Some Political Reflections
Here’s a simple way of crystallizing the politics of “anthropomorphics”: there is a vertical axis and a horizontal axis; on the vertical axis, we start with the minimal model of an issuer of imperatives that include the presumption of obedience (as opposed to imperatives—like those of the petitioner—that include no such presumption) and what one wants is that those imperatives be fulfilled, which means they are “well formed” and situationally “actionable,” are transmitted in a way included in the command, and received by those taking the intention implicit in the command as their own. Anything that gets in the way of this process of issuance, transmission and reception is what one is against. Because what’s the point of putting people in charge of things if they’re not in charge of things? The horizontal axis involves the acquisition, collection, assessment, preservation, filtering and communication—ultimately to those who need it and as they need it on the command chain—of data, information, intelligence; here, we want data, information, intelligence produced in the most directly and comprehensible usable form: we want data, information, intelligence such that if you were to assign teams of auditors to check and verify it all the only defects they would find would be due to the time constraints of having the material available when needed (if the decision could have been put off until tomorrow, the intelligence could have been better, but the best available evidence suggested that the decision could not be put off and the checking and verifying could not offer a better reading of the intelligence then available). Anything and anyone mucking around with this process should be opposed and subsequently suspected with the greatest possible prejudice (but this is not necessarily an obvious determination insofar as, to take one example, public falsehoods may be disseminated as part of a broader intelligence operation—part of knowing is knowing who knows).
If we want to then get into institutional analysis (monarchy vs. republic vs. democracy, etc.) it will be in accord with the question of which arrangement will best enable the vertical to intersect with the horizontal. My own way of formulating this intersecting is through the concept of “singularized succession in perpetuity,” because the most direct proof of a operational command structure is being able to select one’s successor, and the most direct proof of an operational intelligence structure is that such selection is most likely to best serve social continuity, making the present a space where the past flows intelligibly into the future. This is a politics that is as radical as one can get, I think, in its implications for transformation, while still observing the principle “do no harm,” since any but the most completely broken institution will have some operational command and intelligence structure, or even just the institutional memory of one. It’s also the most visionary as well as most practical politics, since it would inflame the imagination of anyone thinking of the kinds of people and relationships needed to create such an order while informing what anyone can do right now, at any level or scale. And think about all the pointless arguments over definitions you can avoid by just asking questions like, “ok, so what can this guy that you want to do things actually do”? We can conduct ongoing, unprejudiced audits of the existing order that question the criteria for assessing facts along with the facts themselves—maybe, for example “90%” of the social order can be deemed “healthy” but even so the corrosive effect of that “10%” is yet to be determined.
Let’s now bring this pared down version of an “anthropomorphic” or “originary” politics to bear on what I hope is an equally pared down version of the primary political transformation we have been undergoing in the transition (to first of all speak colloquially and imprecisely) from modernity to postmodernity. The problem posed for modern governance is the deferral of violence under postsacral conditions where the ordering offered by the honor system must be placed, at the very least, under perpetual pressure. This was accomplished through the assignment of rights to the individual as a simulacrum of the sacrality of each soul before God as the correlate of the evacuation of the sacred governing center: one can rule insofar as one’s rule is acknowledged to be provisional and only in the service of protecting the rights of the individual. This mode of governance has generated an entire moral, ethical and aesthetic order of the elaboration of the individual, his/her/their outer and inner components, dilemmas, comedies and tragedies, and so on. It reaches its breaking point as governance is transferred to the bureaucracy as a result of the evacuation of the center, which is still preserved and even enhanced for stage management purposes, along with the increasing distribution of resources to meeting the needs of all of the individuals (at least sufficiently so as to ward off catastrophe) in the social order so that considering individuals qua individuals becomes impossible. Individuals can only be represented as victims of some usurpation of the center, or as brave opponents of some such usurpation—any claim to centrality, or to be intrinsically centered, is highly suspicious, to say the least. And all of this corresponds less and less to the actual workings of governance.
What governance, with the aid of social media and increasingly sophisticated forms of automation, now aims at, is a probabilistic approach, which involves the generation and ranking of probabilities through the gathering and analysis of data. When I say that “governance aims at” this, I mean that anyone given genuine responsibility over large scale human activities will have to resort to such means or will very much want to. We see enormous resistance to the emergence of this mode of governance, and while the most sensational form of resistance is from right-wing opposition to globalism, the most effective resistance is in fact from the left, which is a kind of internal resistance deforming the process (and thereby producing results especially objectionable to conservatives, traditionalists and nationalists). The probabilistic approach to governance (algorithmic governance) poses enormous, maybe even existential challenges to the juridical form, which is the source of and is reinforced by the institutionalization of the sacrality of the individual. One vector of science fiction has zeroed in on this challenge (e.g., “Minority Report”), and I have mentioned it in previous posts as well: if we can determine through, say, some socio-biological science which people are most likely to act anti-socially and disruptively, who will under-perform and who will perform well, why not act directly on that knowledge to steer and constrain groups and individuals before waiting for them to do the kinds of things we already know they will do? The only thing standing in the way here is our juridically derived commitment to judging individuals in accord with their own actions rather than in accord with some stereotype they fit.
