lukas likavcan, the governance of habitability — review
in this short interview, likavcan and his question-posers go a long way to try to unpack what lukas actual ideas are and where they are categorically situated, and by the time it picks up, it suddenly ends.
its unclear whether lukas has a hard time deciding what to talk about first, or whether the questions are just slowly edging us into a flat plane, or whether the perceived density of the material requires enough time to extrapolate what it is before we can even get started to the point where there isnt any more space for anything else, but lukas and his associates here brush up against ideas without necessarily forming any, philosophers are picked up and dropped, surfaces are touched, but there arent any philosophical moments of great rhetorcial ability, the long passages act more as naturalistic observations and media research analysis than one single large sweep across. this in itself isnt so bad, because philosophy is already dense and complicated, and it isnt so easy having to even decide where to go half the time, especially when you're in the domain of compromise and introduction.
distinctions dont really help us get anywhere if all we get is more of them without the polemical strenght to carry them across, however, it isnt necessarily likavcans fault, as it seems the intervieews themselves are sometimes getting accustomed to his own ideas during the interview rather than posing challenges to it. it also has the officializing rhetoric of an interview, meaning its a socratic dialogue where only socrates gets to speak in any worthwhile level of volume, which is great for, say, american politics, but maybe not as convincing for theory. surely though this is also calculated pre-emptively, serving as more of an elementary introduction than anything else. it by no means is bad material, it just feels improperly chewed.
that being said, the actual categorical and functional concepts lukas is working in and around seem entirely promising and act as a mediator between the post-latourian and brattonian era, lukas seems to know more about this than almost anyone else, and hes even ready to soft-fight bratton for it, which means if anything happens to that project, he'll be right behind to help either backstab or propel it. hes excellent at pushing us upwards, towards more extrapolation, more divides, more elaboration, he seems to be careful where it matters, and has some big ideas behind his name, hes left home with a big load, a whole horse-carriage of stuff, not just one tiny little stick with a plastic bag wrapped around it, but a whole reportoire of important conceptual distinctions.
sadly, this requires reading lukas actual works to piece it together, since it feels like half the time density is being projected, as if the project is so complex that even getting into what it is entails having to endlessly cycle around the points without entering into any of them. the interview that follows between goodman, lawrence and greenspan does the exact same thing here towards the start, but somehow even more stagnatory, to the point i had to ragequit it at first read, which is my fault more than it is theirs.
in the first part of the interview, lukas is being tortured into talking about topology. ag asks lukas to talk about it, he answers with a few references, the name of his work, and the aesthetic experience you would get as affective discharge after spending time with a topological investigator. afterwards, ag asks him about the space of reason as a concept, but he just says that metaphors can become models. afterwards, bogna shoots her shot about topology, to which he replies that topoogy is about continous fields rather than fixed discrete points, which he later mirrors is his main argument about networks also.
right after this he also mentions vertifcal overflows, layering or conflicts also as potential problems, giving the firewall as a classic example of a virtual border. anna then basically explains the phenomenology of the stack, but she gives a very straightforward read of it, and lukas gives an even more bland answer because he just explains that political regulations prevent any aesthetic reading from manifesting through topology.
this in itself is okay, because lukas argues for the vertical model in a way that challenges precisely this fixed understanding of political entities and their aspects to unconsciously proliferate before we understand their unconscious and dynamic character, however this risks collapsing the dialectic, and turning lukas ontology into an apolitical politics without any creative attempt. i assure you, if you asked fourier, hell, even ireland herself the same question today, they would ground it in a scientific manuscript that tries to be both empirical and creative. this again isnt a weakness of lukas, because he is put in the position where he has to enact the real authority of planetary dynamics in real time, so he cant risk giving us aesthetic slop, his polemical negation does serve to reintegrate his project, its just, well, dissapointing i suppose.
all of this talk on topology doesnt do much before it abruptly gets pushed into being about ai governance by bogna instead - which is fair because it is a slightly more interesting conversation anyways, which means ill have to interact with his work on comparative planetology directly later in my review to actually piece together my preliminary thoughts on his work. the last point about topology right before he goes into ai governance however is about how he wants to understand cloud computing as a concrete metaphor or a real abstraction that can be used to destabilize nodes into flows, where the implementation of ai is directly correlated to non-locality as a lived ontological state.
