The Luddite

Source images here and here.

This is a review of Joshua Dávila's Blockchain Radicals: How Capitalism Ruined Crypto and How to Fix It in 2 parts. The first critiques the book's approach to argumentation. It is, if you like, a theory of the book. The second part looks at Dávila's own Breadchain Cooperative as an example of the theory put into practice.

Part 1: Is theory software for humans?

Inside Joshua Dávila's Blockchain Radicals, there lives a better, much shorter book, maybe titled A Leftist's Guide to Crypto. I would have many quibbles with that book, but Dávila both knows and explains crypto very well. I have no doubt that he is an exceptional developer, or that his politics are sincerely held and, in the end, not too dissimilar from mine. Because of this latter point, I hesitated to write this review. Ultimately, I decided to do so because I get the sense (or, at least, I hope) that tech workers are developing a nascent class consciousness. Historically, we in tech have preferred to think of ourselves as one lucky break away from founding a unicorn, but more and more of us are realizing that, actually, we are one misfortune away from poverty. Tech workers have consistently allied with capital to degrade working conditions, like in so-called gig work. For those looking for an alternative application of our skills, Blockchain Radicals might open an appealing route to get involved in socialist politics because, as I'll soon argue, it does political theory the way that developers write software. The book pairs in-depth and sophisticated discussions of technical systems with facile, even nonsensical interrogations into social and political theory. Were I not a software developer myself, I might have assumed that there are two authors: one a strong technical writer, and another who does not understand the subject matter or how to engage with it. This review is an intervention directed at this mode of thinking.

The introduction frames the book as an attempt to correct what is, for Dávila, the left's unfair and ultimately self-limiting conception of crypto as intrinsically right-wing. "Critique Economy" writers, he argues, in their efforts to "enshrine themselves as gatekeepers of what constitutes appropriate left-wing endeavours and resources," have fallen for a trap to "keep the left out of the sphere of influence, limiting its capacity to decide on the direction that this new technological space should take."

Those of us who spend a lot of time online probably recognize this argument structure. Similar to saying that we need neither growth nor degrowth, neither patriarchy nor feminism, or neither capitalism nor socialism, these arguments often display the author's ignorance of one or both positions. It's a sort of false dialectic: It presents a thesis, then a caricature of the antithesis, finally resulting in a synthesis that is not a new idea, but a diluted rearticulation of the actual (as opposed to caricatured) antithesis.

In this case, the "pessimistic techno-determinism" claim is a strawman. Take, for example, Molly White, probably crypto's single biggest critic. She is both a software developer and frequent contributor to Wikipedia. This is not the work of a pessimistic techno-determinist, but of a sophisticated technologist navigating the complex intersection of social relations and technology. If she, and others like her (I include myself here), have taken a negative stance on crypto, that merits serious treatment. The book doesn't engage with these critics or the various frameworks that they expound, like, for example, Luddism. Had it done so, it could have avoided some of its most egregious analytical errors. Instead, it takes a defensive posture, dismissing them all in the introduction and rarely looking back.

Most of the book is structured around what are, for Dávila, 3 incorrect but increasingly more useful ways of thinking about crypto: crypto as money, crypto as finance, and crypto as coordination. Referencing Deleuze's critique of representational thinking in Difference and Repetition, he argues that most people's understanding of crypto is trapped within these frameworks. I don't disagree, and I found this to be a productive organizational scheme, but using Deleuze to explain this idea is like referencing Einstein's general relativity to explain why you slipped and fell: It's not necessarily wrong, but it's overpowered and not particularly salient. You can easily explain this without Deleuze, as I did.

Normally, I would let this slide, but he does this often, and it's crucial to understanding the book. He frequently starts his arguments by introducing a big idea — Deleuze, art, housing justice, cybernetics, climate change, international economic coordination — only to carelessly bat it around for a few paragraphs like a bored cat. Then, instead of napping in a sunny window, he returns to the familiar routine of describing crypto projects in detail, which, again, he does quite well. Here's a quick, inconsequential example, just to illustrate what I mean. I chose it because it's uncharacteristically short and its self-evident absurdity requires no further comment:

It's a lot easier to make these messes than to clean them up, which makes this a challenging post to write. With that in mind, in part 1, I will focus on 3 excerpts, which I will reproduce almost in their entirety. Blockchain Radicals is not a short book, and, given its idiosyncrasies, there's no way to exhaustivley rebut it, nor is that my intention. My excerpts will not be a representative sample. Instead, because I have seen this book title bouncing around in my circles, and have noticed leftists becoming more crypto-curious, my goal is to point out its structural deficiencies such as to cast doubt on the soundness of the project. I want to reiterate that much of the book, especially the technical parts, is well researched and well written, and Dávila brings an impressive knowledge of crypto to the project. On the other hand, for every excerpt that I critique here, I have several more that it pains me to leave out.