The left is closely aligned with and even embedded in the emergent institutions of algorithmic governance so its completely rational concerns here can never be made sufficiently explicit to be discussed openly. If we allow for a strictly probabilistic model, the victim groups defended by the left will be disproportionately disadvantaged. The left tries to get around this by attributing such results (which we’ve already seen enough of to elicit by now stereotyped responses) to something like “racist capitalism,” but they don’t really believe that a more fine-tuned and nuanced programming of the algorithms will produce more favorable results. At the same time, there is much in algorithmic governance that they see much use for, especially in the surveillance field. They need to work out the right mode of control but getting control of the “bad” results might interfere with at least some of the “good” results for some major players. But the left is correct to assume that lowering the threshold of the juridical will lead, not to a more properly scientific creation of social order, but to openings for various returns of the honor system, i.e., the vendetta, and this from both right and left. And, in fact, we are not able to imagine the abolition of the juridical and a full move into disciplinary governance, and we may not ever be—to imagine the uprooting of juridical terms (justice, freedom, equality, rights, impartiality, etc.) is to image the creation of a new language, which can’t be done within the language we have.
It remains the case that even if there’s a 10x higher likelihood that members of group A will commit a crime than that members of group B will do so we don’t know whether this member of group A will commit this crime here and now—and this is still a “higher” insight than any we could glean from the most advanced algorithm, simply because removing this assumption narrows and deforms the data we will receive for reworking the algorithm. But the juridical can be not only preserved but enhanced by acknowledging the implications of probabilistic knowledge for auxiliary preparations meant to remain in the background as much as possible until necessary. When encountering a member of group A, alarm bells, at a certain volume, can be set off and tactical reserves carefully and unobtrusively mobilized. And if this particular member of group A does nothing to increase the volume of the alarm bells or initiate activation of those tactical reserves, he/she/they need ever feel (even if everyone will know) that they were set off. And, of course, those tactical measures can be made as transient in their effects as possible (pre-empting rather than punishing crimes is preferable). These are all things those responsible for public safety could work on, that would in the long run be more transparent, more conducive to public trust and solidarity, and evidence of a mature social order. And what counts as appropriate measures could itself be assessed on juridical as well as disciplinary terms. We could imagine behaviors (and therefore algorithms) changing over time as the more productive members of the various group would be influencing the more anti-social or low performing members, whereas the reverse tends to be the case now. Working towards such an arrangement would be an excellent example of the mode of politics described above, which is therefore well-suited toward ushering us in to the age of algorithmic governance—but I don’t think I’d have to convince anyone that this would not be an easy politics. The left currently stalls such a politics with a hyper-juridicalization that draws upon the early results of algorithmic governance to radicalize claims of discriminatory institutional structures and thereby using the discipline to launder power in an acceleration of turnover of occupancy of the center conjoined with the intensification of centralized means of control. The various strains of the anti-globalist right, meanwhile, will have to determine whether they can see a way clear to an embrace of algorithmic governance, especially, in the wake of their repudiation of the COVID regime, in the area of bio-medical politics. Must all bio-medical politics be “tyranny”? More effective and less intrusive modes of provision of public health must be imaginable, and in fact one often sees explorations in that direction by anti-globalists—it will be a question of, as I attempt above, reconciling such measures with some reconstruction of juridical categories, rather than a reactive use of existing juridical categories to bolster opposition.
The greatest challenge, though, will probably be engaging surveillance itself—the kind of “casual” surveillance that results from the fact that every means of communicating, transporting and transacting we have today leaves digital traces that could be collected and used by corporations and governments. The current approach, one that with no exceptions I can think of is shared by left and right alike, is to deploy existing juridical categories to denounce and seek to minimize and control this pervasive surveillance. Indeed, much of the existing surveillance exceeds existing legal (and moral) provisions, which were not designed with the existing technological means in mind. All our assumptions about “privacy,” for example, are outmoded and reactionary, modeled on mid-20th century urban anonymity. But if we mistrust our rulers so much that we don’t want them to have any information about any of us other than what could be obtained through a properly secured search warrant how could we imagine that any rules we could somehow force upon them would in the end be observed? Indeed, who other than these same rulers would be enforcing those rules? Making the social more intelligent implies more intelligence. Here in particular is where a more open and even adventurous approach to the juridical is necessary, where energy could be focused on enhancing individual responsibility for actions on the model of what was once routine: suing individual law enforcement agents for unjustified searches and arrests. This practice, in fact, was part of the objection to the exclusionary rule derived from the Miranda case—if the officer stumbles, why does that mean the perpetrator should walk? Address the officer’s impropriety separately—if the evidence, even if improperly collected, does point to the guilt of the accused, it should be used, even if the officer is liable. Instead of a polarity of tyrannical government vs. innocent individual, we’d be involved in sorting out cases where the government is engaged in fulfilling its responsibilities employing the means it sees fit in a given situation from cases of incompetence or persecution, where information is misused. This would have a de-bureaucratizing effect, and relativize distinctions between official and non-official modes of enforcement while also accentuating the paradoxically self-abolishing nature of the juridical: after all, if every criminal were arrested, convicted and jailed according to the legally determined penalties there would soon be no more crime; if every transaction was entered into and conducted with good faith there would be no lawsuits, etc., and this is exactly the result the law is designed to encourage—so, in a sense we can not so much imagine as get a glimpse of the post-juridical by re-imagining the juridical. Just as developments in AI make the element of human control and crucial nodes even more important, the pressure placed on the juridical by the disciplinary in algorithmic governance can lead to re-inventions of the juridical and even its expanded use—and then, new versions, moral, ethical and aesthetic, of the “individual,” will emerge in its wake.