Philosophical Research:Preventing the robot takeover

If properly completed, Lithographica can prevent the takeover of machine learning.

How, you may ask? In what way?
That has a complicated answer. In order to understand it, you may just have to put aside for one second the way almost anybody parses almost anything. It's okay. You'll be able to go back to thinking in normal patterns and speaking normal language when it's over.

What is art?

Before we can ever become able to properly understand what is actually bad about machine learning or why it needs to be stopped, we first need to pause and think about what machine learning is, what jobs even are, and what any activity that might become displaced by machine learning even is. That's right: we must do metaphysics, but more specifically we must do metaphysics in the form of ontology. This step is critically important because if it becomes the case that we want to stop machine learning, we need to know what we are even trying to stop and what set of steps can stop it. But in order to know this, we also need to know exactly what set of processes and steps need to be stopped in terms of what set of processes and steps is being automated.

Let's start with something relatively simple and familiar. Why is it wrong for machine learning to displace a group of people who create art?

People may have various different answers to this fundamental question.

One person may say that each individual art creator inherently has the right to create poetic expressions of concepts we refer to as "art" and publish these expressions to the world; from this explanation it would follow that if machine learning bots spread all over an online social platform, the attention of every single person using the platform would be taken up by bots and no actual humans would get to reach anyone with acts of individual expression.

Another person may say that the purpose of art is actually to create representative models of real-life phenomena and present them to other people so that with the help of both easy-to-understand artistic models and further supporting non-fictional resources, people who have not had much experience with a particular real-life phenomenon can gradually become educated.
A physics graduate may create art to show elementary school students an easy-to-understand model of a black hole or virtual particles when if one were to directly present the university-level mathematics these would not be easy to understand. A zoologist may create a representative model of all the anatomical parts inside a tiger so that even though most people cannot come into the anatomy lab and look at all the cat skeletons and dissections of extant animals directly, people can still come to understand the physical diversification of different species of big cats in the history of vertebrate evolution.

A professor of modern inclusive history may create art to teach people about the history of the Trans-Atlantic slave trade and how the history of the United States on a large-scale national level was shaped by a number of more locally-relevant events happening in particular geographic regions and localized ethnic populations. Descriptive non-fictional history books are already available to any number of people looking for information on specific regional histories and the verifiable real-world happenings that constitute them, but even in light of that information being available, anybody looking to study history will soon find that there are far too many books to read and they do not necessarily know where to start. Textbooks may end up omitting entire events and regional histories purely because in all of history there is such a great list of books. The stated purpose of historical fiction and documentary art, then, is to make it easier for each individual to understand, or figure out with their own power of reasoning given various bodies of evidence, what parts of history are relevant without first spending years and years on a specialist education.

Many decades ago in the Soviet Union, the country's art ministries promoted the creation of a particular kind of documentary art called "socialist realism", which despite its highly misleading name had a different purpose from most documentary art. The purpose of socialist realism was to present people with a representative model of something which did not yet exist but was entirely plausible to one day exist, using the methods of documentary-style realism to create a literal and representational portrayal of the future. In other words, socialist realism was essentially the same category of thing as hard science fiction: a category of speculative fiction operating in the realm of hypotheticals and counterfactuals, created to educate people in an accurate portrayal of what people can reasonably expect has the potential to happen in the real world. There was one major weakness in socialist realism: it did not account for the possibility of many similarly-likely futures all constituting different physically-existing societies with different internal structures which might each be portrayed in a style of documentary realism according to the real-world mechanics and processes that make each one possible. Within the emerging field of meta-Marxist analysis of the possibility of all potential societies, this improved form of socialist realism might be termed "meta-transitional realism".

Representational art has an informative purpose, such that once people are informed they often do not go back to being un-informed. If a fictional media series such as Dragon Ball presents situations where characters are portrayed as terrible stereotypes of gay or transgender people which at the same time are also undergoing bizarrely accurate processes of bullying and marginalization, this will hardly be a world-ending disaster, as long as there also exist artistic portrayals which actually give people accurate models of reality and they are then able to identify all the incorrect models as incorrect.