lukas begins with a very whataboutist definition of non-localities as emergent properties that are constantly changing terrains and end up in unrelated spheres of influence, however this is preceded by the existing definition of the economy itself as an emergent domain, so there isnt much of a difference between non-localities in networks and in stacks or other ontological entities such as plantery scale emergent or dynamic computation, consciousness or etc, the main ontological domain lukas wants to constrast networks with is kind of an association of a couple of different terms (but "the plantery" as such on its own seems to be the essential construct). then greenspan asks him the important questions, but he simply asserts that networks have fixed nodes, which doesnt actually deal with the problem of how dynamics shift in space or why nodes persist despite constant dynamic altercations and interactions in space. also the important question of whether, to which extent, and at what relative speed do ontological shifts happen to large/hyperconnected as opposed to small ones isnt actually answered, which leads to a trivial conceptual binary.
one thing that lukas does help us explore are contemporary contradictions that do in fact transpose or move past a definitive divide between what he calls horizontal and vertical space, so that multidimensionality itself becomes an allegory for contradiction, literally the stack has multiple real abstractions that both work as universals that want to achieve totalizing hegemony, but at the same time only use this operating fantasy to their practical rather than so-called "ideal" advantage if that makes sense, and in this sense his basic questions about what entity belongs to which domain is latourian in its schizophrenia.
for the last part of his interview, lukas deals with the contradiction between an instituion and an infrastructure, as well as with governance against principle. to cut it all short and into its most essential tenants, this allows him to seperate the origin of something from its purpose, whilst seperating the contradiction inherent in something from its actual deployment. this is important precisely because ai is the first passive entity that can be deployed into a type of activity that resembles an agent. this means that lukas successfully defends an original idea of ai, but at the cost of a massive dilemma that appears at the end of the book, which for some reason also becomes the title of the essay even though a large majority of the text doesnt deal with this concept and no dialogue is generated about it past his single statement.
but anyways, if habitability is a less ego- or anthropo- centric version of sustainability discourse, and ai regulation is allegorically equivalent to a volcanic activity changing atmospheric conditions, then technically speaking, the earth itself becomes a type of post-game lobby, it essentially gets equivalized to the air conditioning system inside of a room. in a certain sense, for lukas, the planet is one giant room, the plantery means maintaing trivial control over earthly conditions by a type of positivistic construct of balance, no matter what this balance actually entails or the type of balance it guarantees.
this can be thought of as an inversion of german idealism's understanding of the spirit as a type of inversion of nature, a re-inversion where nature itself is an equalized form of agentiality - if ai is a type of natural control of dynamic interchangable factors in the world, then nature itself is simply its own type of closed-circuit regulatory mechanism, where the concept of "evolutionary blindness" is finally removed in favor of a type of passive agentiality, essentially nature's closed rule book is its own fully self-justifiable form of governability.
but the problem is that in nature, the chance that everything goes wrong is pre-emptively factored into the total fact of its soverignity, so in a sense the concept of imbalance exists, the imbalance of an accident. the ai doesnt make accidents, but it can still "accidentalize" everything, it can decrease everything to any and all base-states, it is a base-state lottery system. the base state of the world can be anything as long as its true to its own self - which arbitrarily also becomes the decision of the ai itself.
in lukas, machine decision is absolutely final even if he argues in other works that it absolutely isnt, because the equilibrium-state of all being is removed from accident, therefore from contigency and chance. i am not trying to moralize or ethicize his system, im taking it at its most base point and hyperficially (superficial + hyper) allegorically and speculatively inflating it to its highest degree. you could argue mathematically that you could do a form of sustianability that preserves all originals including their inherent logics, so that in a certain sense habitability is a type of slow athropy - a constant preservation mechanism that values every pre-determinate against every post-determinate, but this is a freeze-locked zone.
no matter how you spin it to me, cosmology is tied to ethics for this reason, because if you want ultimate control, you must choose infintie triviality. the power to command is equivalent to the power to contaminate, so lukas system to me is nonsensical, at least the one supplied in this interview. now i will deal with some sections of his book, just enough to spice up my own review, but not enough to serve as even a partial consideration of that book on its own. i want to quickly point out that my antagonism here isnt over what is likely to be his philosophy, but only over what he technically said in the interview, or only my narrow reading of this particular segment in the thusly-presented wide-spanning manner, however, with partial speculations of an overarching criticism of ai positivism.
now, even though this also works as its own review, i found it appropriate to keep my review of lukas' work on planetology in this book anyways, as originally my light review of this work was tied to the project of writing the counter-text to "machine decision is not final", and also serves as a fairer overview of my counter to lukas.