What We Talk About When We Talk About Blockchain Technology

A Thought Experiment

Imagine if you will: it’s the year 2000. You’re making Y2K jokes, reeling over the Star Wars prequels, witnessing an American presidential election that by your naive, doe-eyed standards feels “stressful,” etc, etc.

You’re also bearing witness to something else: this new, strange, emergent phenomenon called “peer-to-peer file sharing networks.” Although you’ve yet to take the time to learn how this P2P technology functions, a few things are clear to you:

1. It works. You and your friends have been using Napster, Kazaa, Gnutella, and others, and have been able to download music and movies at no cost. Nobody needs to sit down and explain the value of this to you, because you’ve perceived it with your own eyes and ears.

2. It all feels a bit Wild Wild West (a movie, incidentally, you walked out of six months ago). The software interfaces for these tools are usable to non-tech geeks, but just barely; they certainly don’t have the level of shine and gloss of the UIs you’re accustomed to on your new, fancy Dell running Windows 98. The whole experience of participating in these strange networks — from finding and installing the software itself to seeking out and eventually downloading digital content — feels less like the increasingly polished world of web-commerce than the some weird dark-web / hacker/ cypher-punk shenanigans (a culture which, up until this point, you’ve only heard about from an intelligent but increasingly suspicious friend of yours.)

3. There’s good reason to think that many people with money and power desperately want to stop this technology from spreading — namely, the music and movie industries, for the obvious reason that people across the world having access to their product for free severely undercuts their business model, and the federal government, because — let’s be frank — this is clearly illegal as all hell. Any yet, despite all of this money and political muscle being utilized to attempt to shut down these networks (including the arrest of a 12-year-old girl, you could’ve sworn you read), they persist virtually unscathed. From this, you conclude that however this technology is working, it’s doing something quite powerful and unprecedented.

Now let’s suppose, in this hypothetical thought experiment of ours, somebody approaches you one day, let’s call him Joe Finance Guy. Joe is in his mid-50s, and prides himself on his knack for detecting the latest trends in the tech industry and capitalize on them, a claim whose validity is evinced by his gains in his investments during the dot-com boom. While he readily admits he has limited technical knowledge of computer science, cyber-security, or really anything software related, he still insists that that’s no hindrance for him, given his penchant for surrounding himself with the right people.

Needless to say, Joe has had his eyes on the P2P revolution. Spotting what he suspects is the “next big thing,” and noticing that, in its current state, no single entity seems to be making vast sums of money off of it, Joe is feeling his business-minded ears perk up. So today, he’s called you into his office for a modest proposal.

Here, he says, is the plan: We’re starting a business! Our company will advance P2P technology to its next phase — digital content sharing 2.0, if you will. For the first time, the technology will be rescued from the scary realm of hackers and piracy and so on, and into the world of mainstream institutions. What we’ll do is set up our own peer-to-peer network in which our private, registered company controls all of the nodes in the system. What this means is that things will work much more smoothly — search queries and network traffic can be coordinated by our central servers, making things far more efficient. We’ll hire the best designers and UX experts we can find, so the whole thing will feel cleaner, polished, and inviting. And of course, we’ll work with authorities and regulators to smooth out all those messy legal issues. So we eliminate all the downsides of P2P networks by keeping it all under our control, but, because we’re still utilizing the same technological infrastructure, we’ll preserve the benefits! After all, the real power here is the technology, remember? You said it yourself.

What are you to make of this proposal? All upsides with no downsides certainly sounds enticing — is this your chance to cash in big on a paradigm shifting trend at just the right moment?

Never one to shirk due diligence, you go home and embark on your first real research deep-dive into P2P network architecture. And here’s what you come to realize: Joe is completely wrong. His idea rests entirely on a lack of understanding of the technology he is talking about. In fact, it could be argued that, despite his slick presentation, what he proposed actually offers nothing but downsides.

In your second official meeting with Joe, you take your best shot at explaining:

Technology is often best understood by what motivates its creation. We’ve had open internet access for years, but it wasn’t until the emergence of P2P networks that this proliferation of free file sharing was made possible. What is it, then, that this new technology is actually offering? Or, conversely, why did we never see a similar outcome (free digital content sharing) achieved with the traditional web-toolkit (client/server architecture)? What were the impediments?

For simplification, we can distill these impediments down to two:

1. Limited Hard Disk Space: Image, audio, and video files take up an awful lot of hard drive space. For one company to maintain a database of every song that you’ve ever heard of would be prohibitively expensive, to say the least.

2. Risk of Censorship: For better and/or worse, there exists these things called intellectual property laws; if a company is illegally distributing massive amounts of data illegally, the FBI just has to show up at their door, and down it all goes.