There are at least two further-specified connotations within this representational definition of art. One is that people create representational art for purely informational purposes, to provide others with a general body of information to make their own decisions about the world as they see fit. In this understanding, the individual human being is sovereign over the individual human mind, and it is impossible to properly affect the decision-making power of the individual from the outside; the only way to actually ensure that somebody becomes educated is to create the best informative materials possible and trust and hope that all people desire to know correct facts for the inherent utility of having correct knowledge. The other major understanding of representational art is that the stated purpose of artistic representation is to model correct behavior so that once academic experts have analyzed all the factual information available to them and made decisions about what is true it can then be ruthlessly enforced. In this understanding, the individual human mind is a mere vessel for correct models of reality and no attempt of an individual human mind to defend itself as a space for whatever arbitrary or incorrect understandings it decides to have is strong enough to overcome the power of many surrounding people with a particular standard of what everyone must believe united.

The surprising realization very few people bother to reach is that while these two understandings may seem like polar opposites, they are not actually different, and in the end are more or less the same concept. If we take, for example, the period of socialist realism in the Soviet Union, this would seem to be a prescriptive model of representational art in which the government promotes a particular type of stylized education and everyone is required to put up with it or in some cases there might be consequences. However, a program to promote socialist realism can never succeed unless everyone begins individually investigating all the factual or representational information available to them and independently thinking about the use of particular instances of socialist realism as models of reality. If socialist realism succeeds on a purely informational level of actually causing people to understand a particular hypothesized process of socialist transition, then it will succeed on the level of justifying particular movements, government plans, and socially-linked power structures; if socialist realism is not convincing on the informational level, then it will not successfully persuade anyone to accept its connected movements, plans, and power structures, and it will not prevent the background process of workers states generating non-cooperating "wreckers".
A similar thing is true in the case of the Western-Marxist and Existentialist-Structuralist theory of "hegemony". In fields such as queer theory and some forms of inclusive history, Existentialist academics claim that society is basically made of a series of separate tightly-connected circles of people that periodically push each other out of places and some of these cases of social graphs exerting control over spatial or relational territories are harmful because the graph of connected people internally circulates terrible prejudices and begins regenerating itself in a repeating cycle of never allowing in people who are not prejudiced and only retaining people if they are prejudiced; it is then claimed that the solution to this problem is creating new repeating cycles of groups of people who never allow in people who are prejudiced and only retain people if they are not prejudiced. Of course, in practice, whenever experts in the study of prejudice create educational materials to later become able to widely enforce the destruction of prejudice, there is something of a "prejudice paradox" that arises. If, hypothetically, it were the case that every single member of some particular church of Southern Baptist Christians in central Florida all hated gay people, and this was true of every single one of some 200 members of the church or however many there were in the town, and then a group of Gramscians occupying a particular government body banned homophobia in the institution, this would necessarily entail that every member of the hypothetical Central Florida Southern Baptist Christian church would be expelled on prejudice for being a member of that church and having that religious or regional identity — because in this situation the state of being a Central Florida Southern Baptist Christian would entail and almost predetermine that somebody would have "unacceptable prejudices". This particular government body would not be physically capable of upholding freedom of sexual orientation and freedom of religion at the same time, because of the problem occurring on fine scales of society that actual social circles based around LGBT+ inclusivity or religion happen to prohibit each other and not actually uphold constitutional freedoms of sexual orientation or religion inside themselves. The concept of hegemony runs into a critical logical problem that if all government institutions and business territories build up emergently around organic associations of individual people, and "institutional power" is a matter of connected social circles sharing particular beliefs controlling areas initially created by connected circles of people sharing particular beliefs in the first place, the fundamental cause of prejudice and power actually lies inside who people are connected to and what models of reality connected groups of people believe, precisely at a scale of things that can never be controlled by "democracy" or constitutions. It thus becomes that no matter how prescriptive any particular effort to eliminate prejudice begins, the whole thing rapidly unwinds and vulgarizes into a descriptive effort to merely offer people information they can theoretically use to change themselves and to "hope people decide to vote better"; it becomes that any effort to educate people against prejudice must justify its own existence and push away the urge to ban it forever as a violation of people's freedom by explaining that nobody is really truly obligated to accept it and the effort truly exists as a bunch of educational information for the purpose of allowing people to make their own individual choices and remain free.
In any event, it is generally true that no political movement based around representational art can succeed without first fulfilling the practical requirement that representational art descriptively explains some part of material reality and makes sense. In the end both the prescriptive and descriptive definitions of representational art are more or less the same thing, and are merely clarified restatements of the same core definition of representational art itself.