To call this idea underbaked would be misleading, because the dough isn't even assembled. In some ways, this is an uncharitable choice of excerpt, because the technical part is uncharacteristically weak, and there are many examples in the book featuring quite sophisticated crypto schemes, but, ultimately, that complexity is only a distraction from the structural problem that we'll see repeated throughout: We get some big ideas about, in this case, housing and human rights, then an associated crypto scheme, but that crypto scheme has no real tie back to those big ideas.

To zoom in on how weak these threads of argumentation tend to be, here's another excerpt. This is a section titled "Expanding Democratic Expression," reproduced almost in its entirety, meaning that there is no additional context that helps it make sense. Note that it mentions DAOs, but all you need to know is that they're a kind of organization whose core governance functions are defined on and mediated through blockchain operations:

Here's an outline of what's happening, so that I can refer to the separate points by number.

1. Introduce POSIWID
2. Current government systems are bad in a POSIWIDian sense
3. The rise of populism shows that many people of diverse political orientations agree that something is wrong
4. To change the system, those people need to find each other
5. Introduce Schelling point, a place where people who can't communicate with each other can find each other
6. Zucotti park was a Schelling point Occupy Wall Street; Sanders and Corbyn campaigns were Schelling points anticapitalists
7. Gitcoin considers Ethereum a Schelling point for people who want to do good things
8. Description of stuff Gitcoin did
9. Crypto makes democratic stuff possible that can't be done without a blockchain

First, the concept of POSIWID adds nothing to the analysis. Starting at 3 would leave the argument, such as it is, intact, once again showcasing Dávila's conceptual profligacy. 3, then, is a sort of problem statement, and 4 is the vague and dubious transition to crypto, which I bolded in the original text. In one single sentence, we gloss over the entirety of how social change happens. What does it mean to recognize each other? More importantly, what is stopping these people from communicating? Something must be; otherwise the jump to 5 makes no sense, since Schelling points are where people who cannot communicate try to meet. Even if we grant that something is keeping them from communicating, point 6 doesn't follow, because we know for a fact that participants of Occupy Wall Street and supporters of those campaigns are part of a pre-existing web of communication, through shared news sources and social media. Gitcoin's claims in 7 are equally confusing, since they presumably wrote this literally in a communication to people, making it just a regular meeting point, not a Schelling point. 8 is just a description of stuff Gitcoin did, and the tie back to the problem statement is, at this point, severed several times over.

I want to drill down further on that bolded section in the second-to-last paragraph as an example of how this argument breaks down on multiple scales, almost always at the connection between crypto and stated social goals. Consider this language: "many of them near impossible to try without something like a blockchain." "[M]any" means that the claim only applies to some of the items, and "something like" renders it meaningless. Even attempts to evaluate it with other evidence in the text don't help. If you consider the three examples that he provides in text (conviction voting, liquid democracy, and quadratic funding), it becomes clear that "something like blockchain" means some kind of computation/information system, not even necessarily a computer, but maybe just pen and paper (as tedious as that would be). If all you have is a blockchain, everything looks like a DAO.

Here's the last excerpt, after which we'll start to tie things together:

This passage is one of many that completely glosses over what should've been the book's core project: Why, despite all evidence to the contrary, is blockchain useful for the left? That the right captures and redirects people's innate, pro-social, even revolutionary tendencies in ways that ultimately solidifies its power is well-trodden territory, which comes with all sorts of the big heavy concepts with which he was so forthcoming elsewhere, only to be found wanting when he could actually use one. Instead, we get this simplistic analysis about labels and a potentially protosocialist concept decentralization, leading to a prescription that the left ought to get involved with DAOs so that we can reach these people.

