flourishing of the chain they’re building on top of. But private, permissioned, or enterprise blockchains do not have native currencies, or issue monetary units to validators, as the validator set is permissioned and thus has sybil resistance and good-behavior built in, by design.

The reason ‘blockchain’ is such seductive marketing, in this critic’s opinion, is the subtle implication that the data structure alone — absent PoW or open validation — could convey the same benefits as Bitcoin, without the token or costly anti-sybil protection.

Patri Friedman puts it well:

Bitcoin is a monolith which colors the way investors and corporate R&D offices think about similar projects. Would Ripple have been as popular without Bitcoin ever having been invented? What about Corda and Hyperledger? Litecoin? ICOs generally? Ethereum? It is hard to even imagine the alternative history, but I suspect the answer is no in all cases. Bitcoin is a juggernaut which carried with it a set of assumptions that were ported over to projects with a passing resemblance, rightly or wrongly.

Thus, I would be suspicious of anyone that routinely uses ‘blockchain’, especially if they are trying to sell you something. Overuse of ‘blockchain’, especially in a general context, and without qualifiers, reveals one of three things about a person. They are most likely:

  • Well-meaning people who are forced by convention to use subpar linguistic tools
  • A bit muddled, and trying to mask their ignorance with technobabble
  • Trying to posture as experts in an industry which realistically has no experts

I firmly believe that the misuse of the term traces back to a desire to create (or market) systems which are intended as Bitcoin-like, without the unsavory bits. But that misses the point — Bitcoin’s blockchain is just a part, and not its essence.

Bitcoin and its blockchain

If I could have you take anything away from this piece, it’s that referring to Bitcoin as a blockchain is like referring to car as a transmission. Yes, that’s a key element of the system, but does not represent it in its totality. ‘Blockchain’ is a metonymy — a part used to refer to the whole. There’s nothing wrong with that, intrinsically. The conceptual tangle comes when one decides that the blockchain is Bitcoin’s essence, and is owed the credit for its success.

Bitcoin relies on a linked list, indeed. But it also relies on a p2p network, an open source and leaderless project, a replicated database, a self-supporting incentive system, a heaviest chain consensus rule, and a proof-of-work (PoW) scheme which gives block proposals unforgeable costliness. (Unforgeable costliness in simple terms: it’s impossible to fake a block submission — you would have to have allocated a good chunk of computing power (energy) to the task. It is therefore hard to create new Bitcoins, but easy for anyone to verify that you worked hard at it.)

These inputs combine nicely to create a system which has certain qualities: provable scarcity, auditability, tamper-resistance, fair-ish distribution, almost perfect supply inelasticity (rising price does not — can not — cause production to accelerate), free participation (no one can stop you broadcasting a Bitcoin transaction), and so on. These qualities make Bitcoin unique relative to, say, Paypal or Visa. They are its differentiators. Without the p2p nature, the open source collaboration, the voluntarist developers, and crucially the PoW block proposal method, Bitcoin does not exist. The below chart, created by David Puell based on ideas from Pierre Rochard, is an attempt to capture Bitcoin’s essence. Notice that the chain of blocks, while necessary to make the system work, is not sufficient on its own. Bitcoin relies on more.

David Puell’s laudable attempt to characterize Bitcoin’s essence in one chart

I can’t tell you exactly what the essence of Bitcoin is, but to limit it to a chain of blocks is reductionist in the extreme. In short, the soul of Bitcoin is not the blockchain. And if you pull the blockchain out of Bitcoin, you get something rather empty.

Why rail against ‘blockchain’?

My objective is to clarify popular discourse in this industry, with the view that better conceptual frameworks will lead to better outcomes. Orwell believed that the words we use directly affect the way we see the world. He even intimated that scrubbing words from popular usage could eliminate their referents — the very concepts they sought to represent.

The purpose of Newspeak was not only to provide a medium of expression for the world-view and mental habits proper to the devotees of IngSoc, but to make all other modes of thought impossible. It was intended that when Newspeak had been adopted once and for all and Oldspeak forgotten, a heretical thought — that is, a thought diverging from the principles of IngSoc — should be literally unthinkable, at least so far as thought is dependent on words.