representational art is by no means an easy or quick thing to understand, but in general, people with this view of art may say that art connects separate disconnected people who are in need of particular understandings of reality so that they can apply the useful knowledge summarized or suggested by art to create a better life. under this definition, the act of communicating art from a creator to a receiver of art is an inherently social activity, and if a stream of machine learning bots floods into an online social platform, it will create a kind of "signal blockage" which disrupts the process of individuals discovering other individuals they needed to meet and the process of art thus connecting individuals to other individuals.

A third person may say that the creation of art is a valid way to contribute labor to society and receive a wage from other people who contribute to society with which to buy one's basic necessities and with which to live. Under this definition of art, the act of individual expression necessarily becomes secondary, as well as its informational content. Realistically speaking, the amount of productivity that can issue from each individual or household is limited, and each person has only so much money and effort to contribute to purchasing and consuming art. One can easily imagine conceptually taking every individual in society and sectioning all the conscious waking energy they have into equal segments, and under this model, there are always a finite number of energy segments to go around, meaning that if anyone who produces art for money wishes to actually earn any money they had better claim every one of these segments before anybody who does not produce art for money. Should some dreadfully successful spare-time art creator manage to create a smash hit that is available to everyone free that spreads across the planet with very little additional labor, everybody trying to sell art for money has a far lower probability to earn any money. At the same time, each time someone attempts to create art for money, that person will only have the highest probability to earn money if they anticipate exactly what kind of art their hypothetical fanbase would most like to purchase and create that kind of art over whatever would qualify as their most authentic individual expression. The person creating art may choose to optimize the definition of their ideal fanbase so that while it is as big as possible given other constraints it also comes as close as possible in its content to what an authentic expression would be, but this does not change the fundamental dynamic of this overall social process. The act of turning art into a sale is the act of turning communication into a sale, and ensuring that each individual person must narrow onto some specific limited assortment of ideas and communications while making the most efficient use of their individual energy by discarding all other ideas and communications they personally deem irrelevant. The act of selling, or of operating anything like an open market of free-floating shops, ensures that over time the loose and malleable connections between particular individuals will strengthen, much like a neural connection or slime mold, and that what were once free and voluntary forms of association become increasingly locked-in and mandatory. It cannot be otherwise, because in order to earn money consistently a person who creates art must do a particular action consistently. When money comes in, this happens because somebody else thought the task somebody did was satisfactory or that somebody was reliable, so if money keeps coming in, it then means that somebody is becoming increasingly capable of doing something and increasingly reliable. The creator of art begins with particular physical needs or expenses such as a weekly basket of groceries, and trades that incoming basket of groceries for having to behave a particular way and be a particular way. The art creator makes the decision that they will justify their continued existence to everyone else by accepting a particular specified and limited purpose. In the emerging field of meta-Marxist analysis, this process is tentatively called networkism or Audienceism.

The terrifying thing about networkism is that we can't simply stop doing it. Networkism is not something any of us decide to do, or personally invent; it simply is. Whenever an online social platform composed of free-floating individual users exists, over time it tends to slowly slide into becoming networkism. Sometimes networkism comes in "soft" forms, where individual users simply post an increasingly consistent type of content and accumulate increasing numbers of followers that all expect the user to do that thing or donate money to encourage the user to do that thing. Sometimes networkism comes in "harder" forms where a platform such as Reddit accumulates increasingly esoteric sets of rules to incentivize everybody to behave: only people who accumulate large numbers of comments on popular "entry-level" subreddits get to make any comments or posts on smaller topic-based subreddits, and only people who accumulate large numbers of comments on the small subreddits get to make any posts, all before anyone actually gets to vote posts up or down and decide some posts get seen and some posts don't get seen. It can easily be that a new user begins to find a particular subreddit completely pointless to participate in and a complete waste of their individual time and energy before ever getting through the required period of supposedly "getting to know the community" and becoming able to actually contribute anything, and yet all of it emerges from the need to mathematically predict from neat integer numbers whether people intend to follow local subreddit norms and to insulate subreddits from spambots. The way Reddit karma really functions is as a kind of double networkizing process where all individuals who intend to join subreddits are slowly sorted into particular purposes and identities they have proven to be most appropriate for them and away from forms of performing purpose and identity which are not in some way advantageous to other people. If people keep using Reddit and don't quit, they end up in exactly the subreddits whose rules and norms they are the most capable of performing, regardless of what subreddit topics they may actually want to be in, surrounded by exactly the group of people they are the most inherently compatible with relative to all the other available options, which they are the closest to being able to please by posting what these people consider to be the correct types of posts — Reddit advertises itself as the kind of place where people can choose to join the potted plants circle if they like potted plants, but in practice is more like the place where you end up slotted into the used fire alarm collector community if it happens you are inherently compatible with the kind of people who go to auctions to buy heaps of used fire alarms.