Dávila is right that these people belong in our coalition. Almost everyone does. That's the whole point of socialism. He doesn't actually present an argument that DAOs are a valuable tool for the left. Instead, he both says and somehow misses that the political movement behind DAOs is capital, which has found DAOs to be an effective tool to siphon energy from the socialist project. Dávila undertakes no serious effort to understand why, which we will do in part 2, instead blaming the leftist commentariat for its "lack of clear messaging." The left's messaging, however, has been quite clear, and, as we'll soon see, Dávila's own Breadchain Cooperative provides corroborating evidence for its thesis.

This is dishonest, even fraudulent, but I argue that Dávila does not intend to deceive. To understand, we must take a detour to examine the act of software production, something that both he and I do for much of our waking life, and the conditions under which we do it.

We talk about writing software as writing software, but this is an antiquated misnomer, left over from the early days. It would be more accurate to say that modern software is assembled. The vast majority of software is pre-written. If I had to give a conservative estimate, I'd say that, in any software product, 99.9% of the software that it contains is third party "dependencies" and only .1% is custom. This is probably too conservative by at least an order of magnitude.

For the simplest possible example of what that looks like, here is a python script that prints a random integer between 1 and 10:

  from random import random
  my_number = random()
  my_number =  round(my_number * 10)
  print(my_number)

For computers, generating random numbers is hard. We technically refer to them as "pseudo-random," because, in reality, they're algorithmically generated. Thankfully, none of that matters for our script, because someone else already took care of it and made a package called "random," which we can simply import. I just have to know how to use it to do what I want to do, so I went to the docs, and I saw that the very first function they list, called "random," gives you a random number between 0 and 1. To call that function, I import the "random" function from the "random" module, multiply the result by 10, then round it to the nearest integer. That works.

Now consider the structure of Dávila's arguments. As we've seen, they tend to start by importing big ideas, but he doesn't engage with them; he uses them. In just the few examples in this post, we've seen him do this for Deleuzian philosophy, art, housing access, POSIWID, cybernetics, Schelling point, and cooperatives. To use these ideas, he doesn't need to understand them in the same way that I didn't have to understand pseudo-random number generation (or rounding, or even multiplication, if you really think about it). He just picks the piece that matters to him, pulls it out, and uses it to argue what he wants, which is that blockchains are useful for the left. Unfortunately, the attributes of, say, a cooperative can't be individually imported in the same way as the modules of a software library. This way of using, creating, and redefining ideas until they suit your code is common practice in software, as we've discussed before. When you write software, you're the ultimate arbiter of how a term is defined. Consider what it means to "rideshare" today versus pre-Uber, or to "match" with someone now compared to pre-Tinder dating.¹

Similarly, recall the way that Dávila dismisses his critics. When you set out to write code, you have a goal in mind. When I wanted to generate a random number between 1 and 10, had I come across a library that printed random names of fruit, I would have simply disregarded it. Using it wouldn't make any sense. If, after the search, I couldn't find a library that suited my needs, I'd have to write my own. Dávila's research, I suspect, set out to find things to prove that the left should care about blockchain. He was looking for tools, not to understand. He only engaged with what he found insofar as it served that end, and he simply disregarded that which wasn't useful to him. When he couldn't find existing packages to his liking, he wrote his own: "techno-probabilism."

Software is teleological. You have a goal, and the code is a set of steps that arrive at your desired result, which is an outcome of the design process and entirely open-ended. Theorizing, on the other hand, isn't less or more logical, but it's different. It's often dialectical, but it doesn't have to be. Code executes on a computer, which grounds it in math and physics. Conversely, that you can make arguments in human language and have it mean anything at all is miraculous and itself the subject of theorizing since, at least, Plato and Aristotle.

Even the way that he laid out the book, which I complimented at the beginning, fits this pattern. Organizing writing is hard, and a lot of books get it wrong. Organizing huge, complex flows of logic is a core task of software developers, so it's not surprising to me that he did that well. If you look at the structure, and squint only a little, you might argue that it's objected-oriented, like most modern code, and there are three main classes: money, finance, and coordination. The chapters are then subdivided into relatively short sections, much like a class might have methods. This is not that common in long-form writing, but it's how almost all code I write is organized. Sometimes, in these sections, he "imports" a big concept like POSIWID, but then, in other sections, he reasons in very non-POSIWID ways. This suggests a concept of "scope": In code, packages imported into one section live in that section's scope and won't affect another.

