Tag Archives: protocols

platforms make markets

Rob Horning again, on the hustle economy the gig economy the ubiquity of platforms as obligatory labour intermediaries:

The “hustle” platform seems like a mere means of distribution for the “creators” or “entrepreneurs” who own the means of production (their own bodies). But in fact the “hustle economy” scenario is not so different from working in a factory. The creators have labor power, which they effectively sell to the platforms, which control the means of producing the “things” that matter here: reaching potential customers at scale, sustaining a means of limiting access to goods, maintaining the business infrastructure that facilitates exchange, aggregating the supply of “creators” to create a sort of network effect that subordinates the value of individual creators to that of the entire catalog that a platform owns. Maybe the simpler way of saying that is platforms make markets. That is their product. They serve neoliberalism’s tendency to marketize any aspect of existence that has hitherto resisted it in the name of bringing the market’s “discipline” to wayward aspects of sociability.  In submitting to a platform, whatever “existing skills” a creator has are reformatted, warped to fit the scheme devised to exploit them for profit. I think Marx calls that “formal subsumption,” when a capitalist organization of production appropriates pre-existing skills and reshapes them so as to primarily valorize capital. Then, when the existence of Patreon, etc., is taken for granted, the new skills that can be acquired take the shape demanded or anticipated by those platforms, on their terms and to their benefit and for their continued survival (“real subsumption”). The platforms begin to produce the sorts of creators they need to profit.

It’s not at all different to working in a factory—in fact, it’s basically piecework, but piecework done for factories that have given up any pretense of producing a specific product or type of product to the end of capital accumulation; they’ll flog whatever the pieceworkers are able to produce. Or rather, they’ll provide a system whereby the pieceworker can flog that thing—and flog themselves—and cream off whatever the market will bear.

Work itself, when organized through platforms, is no longer “skilled,” but just abstract spare capacity that capital can take or leave when it needs it, dipping into the pool of reserve labor on its own terms. Patreon et al. bring that model to forms of work that otherwise seem resistant — the kinds of things contingent on one’s reputation or personal brand. “Anyone with noncommoditized skills can do this,” venture capitalist Li Jin tells Dewey — anyone can sell themselves on/to a platform. It’s no accident that VC types champion these models with deceptive rhetoric about inclusion and empowerment; the models are designed to line their pockets while impoverishing the world, and they work better when every possible person is exploited. That’s the beauty of scale.

Oooh, look: it’s our friend scale again! Which is, I think, the fundamental feature of platforms as a business model… and that’s related to the natural-monopoly tendency of infrastructures of distribution.

In short, all platforms are labor platforms. They turn “life” into “work” wherever possible, for “creators” and consumers alike, tracking and redistributing user behavior on the platforms’ terms. Commodification makes inroads in more and more aspects of everyday life and comes to dictate more and more of social relations. Why make friends when you can have followers? Why speak from outside a company’s paywall? The platforms are not offering a way to survive in a crisis; they are exploiting a crisis to introduce a regime that makes immiseration a permanent condition. The tides rise but never recede.

To return to a point I keep drifting away from (because easily distracted), the platform model is not the inevitable outcome of the natural-monopoly tendency in distributive systems—or at least it doesn’t have to be. (Capitalism’s incentive structures make it effectively inevitable, as Horning is pointing out above: “platforms make markets”.) However, the protocol model could achieve the same opportunities for the pieceworker, but without the proprietary capture and rake-off associated with a successfully scaled monopoly platform; the platform model in this case would look like a suite of open-source tools for website building and online payments, a fairly-priced postal service, and a search function that wasn’t massively biased in favour of platforms already within the search engine operator’s own value chains (for both uses of the term “value”). Protocols also make markets—but they make markets with freedoms-from as well as freedoms-to.

With the hopefully assumed caveat that it’s far from being a silver-bullet solution to this or any other socioeconomic challenge, we could really do with remembering what the rail barons were, and why so many states eventually decided to nationalise those early systems of distribution (which, you may recall, also included telegraphy, the forerunner of the communications infrastructures which we still erroneously refer to as “the internet”).

The natural-monopoly tendency of distribution networks can be a problem, but it’s also the source of their usefulness—their utility, you might say. Nationalisation, or some other form of common ownership, gives you the horizontality of the platform without the verticality of the associated value extraction chain; open protocols provide a level playing field that still permits for innovation at the producer level with regard to products and services, and for middle-men operators who have genuine value to add to the process.

But: you have to regulate the system in such a way that the middle-men can’t make themselves into an obligatory passage point—which is to say, you have to keep the infrastructure layer free from commercial interference, at minimum. Well-designed and constantly revised standards and regulations at that level can, in theory, make sure that the interface layer doesn’t ossify into a platform model.

(For an example of this sort of set-up, think of the road network, at least in the UK. It’s not without its flaws, certainly; nor has it avoided becoming a site of ideological economic interference. But in terms of being a transportation protocol which is effectively open to all users, provided they stay within the standards and regulations, it’s so exemplary as to be almost invisible. Admittedly, the neoliberal era has seen the assembly of de facto platform business models upon the road network, but that was achieved more through stock-market malarkey than actual competition. Effective and reliable distribution requires detailed knowledge of the last mile as well as the system as a whole; a national-scale firm can’t really do that, unless making a good job of the last mile is an imperative of its reason for existing. And because the last mile is the expensive bit, as well as the difficult bit, no national-scale distribution firm oriented toward shareholder returns will ever do it well, because it’s the easiest place to shave off your overheads. Effective and reliable networks of distribution, when seen as entire systems, are almost inevitably loss-making; if there is any point at all to the existence of the state, then providing that fundamental infrastructure for economic and informational exchange, and underwriting those losses as the unavoidable cost of having an economy, is surely it. I might even go so far as to say that the state is its infrastructure; all the other stuff is just platforms and protocols running on that network.)