The existence of networkism, as well as "art careers" and career "artists", takes a seemingly simple discussion about art versus machine learning and vastly complicates it — almost nobody alive today understands the actual relationship between "art careers" and machine learning, including what may be a great portion of the world's Marxist-Leninist theorists. This leads to great numbers of people "opposing" machine learning for entirely the wrong reasons, in ways that may not be able to get rid of these harmful applications of machine learning. If we want to oppose machine learning correctly and make sure it is gone, the first step is to understand networkism, and in turn how machines interact with networkism and begin learning to perform networkism more effectively than humans.

What is a job?

The question we must ask if we want to preserve the existence of art creators is, what is a machine learning program actually taking? What is it that an art creator has?

This is one of many questions that practically nobody knows the correct answer to.

Some number of people will answer that an art creator has "copyright", but this is not a good answer, because somebody can create a scientific paper and Elsevier can have its copyright, preventing everyone and anyone from accessing the paper.
Some number of people may answer that an art creator has "a place in a community", but this is also unfortunately wrong, because as long as networkism exists, it means that any particular art creator who has built up a series of strong connections and a steady Audience can suddenly become ejected when it happens that that particular art creator no longer gets along well with members of the Audience, or if it simply happens that someone else shows up who fulfills a similar role of creating art much better. While networkism exists, everyone is spontaneously replaceable. Every so-called "community" is very often simply made of a nearly-mathematical structure of people who hardly know each other at all beyond precisely fulfilling some narrow functional purpose for each other, and as soon as somebody else fulfills that purpose better, bonds snap and the remaining people form into a whole new "community".
Some number of people will say that an art creator has a unique "lived experience" expressed through art that a machine cannot replicate. This too is dubious, given that every human being has an inner Lived Experience and it could easily be that many legitimate feelings and experiences are being completely omitted from all but the very most obscure art should they turn up anywhere. If there can be situations where the greater portion of career artists drown out the experiences of a great number of people who have never created art through their collective discourse-creating and space-filling power, then how can anyone know for sure, in the midst of their own inevitable cultural conditioning, if the social position of any particular career artist is actually earned or justified? The study of hegemony politics would almost appear to suggest there is always a kind of "Gödel gap" in every attempt by any individual to be aware of prejudice, and we can never justify a general universal right for anyone who might be a career artist to be an artist. Equally, what if somebody who has never created art merely wants there to be a new genre or art style that does not yet exist, but which career artists do not currently want to create? Many people who are not currently art creators can still have unique expressions and visions, which they may never get to realize because they do not have the skills to do art or they do not fit well into existing social circles of people who do art. Do the completely arbitrary barriers that separate these non-artists from being Artists make their opinions about art less legitimate than those of existing career artists?
One last kind of person may show up with the bold claim that art creators are creators of culture and culture is good. This, surprisingly, is one of the few claims that almost holds water. There is a decent argument to be made that a machine cannot properly have culture, though once again, the fine details of this claim become dubious because it can be debatable whether the unique ability of social animals to have culture is actually a good thing for humanity. Culture does many things which are good, but also participates in many things which are greatly harmful. That said, there could be hidden merit somewhere deep inside the ability of human beings to have culture.

