In the meantime, the design dilemmas of algorithmic governance are undecided but of pressing importance.
The infrastructural determinism of platform economics shifts where politics might be located and what its compositional limits might be in relation to markets, agora, and polities is solved differently by different systems. The economic architectures of China, the United Arab Emirates, Wall Street, and Silicon Valley (for example) all represent different genres of capitalism, differently expressed by similar computational substrates, enrolling or refusing available layers of The Stack. More important, components of those architectures are as available without prejudice to geodesign as much as the legacies of red futurism. We are interested in systems that can adjudicate the “ socialist pricing problem ” (knowing that centralized systems can be too slow to sense and calculate individual price signals) and as well as the capitalist pricing problem: that market models can confuse the emergent effects of transaction liquidity with systems planning and do so at the expense of artificially segregating and suppressing the real costs of near- and long-term “ externalities. ” We allow, to the pronounced consternation of both socialist and capitalist realists, that some polypaternal supercomputational descendant of Google Gosplan might provide a mechanism of projection, response, optimization, automation, not to mention valuation and accounting, beholden neither to market idiocracy nor dim bureaucratic inertia, but to the appetite and expression of a curated algorithmic phyla and its motivated Users. Perhaps to an extent, it already does. As suggested, such a platform-of-platforms might be idealized as pure governing medium, creating almost no original content in and of itself, because its fundamental role would be to interface, sort, and optimize the logistics of physical and virtual value, moving in and out. Its politics are fought over the naming, archiving, optimizing, disseminating, and recapitalizing what has been or will be produced by User polities that can possibly be Addressed. Its vision, “ seeing like a state, ” would be defined by the particular constellation of addressable objects and processes that constitute its world and by how it gives interfacial narrative to these through binding metadata about its uniquely cohered archive(s). Now, consider not just one of these universal apparatuses at work, but several, each interlocking, opposing, or ignoring the others, each social totality of planetary-scale computation enrolling the whole globe at once, along with the others. Comparative planetology, we now realize, does not necessarily require going to Mars. <sup>59</sup>
The calculation problem for centralized plans or decentralized markets or any combination of these depends on the availability of data, which itself depends not only on the worldly divulgence and figuration of information from signal to sensor, but on how any platform apparatus is constructed to construe the world as an Addressable surface in the first place. A old growth redwood falling in a forest is not registered as a governable signal if the sensing apparatus is ignorant of its existence or relevance. It is not data until it is made so. Sociology has been aware of the potential and perils of diagramming the world by immanent statistical self-transparency going back at least to
The Stack to Come
333