Eliminate the word “freedom” from popular use, and eliminate the desire for freedom entirely, so the theory goes. Additionally, he strongly felt that sloppy language was indicative of muddled thought, and used as a way to sneak indefensible assertions past an unwitting reader.

My point here is that in elevating the linked list to the exalted status of Blockchain, we overrate it; and in insinuating that Bitcoin is just a blockchain, or simply the origin of the more interesting underlying technology, we do it a disservice. In constructing the Blockchain artifice in the popular consciousness, we enable a sloppiness of thought and do away with rigor. ‘Blockchain’ dilutes the importance of a tremendously important and valuable innovation — a trust-minimized monetary system — and abases it by putting it to work to generate efficiencies, real or imagined, in enterprise supply chain management.

The path forward

If I may, a few simple suggestions:

To permissioned/enterprise blockchainers: be honest about what you’re building and in your marketing! If you’re building a database controlled by a consortium of pre-permissioned entities, don’t claim or imply it will have similar reliability characteristics to systems that are designed to live in far more adversarial environments. Imagine how you would market your system had Bitcoin not been invented. And — let the public blockchains be. You aren’t competing with them. Your systems have totally different goals. If you do want to persist with the blockchain moniker, I encourage you to very carefully define what you mean by ‘blockchain,’ and be sure to distinguish your system from open, public blockchains. And for god’s sake, give blockchain back its article (refer to a or the blockchain, please).

To the computer scientists : stop mocking non-technical people for using ‘Blockchain’. You’re missing the point! They aren’t really referring to the data structure. So it’s besides the point to say “just use MySQL.” Blockchain, for better or for worse (definitely for worse) has become a term of art which is typically used to refer to the whole system — economic and social — rather than just the data structure.

To regulators : please do not try to define ‘Blockchain’ or create blockchain regulation. You will fail, not due to your lack of astuteness, but because blockchain is so semantically dispersed as to be un-definable. Definitions need to be specific and useful, and also general enough to encompass all of the members of the set. However these tensions tear ‘blockchain’ apart. It is used too generally to be useful.

Instead, dis-aggregate. Recognize that legislation that covers cryptocurrency probably cannot cover security tokens, ‘utility tokens’ and permissioned blockchains too. Private or enterprise blockchains aren’t just “Bitcoin in a suit” — they’re totally different. The two really have nothing in common.

To everyone else : please join me in spurning ‘blockchain’ at every opportunity! Let’s try and devise new terms which are specific enough to be useful, and do justice to their referents. I currently use ‘public blockchain’ to describe permissionless, open networks like Bitcoin and Ethereum, but I would love to use a different term which doesn’t rely on the B word. If you do insist on using it, I like Peter Todd’s definition best:

Using it a minimalistic and directed way removes some of its conceptual weight. This eliminates its ability to implicitly promise amazing consistency, reliability, and uptime. The more honest your definition of ‘blockchain,’ the less it lends itself to exciting press releases. The blockchain, in Todd’s definition, is really just a way to arrange data. And that is supremely unsatisfying, given how it’s used today.

If you want to read more about dis-aggregating these systems, I recommend the DLT Systems Conceptual Framework, published by an interdisciplinary set of practitioners and academics under the aegis of the Cambridge Center for Alternative Finance.

Canny readers will notice that I cofounded a firm that invests in blockchain startups. This is true. I am humiliated. But our usage of the term is a matter of practical reality. Such is its proliferation that it has become a Schelling point — an easy meeting place where technologists and allocators can communicate. Out of convenience, and so we are understood, we use the term. But we’d love to abandon it. It encompasses many distinct concepts, some of which we love, and some of which we hold in contempt.

I believe that in five or ten years, we will look back at the popularity of ‘blockchain’ and be slightly embarrassed. The term will seem as archaic as “surfing the world wide web” or using the “information superhighway.” Consider this an open solicitation for replacements to the term. Let’s move on.

Postscript: In researching this article (basically, this involved watching IBM’s cringeworthy blockchain ads) I was invited to give IBM some feedback.