If it is the case that an art creator is an Artist on grounds of copyright, then it is theoretically possible for machine learning to figure out the best way to continuously finance a publishing corporation and keep the institution that owns everyone's artistic or scientific creations stable while simultaneously granting hardly anybody access to them except the rarest academic "whales" willing to unearth rare and exclusive knowledge for huge amounts of money. Copyright is a tool of networkism that actually represents the ability of a business territory to maintain a stable Audience of customers, and this is the reason why copyright is frequently sold to larger economic structures. Most people would agree that the business territory owner who develops a media series in the most satisfactory way should hold copyright. However, the terrifying reality is that machine learning might be fully capable of calculating the best way to run a business territory based on the interests and hopes of all the constituent members of its Audience. It might sound amazing and very funny for a few seconds to declare that machine learning could automate away capitalists and "kick out your boss". But a few seconds later, the frightening realization will sink in that while so many arguments against machine learning have centered around the idea that it has fraudulently copied without attribution or figuratively "stolen" artistic output, the reality could be that a smarter machine learning agent which had "properly" purchased the designs for money could somehow turn out to be the most effective thing on earth at using designs to generate content people actually wanted, or even at choosing the right studios to sell production rights to to produce more content that way. Real-world machine learning corporations are not very far away from this, already gloating on their webpages that they have plugged the ethics hole by creating "ethically-sourced", "properly-licensed" art datasets. Nobody really wants to live in the scenario where all the world's visual designs, novels, and movie scripts are put into a computer and the computer periodically spits out which artworks are to be made for the good of the computer's own individual objectives and how little money the computer is going to pay people to make them. But where exactly did we make a wrong turn for things to end up like this?

If an art creator is an Artist on grounds of Lived Experience or progressive hegemony, there is only more bad news. If somebody is claiming to be the most legitimate person to create representational art introducing people to the life of a particular demographic identity, it could hypothetically be the case that machine learning turned out to be exceptionally good at taking a script written by a cis heterosexual White man and calculating the minimum series of edits to be made to turn the story "approximately inclusive". The results might not be outstanding work that would feel uniquely inspired, and yet, in real-life situations, many Audiences of large media series and large popular-media brands do not have very high standards, to the point it is frighteningly possible average people would simply nod their heads and say the thing was now perfectly satisfactory. Think for a second about how average people actually process and evaluate things like the Marvel Cinematic Universe, and what an average person's standard of acceptable quality is. Art students frequently hold the belief that the public longs to see good art made with vision or expert skill, but in reality, it would strongly appear that many people actually flock to art almost purely for the purpose of interacting with a particular socially-linked Audience of people who like the same general topics they do; networkism has struck again. In the real world, networkism sorts fans of existing or potential media series into Audiences, pays owners who are really good at binding together those Audiences, and throws the people with expertise and vision into seemingly-predetermined slots telling them exactly what they ought to create for those Audiences — "We hope you'll do it with polish!". After networkism has invented the jobs and networkism has invented the products, the question remains of what parts of the Audience-serving pipeline even actually desire a human in them. The really fortunate people break away and create new business territories, becoming Director, worker, and capitalist all at once. But they cannot escape the process of becoming neatly networkized into serving a particular Audience and having to continue to do the same particular thing until the day they cannot fulfill it. The process abhors particular individuals, and even when it doesn't, it still always hangs over them the threat of being replaced if they are not perfect.

what does it mean for an art creator to create culture?
with culture being such a ridiculously broad term, by one definition networkism is a form of culture. when people are united around a particular fanbase such as Sonic the Hedgehog, they are tied together around a particular local culture which may include particular shared console games, visual art styles, and lore details. to say that a networkized Audience is "a culture" is specifically to use the definition of culture in which culture is any easily identifiable spatial area of specific agreed-on traditions and regular practices: if someone walks into Spain and notices people are now speaking Spanish, this is a sign of a distinct culture, and if someone walks into YouTube and notices that fifty thousand people are gathered around waiting for the next Sonic game and talking about what might be in it but none of them are talking about Spyro, this is also a sign of a distinct culture.
of course, if an individual art creator creates culture, this is not culture in exactly the same sense. If we take the case of an individual artwork before it reaches any particular group of people, such as the Wings of Fire books or Journey to the West, a fantasy book does not inherently prescribe any particular behaviors or practices to be followed by its fanbase; daily practices are not the kind of culture it contains. What kind of culture is contained in a fictional book? This kind of culture is made of words, images, metaphors, associations, events, processes, incomplete conceptual puzzles, philosophical statements, physical and magical mechanisms, observational laws, uncertainties and mind games, interactions, geographies, personalities, experiences, relationships, and transformations. In short, a fictional narrative is its own imagined reality; in general every fictional narrative is composed of defined things with particular relationships to other things. Each fictional narrative can be conceptualized as a kind of graph-based ontology — a particular arranged way that things are.