We learned all this once before, in the era of the rail barons, but we allowed ourselves to forget it—because, as a friend put it earlier today, history never actually ended, per Fukuyama, but everyone nonetheless stopped paying it any heed. But it feels like we’re starting to remember it again, collectively, and to find ways of narrating this systemic causality that might help build the grass-roots political will for busting up the platform barons of the Valley, and the ideological scaffolding that they stand upon. The struggle will be long, and victory is far from assured—but nothing worth doing is ever easy, right?

protocols > platforms

I used to read Techdirt, and Mike Masnick in particular, with genuine reverence when Futurismic was still a running concern. He was tech-critical long before it was fashionable to be so, but from a position that challenged my own thinking quite a bit, and still does: in short, Masnick’s about as close to First Amendment Fundamentalism as I’ll go. (I used to go a fair bit closer than I do these days; insert your own wistful reminiscences of more innocent times on the still-somewhat-utopian internet of the Noughties here.)

These days my greatest and enduring interest in Masnick’s work comes from his maintaining the old (and seemingly all-but-forgotten) distinction between protocols and platforms. As presumably foreshadowed by the preceding paragraph, this is largely in service of “free speech” — though Masnick’s notion of free speech is notably more thought-through and pre-problematised than the one you encounter most often. I guess we might say he’s one of the last intellectually honest techno-libertarians.

Here’s Masnick introducing a long piece at Columbia U’s Knight First Amendment Institute, where he argues in favour of a return to an internet based on open protocols rather than proprietary platforms:

Moving to a world where protocols and not proprietary platforms dominate would solve many issues currently facing the internet today. Rather than relying on a few giant platforms to police speech online, there could be widespread competition, in which anyone could design their own interfaces, filters, and additional services, allowing whichever ones work best to succeed, without having to resort to outright censorship for certain voices. It would allow end users to determine their own tolerances for different types of speech but make it much easier for most people to avoid the most problematic speech, without silencing anyone entirely or having the platforms themselves make the decisions about who is allowed to speak.

In short, it would push the power and decision making out to the ends of the network, rather than keeping it centralized among a small group of very powerful companies.

Now, regular readers will likely have picked up some hints in recent years that I no longer assume competition between commercial actors will necessarily, or even possibly, result in the emergence of an optimal solution. (YA RLY.)

That said, I think there’s an argument to be made that a lot of the sociopolitical issues we’ve seen in the last decade have been exacerbated by the ubiquitous reach of certain proprietary platforms of communication. A very reductive way of putting it might be to observe that if a platform and the set of rules by which it is used (which we might call its interface) are considered as a game, then two things are likely to happen: a) highly motivated people will find a way to win the game by pushing at the edges of the rule-set, and b) highly motivated people with access to money will lean on the curators of the rule-set to make their pushing at its edges more successful. All of this will happen within a broader ecology where almost all the strong incentives to action tend to be driven by and towards the accumulation of profit and/or power.

(Or, more succinctly: Facebook, and its abuse for political and financial influence, is inevitable in a capitalist context which does not actively mitigate against the existence of Facebooks.)

But Masnick’s spot on when it comes to the power of protocols — because both he and my anarchist self recognise that replacing Facebook with a state-run monopoly platform, however well intended, would result in similarly dysfunctional results. (This is another sense in which the big comms/tech companies are infrastructures, and thus very similar to the rapacious railway companies from which they are descended.)

I don’t have the time (and you likely don’t have the patience) to revisit some of the more woolly implications of my PhD thesis today, and in the end the protocol/platform distinction didn’t make as strong a showing in the final work as originally expected — but nonetheless it’s a distinction that matters a great deal in my theoretical model of sociotechnical change. To simplify hugely, the reason why a state-run monopoly platform is little better than a privately-owned monopoly platform is not that they are both monopolies; on the contrary, any geographically inelastic distribution system (be it a railway or a communications network) should, nay MUST be a monopoly, because breaking the network up into subnetworks reduces both its functional and its economical efficiency.

Rather, the problem with platforms is that, to use my preferred theoretical nomenclature, they control both the infrastructure layer and the interface layer. A protocol, by contrast, is provided as an open opportunity (or capacity) by an infrastructure layer in order that all comers are able to to develop their own compatible interfaces thereto; those interfaces will work with the clearly delimited capacities and potentialities of the infrastructure layer, and package them in such a way as to fulfill the teleology of a particular practice-as-performance.

(Regrettably, the full elucidation of my theoretical work has yet to make it into any publications, and I’ve not had the time to write it up for its own sake. I’m hoping that an SI proposed in the aftermath of last year’s RGS conference, plus a paper I’m hoping will be accepted for this year’s Petrocultures, will give me the chance/reason to get this stuff down in print and out in the world.)

Looping back to Masnick, then: my theory broadly agrees with him on the point about “pushing power out to the ends of the network” — but with the proviso that the network (which cannot be disentangled from the organisation charged with running and maintaining it) must necessarily be a closely regulated and functionally restricted monopoly in order for his proposed freedom of use-cases to be possible; the total organisational separation between the infrastructure layer and the interface layer must be maintained. This is not an ideological position, but an argument from function which can be illustrated with pretty much every infrastructural development in history.

And that’s the core thesis one of the handful of books I’d really like someone to pay me to write… but as no one’s gonna even think about paying me until it’s been written, I guess I’d better find the time to write it while I’m being paid to do other things, eh?