Jump to content

9k:RD/ LogicOfTheArts (Q618 #2)

From LithoGraphica
Revision as of 13:40, 3 November 2026 by Reversedragon (talk | contribs) (copy markup from trans4mer page)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Prompt

What is "the logic of an industry"?
One argument I have heard against AI is that it encourages people to avoid learning to do what people in general are still paid to do. Put another way, both skilled and unskilled workers enter their position in industry on the basis of being able to do a certain set of skills without machines, acting themselves as the "rented machine"; proletarian society can begin only at some point after there are enough people able to fill all the positions required to run a proletarian society, and not while people are still peasants or while people are focused on 'graduating out of' the proletariat and trade unions are not committed to defending people's employment in a particular place rather than surrendering to capitalists' decisions on layoffs or closing facilities. Broadly speaking, a hypothetical proletarian society will begin because society hits a point where no amount of innovation can get around the pattern that all production is based on arranging people trained to act as "task machines" into production structures, and it is easier (simpler; less convoluted) for all the people who live on arranging people into production structures to get replaced by more proletarians or more permutations of 'arranging the proletariat' (council of workers arranges workers, delegates of proletarians go to proletarian legislature, etc.) than for society to get rid of every person trained to do a particular task as "expensive" or "inefficient" and think of some other way to do almost every task than hiring workers. People trained as management eventually start to resemble workers trained for a task more than they resemble the owner of the entire corporation. People hired for unusual levels of skill like concept artists or TV show directors who should be capable of being petty bourgeoisie start to resemble workers trained for a task every time they hand over control of a project they participated in and make no effort to be its owner. But the one thing that is inescapable every time somebody becomes part of the proletariat based on the conditions that create the proletariat in the first place is that proletarians are assigned to a task and they can only stay part of industry when they successfully do that task. It's true that these rules were generated by capitalism and not by socialism; in the Soviet Union, there were already measures to move away from that, like requiring businesses to hire unemployed workers so they can try again somewhere else even if they get fired. But, proletarian revolutions crumble whenever someone turns away from the proletariat — as with many cases of betrayal by trade union bureaucrats who value ending conflicts without a bad outcome more than getting a good outcome, or as with various times in the Soviet Union where people turned away from the bureaucrats because they had no principles and valued "happiness" and "freedom" over large-scale societal outcomes ('the Soviet Union is a dystopia because I'm unhappy') rather than because they understood what was actually wrong with Stalinism, and were no help to replacing Stalinism with a Trotskyist society.
Automation in general does not destroy the proletariat, whether or not it replaces highly skilled tasks and whether or not it replaces the easiest tasks (if "printing presses" are arranged using digital images instead of by hand the industry still functions; when toys or statues that used to be individually made are mass-produced with molds, it mostly means there are more items and even if they are cheaper the total monetary value of items being sold can increase). Nor does automation which is not related to the tasks proletarians are actually used for, like the possibility of replacing search engines with AI and giving people some kind of better way to navigate through all the world's media and businesses (ultimately this might even help some amount for stabilizing previously precarious archipelagos of industry and getting all islands of industry in the world to operate as one system). So, put within the context of the prior history of workers' states, the complaint would boil down to AI encouraging people to lean away from "learning the techniques for learning" which are the techniques they would use to learn how to be workers in any particular industry, and to lean toward believing that society itself is made out of technology and technologies that magically solve the problem are themselves the medium for doing tasks — the vantage point of people who own capital and direct workers to do tasks using capital's social power.
There have been many errors inside Stalinist states on what people think "being proletarian" is (creating all cultural values inside China / North Korea and refusing cultural values that were "arbitrarily made up" in North America, etc), but a definition which directly emerges from the thing that proletarians do — contribute specific kinds of labor to produce a product — is more likely to be right than a definition that emerges from "national values" in the sense that "being part of China" or "having Chinese values" are not how proletarians are created.
When everyday people try to rant about how AI "teaches people to stop thinking" and clumsily go through examples of "AI images as art", "essay-writing AI", "using AI to write emails", and so forth to demonstrate it, this is what I see beneath all of that. Everyday people are taught to think that all labor is "property" and to see something "evil" or "criminal" about treading outside an individual's own abilities although all businesses build on what came before and often without permission. But when you look beneath that outer appearance what you see is that one of the real reasons "property" is finite and valuable is that at any given moment the workers that made it are a finite resource. Novel objects like a newly released video game get their initial price based on the fact that the arrangement of workers used to produce the first copy was rare and unique at that time, while as a disc or a file is copied (or a book, or a molded plastic item) the inherent rarity of the arrangement of things required to produce that item goes down. (The exact retail price shown to consumers does not immediately lock itself onto the actual rarity of the item because corporations need to make money, but 5-10 years later an item will often drop in price on secondary markets, catching up with how many have been produced.) The value of the "intellectual property" comes from the value of the arrangement of workers. But really the workers are not items themselves and provide labor-power, which is the future ability to work. The "intellectual property" is added to existing piles of capital based on actually being able to find workers that can form new arrangements of workers and who sell labor-power to be able to create those arrangements. Each new project created is based on new collections of workers and new arrangements of those workers' skills, not directly on the collection of capital that already exists, even if that capital contains information and skills by its nature (as with a room full of scientific journals at a university many decades ago where publishers still charge significant amounts for copies, or an LLM now). The thing that any existing collection of capital does which gives it power within society is arrange workers, which is an important thing to notice because it means that no more knowledge added to the LLM (no more actual, unique knowledge whose relevance or significance in the material world was not previously verified) without first finding workers that know it, at least when arranged properly. Because of this a trained worker is still instrumental in producing anything. So, the conclusion from that is that LLMs threaten to accidentally cut off the circuit of production by leading people away from the knowledge that workers exist based on physically and mentally solving new problems that are not already encoded in the LLM which will then be put into the LLM. If workers do not learn to "work without cheating", or more accurately, do not learn what workers were created for as a class and what a worker is, workers could accidentally destroy their own emerging civilization and render it incapable of taking power while the bourgeoisie tries everything to make everyone into gig workers or small businesses.
Of course, that's just a worst-case scenario. Hypothetically, maybe there is a way for the world to function without any workers that I just haven't thought of yet and industry would just jump past the age where arranging teams of workers is required for designing new products and force Marxist economic theorists to diagram exactly what that society looks like and teach everybody that.
What do I have wrong about capital and "the logic of industry"?

Motifs or claims in prompt (no AI)

Motifs or claims from response

Subjective themes (no AI)

Usage notes

  • You are allowed to post LLM prompts as prototype pages (the part you wrote manually), as long as they stay in one of the pronounced Talk namespaces and there are no direct quotes from generated text. Concisely describing abstract propositions that turn up several times inside response text — or quoting source texts written by humans — is fine.
  • It's recommended you include all of the AI's "references", such that each prompt becomes a sort of 'combination nickel' that becomes a convenient place for collecting twenty or so links that answer the question. If any of the pages referenced are already nickel pages, you should link the nickel page.
  • None of the commentary on these pages after the motif names / arrows is AI.

Ideologies, fields, topics, or search engines

  1. pronounced Trotskyism / Socialism AI