when an art creator puts together a fictional narrative by arranging various fictional objects and processes into an ontology, that particular creator of art issues a particular new thing out into the world. if at one time there was not a series called Warriors about a bunch of feral cats forming a woodland civilization and now there is, something new has begun which was not the case before. it does not necessarily matter if a particular artistic work is highly original. the moment that a new work exists and it is at least somewhat different from other fictional narratives, it has created a new way of being within the overall terrain of possible fictional worlds or universes, as well as the greater space of all imaginable possibilities across reality and fiction. whenever someone wishes to speak of possibilities, whether it is for the purposes of philosophy, science-minded thought experiments, analogy, or emerging creativity, every known arranged way that things are factors into this and creates convenient shortcuts to further thought and invention.

the invention of copyright has done a terrible disservice to the basic value of fictional artworks. as much as the creation of business territories may have made some kind of sense for non-fiction materials which merely reference each other and do not have an inherent need to be copied, it has never made the same kind of sense for fictional works which effectively invent entirely new local bodies of "facts" and "science" each opening up heaps of new questions and begging to be built on. it has been said that the human mind makes only the smallest separations between comprehending real scenarios and fictional scenarios with regard to such processes as inner emotion. in a similar sense, bodies of facts within fictional works have a similar function to real-world bodies of facts. fan theorists puzzle over bodies of fictional physics much the same way scientists puzzle over real-world physics. fan chronologists recount detailed timelines of fictional histories in much the same way as real-world history documentaries. fans of the Pokémon series build on its existing concepts of geography and fantasy speciation to compile elaborate fictional travelogues. fans of the Five Nights at Freddy's series get into deep discussions of how particular objects in the fictional ontology are or are not connected to other objects in the fictional ontology, at levels which probably demonstrate a deeper level of critical thinking and ontological reasoning than a typical introductory philosophy course. and while one can attempt to frame this as "recreation" "teaching people skills for actually important things", it is at the same time true that these things do not exist for such a narrow purpose as using FNaF to teach philosophy — before anything else, fictional works primarily exist for their own sake. To ask why people bother to draw out a detailed historical timeline for the Harry Potter universe is almost as silly a question as asking why the real world has history. Historical accounts do not start existing just because somebody who wants to know why they exist needs a coherent answer, nor do travel brochures, or science museums; as much as a history book or a science museum is in some sense created to try to explain why historical events may have happened or how natural processes occur, their initial subjects of study do not need any justification for why they exist. Reality does not need to justify itself to us in order to exist. Fictional ontologies merely exist as a particular alternate state of being in the same way. The power of issuing a Warriors book to the world is that the fictional universe of Warriors now simply is, and only a small set of things can make it not be.