In the most general sense, we all do this, but it's easier to notice in others than in ourselves. Thinking, like most things, takes practice, and if you mostly practice one way of thinking, that's how you'll think. The truly devout, for example, can only understand the world through their religion. It's not only a question of belief, but of patterns of thought totally alien to the nonbeliever. They had to learn both how to think that way and how to avoid thinking in other ways. This took practice and training.

Here's Philip Agre, an AI researcher writing about his own thinking before what he calls his "awakening":

Here he is post-awakening, describing the experience of talking to others in his field:

It's easy to see Breadchain Cooperative in that bolded section. Dávila does not engage with the leftist consensus on crypto because he hasn't had this critical awakening. He finds his critics frustrating because he thinks that Breadchain is what leftists want, and he's tired of being scolded by what he sees as his own side. He is, after all, out here doing the work, building the world they claim to want, and all they can do is find reasons to complain. As we'll see in part 2, the reality of Breadchain is that, like in his arguments, the technical system he created is unmoored from any new, big ideas. Instead, he recreated the thing that he thinks he critiques, but he lacks the critical capacity to evaluate it as such.

This, finally, is the message that I want to send my fellow developers through this review, especially those just starting their journey of political awakening: Being radical means thinking radically different. It's not a question of simply discarding old labels and donning new ones. It will be uncomfortable and unfamiliar, but, when you get it, it will be fun, rewarding, almost intoxicating. If you think that finally finding that one bug was satisfying, you're going to lose your goddamn mind when you suddenly understand a text that changes how you think. Here's Agre, one last time, explaining his early experiences with that feeling:

Part 2: Garbage in, garbage out

Breadchain, for Dávila, is "a post-capitalist economy within the confines of present-day capitalism." The technical infrastructure is supposed to act as a "membrane" between it and the outside world. Reasonable minds might disagree on how useful it is to think of, say, cooperatives as islands of socialism within capitalism, but technical infrastructure, on its own, cannot make those islands. It could, potentially, facilitate solidarity and organizing, but, as we'll see, Breadchain does not. To be clear, I am not arguing that making these kinds of ecosystems is necessarily bad, rather that, as Breadchain's theoretical grasp of its stated goals is hopelessly confused, so too is its praxis.

To explain how it works, I'll be drawing mostly from the detailed and well written explanation found in the book, but I quote here from a blog post on their site for length:

Here it is outlined and illustrated:

• People buy BREAD with DAI. BREAD can be converted back to DAI at anytime.
• All the DAI goes into a lending pool, which gets loaned out to people doing whatever they do with crypto stuff. Breadchain says that the DAI isn't donated. This is misleading: Taking on all the risk but passing along the interest is close enough to a donation. If and when this whole thing comes crashing down, the not-donors will be left holding the bag, but the member projects will have gotten all the interest.
• The interest from those loans goes into a shared wallet.
• People who own BREAD vote on how to distribute the interest to the various member projects. From the project's voting page: "Your Voting Power is equal to the average amount of BREAD you held during the 30 days up until the current round of voting opened."
• BREAD, like all tokens, can be bought and sold. Because BREAD can be exchanged for DAI, there exists what's called a "liquidity pool" to stabilize their relative prices on the open market.
• BREAD is intended as a currency that can be used between members. There's a marketplace, though it only lists 5 things.

As far as I can tell, there is no publicly available information on the process by which projects are added or removed, but here is the list of member projects, along with their descriptions, taken directly from Breadchain's site in early January 2025:

• ReFi Dao: A global network driving the Regenerative Finance (ReFi) movement, leveraging Web3 to regenerate ecosystems and foster economic systems that prioritize ecological and social well-being.
• Citizen Wallet: An open-source project providing a Web3 tool stack to support community currencies.
• Breadchain Core: The core team developing the tech and design used by Breadchain.
• Crypto Commons Association: An organization producing research and events related to decentralized technologies and the commons.
• Symbiota Coop: A collective of event-focused organisations devoted to new forms of culture and enquiry, structured as a not-for-profit worker co-operative.
• Labor DAO: A collaborative community of researchers, developers, and visionaries dedicated to building innovative tools and solutions that enhance the web3 ecosystem improving experiences for workers and organizations.
• Breadchain Shared Treasury: A co-owned treasury in Breadchain used for grants and sponsorships.