What’s made P2P architecture so paradigm shifting is how it’s solved these two problems: in the case of data storage, no single party has to host it all; anybody with a standard household computer can run the software and seed whatever data they choose. Most people have plenty of excess hard-drive space and bandwidth on their personal computers anyway, and running a node is easy and convenient enough that enough people seem to be doing it to keep the network alive. Thus, instead of a massive burden of financial cost falling on one entity, the cost is distributed widely enough among many parties to make it virtually non-existent.

As for legal censorship, these nodes are numerous enough and geographically distributed enough that it becomes infeasible to shut the network down. Most files — especially files users are most interested in — get duplicated dozens, or even hundreds of times across different nodes. This forces authorities to play a Sisyphean game of whac-a-mole; at best they can occasionally shut down a major node or two, but given how many there are, and the fact there are constantly new ones going live, this does little to stop the network from functioning or even slowing down.

Which is to say: while the benefits of P2P are indeed technological (this is software, after all), analyzing the tech in isolation misses the point: the “technological utility” here is inextricably linked to its utility from an economic, political, and even social perspective. In short, the real-world context this technology lives in is what matters, not just the software itself.

Re-examining Joe’s proposal: where it fundamentally differs from existing P2P networks is that the company will be running the whole system. So consider the burden you are now taking on: they have to pay to keep all nodes in the network running. Which, recall all the duplication/redundancy involved, will be not just pointless (why bother storing 100 copies of a file you own anyway?) but also far more expensive than the already prohibitively expensive (hypothetical) case of storing a giant music library in a query-able database. And additionally, now all the FBI needs to do is confiscate your server keys and smash your hard drives (which, they almost certainly will) and your whole system is toast.

What you’ve done is create a system that’s expensive, slow, and easy to dismantle by hackers, authorities, or anyone in between; you may as well save yourself time and money and build the thing using the client/server architecture of most of the net. In which case, of course, you’d still ultimately run into the same problems, so you’d be required to come up with brand new business models, figure out new legal frameworks for handing intellectual property, and yes, sure, you’d also need some technological advances. And who knows, maybe some giant multinational corporations years from now will manage to do just that, ushering in a future online downloading and streaming of digital tunes and ‘flix. But let’s be clear: those technological advances have essentially nothing to do with those of P2P networks. The idea of a Kazaa-esque P2P network entirely controlled by one entity is non-sense, which is why it hasn’t happened yet, and why it never will.

An Actual Experiment

Fast forward to the year 2018 (sorry, but we must.) Here we all find ourselves, trying to make sense of blockchain technology, and all of the bold, confusing, seemingly contradictory claims surrounding its purported benefits. If you don’t work in blockchain tech but know somebody who does, there’s a good chance that you’re sick of hearing about it. But bear with me — our neat little P2P-tech-proposal-analogy should set us up to better understand some thorny confusion around blockchain about what exactly its value proposition is, and why its supposed limitations — and certain proposed solutions to said limitations — may not be what they appear.

In a word: what blockchains give you is disintermediation. More to the point — they give you disintermediation in managing the ownership and transfer of digital commodities [1]. Unlike physical commodities, which take time and resources to duplicate, digital commodities are represented as bits on a hard-drive, and bits can be duplicated for free, for all intents and purposes. Free duplication is no small issue for commodities, which obviously lose their economic value without scarcity. Thus, in the pre-Blockchain digital Stone Age, managing digital commodities required granting authority to a trusted third party (TTP); the TTP processes all financial transactions and is ultimately responsible of managing and updating the canonical data-ledger , their own digital representation of ownership.

Why might we want alternatives to TTPs? Aren’t they functioning just fine, or at least well enough, as is? Cryptocurrency advocates argue that the use of intermediaries is so fundamentally built into the way we think about electronic payments that many fail to notice the incredible costs we’re constantly absorbing. Even ideological grievances — about financial institutions’ increasing concentration of power rendering them more prone to corruption, etc. — aside, consider purely financial grievences: we may be doing more work than we need to be doing, and are losing time and money in the process.

Nick Szabo has convincingly argued that when it comes to computer security, TTPs inevitably evolve into the slowest, costliest and riskiest component of a protocol; keeping them honest, functional, and accountable requires layers upon layers of oversight and adjudication. These levels of oversight take many forms: auditors, accountants, contracts, insurers, lawyers, and ultimately, police, courts, judges, and hell, even bail-outs; essentially, the gears that typically run human civilization itself. Integrating these layers of human wet-ware (identity verification, transaction verification, regulatory compliance) into the fundamental components of what is otherwise dry software will, by necessity, burden and slow the system down and introduce all sorts of attack vectors. In short, computers are better at interacting with other computers than they are with people.

Now even if you don’t buy into the theoretical underpinning laid out above, if we just temporarily grant that TTPs in a software protocol present risks, we can then categorize blockchains as existing at the extreme ends of the solution space: we start with software that requires the tightest security imaginable (money itself), and we not only ease reliance on third party trust, but attempt to effectively eliminate it entirely.