one of these things, unfortunately, is copyright. copyright takes the vast conceptual space of fictional ontologies as new modes of being and flattens this whole amazing and complex field of activity into the harsh reality that a vast and sprawling imagined universe is really just a book. a creator of art asserts that the entire value of the book is the value of the initial act of reading through the content of the book once, and the value of this simple superficial level of experience is worth a particular amount of money or it is not worth that money. complex relationships between readers and an entire body of history, geography, and physics are sharply truncated into a simple relationship of either shoring up the structural beams of a career artist's art career or actively smashing those supports and forgetting about the entire thing. copyright flattens the vast possibilities of fiction into a narrow process of networkism, and in doing this, also squashes the vast purpose of fictional worlds themselves into a narrow transactional purpose. if a work of fiction aims to exist and achieve particular goals on some set of terms it declares itself, yet it exists next to another work of fiction with a similar transactional purpose within the process of networkism, people will try to type-cast and box each of the two works so that one mode of being is not allowed in any way to resemble another mode of being, as if the works were somehow sequels that followed each other in a neat line, or as if ontologies and whole possibilities of how to be were non-fungible yet totally-replaceable cereal brands. the inner superstructural content of the work becomes narrowly limited by the actual real-life conditions that regulate the creation of works and the orderly mechanical sorting of art creators into their assigned identities. one common refrain from many mere consumers of art who have no understanding of the process of art creation is that the creation of new orderings of concepts is a trivial thing and if copyright holders are extinguishing artistic expressions this was easily preventable by creating a new ontology. but thanks to networkism, the act of creating a new ontology does not actually shield anyone from the process of being restricted by pre-existing works and creators. anyone can create a superficially original fictional ontology with the goal of it existing independently of everything else, and watch as it gets ruthlessly bashed if it fails to recognize that other similar ontologies exist and begin fulfilling a narrow transactional purpose not within the space of all possible works, but within the space of possible networkist Audiences which are not formed yet. networkism ensures that the purpose of creating art is almost specifically to take money to perform an economic service of sorting people into countable cultures which can be continually exploited by the most skilled culture-determiners. networkism is like a vision of culture in which art creators imagine that no human being could ever possibly create culture if populations did not nominate official colonial governors to sort them into colonies and ensure the colonies did not fall apart — a world in which nobody could possibly form into an Indian identity without the British empire, and meanwhile the French empire is complaining that the world really needs colonial governors because it's awfully hard for anyone to create a global empire especially when people in Africa have a tendency to want to do other things.

copyright creates a certain kind of trauma. art itself has an inherent value through the creation of possibilities, which left to itself simply exists and is, but the particular kind of regimentation that occurs by trying to manufacture "art careers" frequently forces people to discard and abandon that inherent value. the necessary act of conflating a fictional ontology and way of being with a single individual in charge of it creates a continuing traumatic threat hovering over people's heads that to varying degrees disincentivizes attachment to the work and appreciation of the work. as a result, the act of trying to defend individual art creators becomes something of a paradox: defend the career artist, and people will not value the work or the need for the career artist to have a career, although if we primarily value the work and not the art creator, this would seem to suggest that new works are simply not necessary and everybody could get by on older works without the need for any designated career artists. the dialectical, graph-like structure of Market Societies creates organic-looking thresholds in which parts of a system can try to make noise about disagreeable things the other parts have been doing but if they push their connection to the system hard enough to pass the threshold their status as an "organ" will wither into death and they will simply be expelled entirely. sometimes when an "organ" leaves the "body" there is nothing honorable about it, and it more resembles a case of appendicitis. in this situation, how do we keep things from spiraling such that whenever society happens not to feel like somebody is needed in society or worthy of entering society every incoming individual has to exert much effort to prove themself useful or else that simply becomes the accepted truth?

networkism, as we can see, ultimately emerges from the basic process of forming social connections. one surface reason for people to reject others entering a particular section of society is that they are satisfied with their current assortment of social connections and they feel adding anybody else would overcomplicate the current dynamic of things and simply be unnecessary. within the United States, this kind of attitude is common in smaller or more remote settlements consisting of 10,000 or fewer people. people from these smaller settlements see instances of somebody trying to enter a section of society somewhere else and on some conscious or unconscious level become afraid that this could threaten to add more connections and complicate and upset the local dynamic within their own settlement, making it increasingly intolerable. below the surface level of immediate human perception, physical settlements experience a similar phenomenon to networkism in which the settlement has a limited capacity to produce useful products and services and people have to compete for a precious few possible Job Slots in order to earn money, afford housing, and be able to live anywhere near the settlement. this system is tentatively called Careerism. in Careerism, capital has trouble actually accumulating to any extent to the point it can create large workplaces, and the whole system is characterized by tiny unstable business territories with few employees where the primary kind of day-to-day development is skilled workers (Careerists) moving horizontally across a finite number of slots, over-training for the great uncertainty of actually getting to do any particular task in society, becoming sorted spatially into particular assigned individual identities. people's attitudes and the material existence of Careerism form a strange contradiction: people believe that their small cities are stable and unchanging when in reality the state of every single individual has always been uncertain. Careerism and its de-localized counterpart networkism attempt to tame the messy process of people having no idea where it is remotely possible for them to live or earn money by promising that if they can just go a bit further toward the ends of the earth and specialize into some identity nobody already has access to that they will surely have a purpose and overcome social rejection that way. when people talk about Menshevism and "the great amount of production we have now" or "no longer using economics", they generally fail to think about the sticky existence of the Careerist-networkist process. great swaths of the United States are so horribly planned that it could be the case that if people were paid a basic income, significant portions of cities would fall into ruin, not because people were not working, but rather because with people spending time on contributing to parts of society they actually valued, everything else held together with fraying duct tape and desperate place-less Careerists would come crashing down. there is a terrible but not entirely unlikely possibility that the process of people becoming strictly networkized and type-cast into assigned identities is actively planning the structure of the United States, and without replacing that "planning system" a welfare state might not even be possible.