DAI is a so-called "stablecoin," meaning that its value is, as their whitepaper explains, "soft-pegged" to USD. They accomplish this with complex financial instruments involving over-collateralized loans and blockchain smart contracts, the details of which are discussed in the whitepaper. In the book, Dávila says that DAI is "pegged" to USD, but I prefer DAI's own language of "soft-pegged." This was probably just intended as a harmless simplification, but the nature of the relationship is qualitatively different: BREAD is pegged to DAI mainly because you exchange DAI to get BREAD, and you can always cash it back out. DAI is pegged to USD through an elaborate scheme of financial machinations, not a reserve of USD that DAI owners can claim. We have not begun to discuss Breadchain Cooperative itself, but the foundation upon which the entirety of the project rests, since the project depends on the value of DAI twice over: For Breadchain to work, DAI must both maintain its value and be desirable enough to loan out with interest.

Atop the foundation, then, is Breadchain's structure, which is not unlike how many kinds of financial endowments, foundations, impact investment funds, and grant-making organizations work: There are some investments, and a governance structure decides how to spend the proceeds to further the organization's goals. Crucially, in Breadchain's governance structure, buying more BREAD gives you more votes in how the proceeds are spent. This is exactly how shareholders vote in actually existing capitalism, even though it's on this side of the membrane. As we've already discussed, this is by definition not a cooperative, but is substantively similar to many kinds of non-cooperative philanthropic organizations. It should be called "Breadchain Foundation."

Now, we must ask the obvious question: If Breadchain is structurally similar to many existing organizations, why even use blockchain? From the book:

The revolution will not be funded by the current rulers of the economy, but if we are going to try to live with the contradictions inherent to building the new in the shell of the old, funding during this awkward stage will certainly not come from the state-regulated banking sector. Only by creating bonds of solidarity and networks of mutual aid for cultivating dual power will we reach true economic democracy and freedom, and doing that will likely include engaging with some parts of shadow banking in a strategic fashion. BREAD is not a governance token like a stock in a company or some other speculative financial product [emphasis added], and it is not just a stablecoin. It is a form of economic expression that shows support for a vision of post-capitalism.

First, that bolded section of the explanation is misleading, if not entirely indefensible: BREAD buys you control over how proceeds are allocated, can be sold on a secondary market, and its value depends on the stability of complex financial machinations. It's not exactly the same, but it's damn close. The main difference, this "form of economic expression," is vibes, or, to put it in marketing terms, it's a branding exercise. Recall that, for Dávila, "well-meaning people" wanting Breadchain to be a cooperative made it so. Here we see a similar, vibes-based theory of revolutionary praxis. The supposed radical nature of the project lives in exactly 2 places: the marketing copy, and the hearts of the individuals. In the actual structure of the project, it is notably absent.

Setting that aside for now, Dávila also argues that the blockchain's alegality is its tactial advantage. In reality, the blockchain offers little to no meaningful protection, but comes at great cost. There is, of course, the aforementioned risk inherent to tying the work of generations (as the socialist project is) to crypto shenanigans. More relevant to his argument, if BREAD ever truly had revolutionary potential, the state would try to shut it down, one way or another. Blockchain's alegality is a fiction born of its current alignment with power.

This is true for all activist organizations, but BREAD is particularly exposed to reactionary forces for at least 2 reasons: First, actual revolutionary organizations work to build power, but what is Breadchain's equivalent of a strike? How do BREAD-holders extract concessions from power? BREAD has no leverage, no way to make demands or negotiate. Second, all coordination on blockchains is, by its nature, exposed to capital. The book somehow argues that this is an advantage:

As an aside, revolutionary organizations are not known for their transparent operations. Leaving that and addressing his point directly, he argues that liberal democracies allow capital to have an "unofficial" channel for governance, but the solution cannot possibly be to turn that unofficial channel into an official one, transparent or otherwise. Because Breadchain's governance is on the blockchain, there exists an amount of money that would be sufficient to collapse the value of BREAD, gain control of voting, or otherwise disrupt its operations with clever financial maneuvers. That Breadchain uses a liquidity pool to protect BREAD from speculation acknowledges this risk, but anyone who can outspend the liquidity pool could disrupt BREAD circulation. They're rookies playing on capitalism's turf. Even if it weren't ideologically confused, it is tactically inexplicable.

