- O ’ Reilly, “ Profile of Tim O’Reilly. ”
- Evgeny Morozov, “ The Meme Hustler, ” Baffler 66 (2013), http://www.thebaffler.com/ articles/the-meme-hustler.
- Among the core design problems for algorithmic governance is the modulation of that symmetry and asymmetry: between “ server side ” and “ client side, ” authority and autonomy, between upstream and downstream decisions, between Users informing algorithms and algorithms governing with the force of law, or perhaps just “ nudging ” Users toward ostensibly desirable outcomes. That is, where it may be news to conventional political science that planetary-scale computation is not only something to govern but where and how governance occurs, it may be equally surprising to others that software is not only how to govern but that it itself needs to be governed, and that the cultural conventions of early twenty-first century North American programming culture cannot be transposed onto all systems without difficult mutual translation and even violent (if necessary) counterdistortions. (As “ design outcomes ” algorithmic governance and governance of algorithms should be tightly paired.) Historical precedent strongly suggests that asymmetries of knowledge and power between governor or and governed (in either direction) may have disastrous consequences. Contemporary concerns over User privacy in relation to state surveillance and platform dataveillance have led to considerable outcry, some very legiti mate and some not. The contravention of surveillance (looking “ down ” onto the governed) with sousveillance (looking “ up ” at the governor) has become the conventional point of distinction for modulating transparency versus opacity, and, by extension, designing the geopolitics of the User. The problem in practice (as well as in theory) is that this distinction relies on notions of who or what is “ up ” and “ down ” in relation to whom and to what that are at best highly debatable. This is so even among the privacy advocacy communities who have come to agree on “ outcomes ” even if the reasons for these are incompatible. For some, large central governing systems must always by definition operate in complete transparency lest they become totalitarian and end Users must always by definition retain rights to complete opacity for the same reason. For others, a phobia regarding public scrutiny is hardly the point, and there is strong motivation to shine the bright light on some “ Users” such as wealthy tax evaders for the benefit of large public governance. Others advocate for maximum societal transparency as a self-correcting good, such as David Brin, while others advocate for maximum opacity and privacy for parallel or opposite goals, such as some cyptoanarchists. One pernicious result is that everyone believes himself to be working from a position of disadvantage and self-defense, and so assumes for himself a limitless license for social violence.
This up-versus-down heuristic will prove to be a far limited model for scaling the complex geopolitical design questions at hand now and in the future. It is born of a presumption that governance is synonymous with states that contain human being citizens, that governments are centralized and distant, but “ re-present ” the will of those citizens, that any space not occupied by government sovereignty is instead the domain of “ the market, ” which has diametrically opposed priorities for the essential (and inaccurate) vision of center versus periphery that underlies parlia mentarian and bureaucratic conceptions of how and where political authority operates, and so which can only vaguely discern how platforms may constitute a very different kind of political geography. As discussed in the City layer chapter, Cloud platforms push what used to be command-and-control infrastructural capabilities to the edges of the network, imbuing humble objects and interfaces with supercomputing capacity, with the Cloud interface device called the “ mobile phone ” being the plainest example, and in doing so, the Cloud absorbs and centralizes other platform regulation duties. In doing so, the object of governance is not just humans but things as well, including and especially the interfaces between humans and things. Who or what is or is not transparent in the first place is not the same for people and objects, and yet both can be Users of one another. As we will see, the geopolitics of the Cloud and composition of the User as a political subject may bring a kind of trauma, as the emergence of an algorithmic governance that is not primarily a governance of humans but also of ecologies, inanimate objects, data sets, other algorithms, of sensing and sensation. This is complicated by the fact that we are not designing one algorithmic-interfacial regime but multiple simultaneous and interweaving totalities. “ Desired outcomes ” may overlap but vary widely. What is legal and consistent for one may be technically or morally impossible for another. That heterogeneity may be another bulwark against totalitarianism, and it may also be a path toward another as yet undefined universal materialism, one that is no more or less totalitarian than the laws of mathematics.
Notes
453