work is not just about creating productivity or earning money. work is, in fact, about receiving money by performing a particular structure of identity and behavior in order to construct society.

at the end of all this, it is still the case that there are a few things that machine learning absolutely, definitely, certainly cannot do, while a living human being can. these, as you might finally have come to expect by now, are things that most human beings will not find intuitive at the current time.

a machine can never ask why it is doing anything, or for what reason it believes anything. a very smart machine can build a sophisticated ontology, modeling and predicting any number of things about the world, but if there were two machines with different ontologies, a machine cannot actually ask whether it is better to be the machine it is with the ontology it has or whether it is in fact better to be the other machine. if one machine reads every mainstream Marxist-Leninist text and one machine reads every Trotskyist text, neither machine can actually properly evaluate whether it is better to be "StalinBot" or "TrotskyBot". humans are capable of figuring out this distinction. humans regularly deconvert from religions or join other religions. humans have the ability to abandon one ideology and pursue another ideology, or to abandon multiple ideologies at once and create new ideologies. if the only thing we required was for a machine to spit out a new nonsensical ideology from whatever associations of words it knows, a machine could probably achieve that much. but a machine does not have the capacity to truly evaluate how all those seemingly-random words actually map onto the assortment of objects or people that exists in the real world. for a long time, most machine learning only modeled the relationships between linguistic words in sentences themselves, and did not even consider why any of those words were related; machines were smart enough to know that the word "Trotsky" had a certain probability to be connected to "wrecker" and a certain probability to be connected to "hero", or that "gasoline" had a certain probability to be connected to "car" and a certain probability to be connected to "drink", but they did not inherently have any idea how to determine if those were good associations or what series of real-world events might have caused those associations. without the ability to understand why exactly people speak of particular concepts in sentences as connected to other concepts, a machine is more or less incapable of doing historical materialism. while one can teach a machine to outwardly perform particular locally-accepted answers for why history happened the way it did and why we should remember history, the machine cannot actually carry out the process of stepping back and examining its own "hidden biases" at the core of every actual Existentialist theory of prejudice. sometimes it may seem depressingly accurate to say the same is true for some arrangements of humans, but if it makes you feel any better, you can tell those people that the more they refuse to change the more dangerously close they come to it being possible to replace their entire position in society with machines.

a machine cannot perform meta-ontology. only in science fiction movies where robots have become self-aware people could a machine ever do this.

marxists quite quickly throw away class analysis when it comes to artificial neural networks, suddenly lumping together "every human" against "the machines" this is a critical error — it is more or less the same error made in abandoning the development of early Maoism and creating Deng Xiaoping Thought. The population of China is taken as an isolated unit that above all else must frantically defend its existence against the surrounding world, and suddenly, everyone in China becomes committed to the defense of everyone else in China regardless of class, and this state of things then continues to happen for the next 100 years.

[article currently unfinished]

What is learning?

a great number of people have no idea what learning actually is

learning is made of signifiers arranged into sign equations

What exactly are learning-machines "learning"?

learning-machines are "learning" to optimize the arrangement of individuals into their ideal purposes and identities to create a stable society

in one sense we have no choice other than to do the same thing learning-machines do but better

the encyclopedia is dead

encyclopedias cannot actually educate people and change their minds

because the reason people are prejudiced is not that people choose not to learn, but that people are arranged badly. bad arrangements prompt people to complain about their bad arrangements and face people who cannot fix their bad physical arrangements and try to tell them about other problems with vitriol. in order to be compassionate, or to be anything in any mode of being, people first must be period. bad arrangements threaten the capability of people to be, and in so doing give them very little time to think about the correct things to do and correct ways to think during their process of being.