Because Breadchain redwashes what are mundane capitalist modes of governance and operation, it has a parasitic relationship to socialist organizing. It is functionally identical to the DAOs backed by venture capital that Dávila criticized in that excerpt in part 1. Breadchain convinces people looking to get involved in radical politics to give them a free loan, buy things using BREAD within the BREAD ecosystem, and participate in a capitalist governance model, then calls it radical praxis. Political education has always been at the heart of socialist organizing, but Breadchain's marketing copy is political disinformation.

Solidarity is something you do, not something you get. "Buying" BREAD, despite their marketing copy, doesn't bring you into solidarity with others who own BREAD, but just gives you the right to vote in fund allocations, and lets you buy things from people that you think are cool.

Breadchain could probably be improved at the edges, but these problems are innate to organizing on the blockchain. Its most important technical achievement, as Dávila explains, is the combination of its "permissionless nature" with Turing completeness. This means that you can write code and have it executed on the blockchain, as you would on any computer, except that the actual computation is being done on a bunch of computers that you don't trust, which also don't trust each other.² I fully agree that this is an impressive feat, especially when you consider how much effort developers expend making sure that the right code, and only the right code, executes in the right places.

Because of this, every use of blockchain applications in socialist organizing has a non-blockchain alternative with more desirable institutional dynamics. We've already seen an example of this in the discussion of legibility and governance structure, but the problem is general. Blockchains can do any kind of computing without trust, but the core of socialist organizing is building trust between members. Organizing on and through the blockchain mediates the resulting social relations through inflexible, anonymous/pseudonymous technical systems. Blockchains can coordinate, but they can't build associations, meaning that you can, for example, vote on the blockchain, but you can't deliberate on it. The former is often the procedural aspect of democracy, but the latter is its substance. Dávila, then, has it exactly backwards: By building a blockchain system and recruiting people to it, he creates an organization whose relations are, by default, mediated through the blockchain. This is an atomizing process, not a unifying one. BREAD holders are consumers of commodified pseudo-solidarity engaged in perfectly ordinary capitalist activities, and there is nothing radical about it.

Towards the very beginning of his book, Dávila gives one use case for the blockchain that I find sound, and the comparison is enlightening:

It is exactly through this mechanism ["shadow banking," i.e. blockchain] that Wikileaks, Sci-Hub, sex workers in the US and those living in sanctioned countries can have basic financial rights...

Pre-existing organizations trying to circumvent government repression can and should use all means available to them, including crypto. To understand why that make sense, note the order of operations: Organizing comes first. If crypto were to disappear tomorrow, those organizations would still be around to figure it out. They are in this position because they have already survived attempts to shut down their finances. Their relationships are what makes them work, not their finances or their tech. Dávila argues that blockchain's rigid legibility is its advantage, but the illegible, flexible human relationships are the true power.

In part 1, I argued that Dávila's analysis is unsuited to a project of revolutionary theory, and here we see that come to fruition. I have written before about how human relationships, by their nature, are illegible to computers. As the tech industry increasingly mediates more and more human relationships, it also degrades them, changing them to suit the needs of computers. This is an alienating process, not a liberating one. Breadchain is more like Tinder than like a labor union or a socialist party. Like Tinder, it reproduces itself by capitalizing on the innate human desire to connect, in this case literally, but the resulting connections are simulacra, not true bonds of solidarity. Dávila argues that blockchain's affordances allow for new kinds of coordination at scales never before possible. This is true in the way that Tinder offers new kinds of dating, or in the same way that, as I've previously argued, hashtag activism and the associated spontaneous mass protests reproduce the undesirable institutional dynamics of social media.

Breadchain dooms its postcapitalist project by allowing its chosen technical infrastructure to dictate its organizational structure. It's a supposedly revolutionary organization, structured in the form most susceptible to capitalist intervention, capitalized by totally-not-donations from would-be activists, which redwashes perfectly ordinary capitalism. It offers solidarity as a commodity, an obvious absurdity that degrades the connections among would-be activists, and it is structured to leave its totally-not-donors holding the bag should the whole thing collapse. Despite how well meaning the people who made it might be, it is, structurally, a grift, to the surprise of absolutely no one who knows anything about crypto.