How can such a system still possibly be secure? Or, put another way — what’s replacing all of those human-civilization-underpinning-levels-of-oversight discussed above? The (bizarre) answer, in essence, is that the burden of all of that human-work is being offloaded onto the software itself.

This “burden” takes many forms; blockchain nodes are limited in the amount of transactions they can process over a given amount of time; an investment of resources must be made to participate, either in the form of raw computational energy (proof of work) or putting up money as collateral (proof of stake). The details of which aren’t important here — what matters is this: these aspects that make blockchains “slow” are an intentional and deliberate way of avoiding reliance on trust and concentration of resources, not just wonky computer science problems to be solved away with fancy algorithms. A typical computer science problem assumes a computer is under the control of the developer, who then tries to optimize for some metric of efficiency. The challenge of blockchain development, however, is to work without the fundamental assumption of control, and figure out a way to best ensure that whoever runs the code will do so honestly; it’s equal parts instructing computers and incentivizing the humans running them.

Which brings us back to poor, hapless Joe. When it comes to both P2P file sharing and blockchain networks, the “inefficiencies” are largely a means to an end, that end being decentralization, disintermediation and censorship resistance. Working around these problems by centralizing isn’t a solution; it’s simply giving up. But it’s worse, because to whatever extent you naively maintain these costly components of the protocol is the extent you’ve now imposed empty costs on yourself. Like Joe thinking that running all 100 nodes in his P2P network himself is good idea, a cryptocurrency on an entirely “centralized blockchain” unnecessarily takes on blockchain’s burden while achieving something that could be far more easily done with a good old fashioned database.

Hence the frustrations, to the ears of a techie, of hearing debates about the extent to which blockchain networks should be “centralized,” as though it’s just one of many factors to consider on your checklist (along with, say, transactions per second, fees, developer team, native features, etc.) It’s a bit like listening to a debate over the merits of various different airplane models, with one of their many equally weighted factors being whether or not the thing can even fly. Surely if it never leaves the ground, then whatever it is, it shouldn’t be called an airplane — can we perhaps agree that your centralized thing shouldn’t be called a blockchain [2]?

Now of course, in the world of digital file-sharing, centralized powerhouses did in fact win out in the end; odds are that you use iTunes, Spotify, and Netflix more often than you do Limewire and BitTorrent. So given that we’re drawing parallels between P2P and Blockchain tech here, it feels only fair to consider the possibility that Bitcoin and Ethereum will ultimately yield way to something similarly centralized and corporate. Exactly what this beast would look like — Amazon issuing its own purely digital token? — is hard to say. But hell, maybe it’ll happen; I’m not here to claim that I can rule it out [3]. But what I submit to you is this: if AmazonCoin comes to be, and Amazon decides they want complete control over its underlying protocol (and does anyone even pretend to imagine they wouldn’t?), Amazon will find they have no use for Blockchain tech itself, and for the same fundamental reasons that iTunes had no use for P2P architecture.

The economic viability of a particular industry and the utility of a particular piece of technology are two very different discussions. It could very well be the cryptocurrency trend will prime the world to embrace privately issued digital commodities; claims about the future of blockchain tech itself, however, are another matter entirely. To many, it instinctually feels like the safest, most conservative bet to assume that the zany and un-regulatable will wield to the boring, institutionalized, and corporate. But when you’re talking about technology whose value-proposition is, in effect, removing reliance on central parties, the “safe bet” turns on its head. Admittedly, there is something weirdly satisfying about witnessing this confusion play out: if blockchain tech can really do what the boldest among us suspect it might, the supposedly naive crypto-anarchists types may be the only ones actually grounded in reality, while the supposedly pragmatic, business-oriented, “it’s about the technology, stupid” crowd may prove to be hopelessly confused.

Speaking of confusion, here is one prediction: projects will continue to emerge that intentionally muddy their true nature; they’ll coops the language of truly open, distributed blockchain networks to obfuscate the fact that their technology positions them to be much more like a highly-power concentrated AmazonCoin-esque player described above. The level of smoke and mirrors on this topic in the space is already astonishing; if, God bless you, you’re some sort of investor for whom sorting the wheat from the chaff is already literally a full time job, I can do little more than to wish you the best of luck.

Welcome to the future of money. Trust no one.

Notes:

1. For the sake of it, let’s just use commodities as catch-all term for “that which has monetary value” so we don’t get lost trying to determine exact line between currency, digital assets, goods, etc.
2. A retort you’ll sometimes hear to the “stop calling it a blockchain” complaint is that we should still call things blockchains so long as they have both “blocks” and “chains”, which, admittedly, is more persuasive than I’d like to admit.
3. Though it feels like it would be dishonest to neglect to admit that the prospect does kind of scare the shit out of me.