Editing Philosophical Research:Preventing the robot takeover (section)

== 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 <em>taking</em>? What is it that an art creator <em>has</em>?

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.<br />
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".<br />
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 <em>every</em> 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 "[[redlink|Gödel gap]]" in every attempt by any individual to be aware of prejudice, and we can <em>never</em> 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?<br />
One last kind of person may show up with the bold claim that art creators are creators of <i>culture</i> and culture is good. This, surprisingly, is one of the few claims that <em>almost</em> 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 "[[redlink to whales|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 [[redlink - Anarchy is the most authoritarian thing there is|being replaced if they are not perfect]].


what does it mean for an art creator to create <i>culture</i>?<br />
with [[redlink|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 <cite>Sonic the Hedgehog</cite>, 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 <cite>Sonic</cite> game and talking about what might be in it but none of them are talking about <cite>Spyro</cite>, this is also a sign of a distinct culture.<br />
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 <cite>Wings of Fire</cite> books or <cite>Journey to the West</cite>, 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 <em>is</em> 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 <cite>Warriors</cite> 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 <em>all imaginable</em> 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 <i>arranged way that things are</i> factors into this and creates [[redlink - the batman meets other things Dinosaur Comic|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 <cite>Pokémon</cite> series build on its existing concepts of geography and fantasy speciation to compile elaborate fictional travelogues. fans of the <cite>Five Nights at Freddy's</cite> 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 <em>for</em> such a narrow purpose as using <cite>FNaF</cite> to teach philosophy — before anything else, fictional works primarily exist <em>for their own sake</em>. 
To ask why people bother to draw out a detailed historical timeline for the <cite>Harry Potter</cite> 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 <em>exist</em> as a particular alternate state of being in the same way. The power of issuing a <cite>Warriors</cite> book to the world is that the fictional universe of <cite>Warriors</cite> now simply <em>is</em>, and only a small set of things can make it <em>not be</em>.

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 <em>worth</em> a particular amount of money or it is <em>not worth</em> 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 [[redlink - S2 transgender people can't own Harry Potter books|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 <em>purpose</em> of fictional worlds themselves into a narrow transactional purpose. if a work of fiction aims to <em>exist</em> 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 <em>space of all possible works</em>, but within the <em>space of possible networkist Audiences</em> 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 <em>exists</em> and <em>is</em>, 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" [[redlink - body without organs|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 <em>by</em> 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 <i>ontology</i>, 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 <em>why</em> 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 [[redlink to herd of cats effect|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]