Someone was saying that even if using machines becomes cheaper than using humans, capitalist will still use humans because
"automation constitutes constant capital and human labour is variable capital
The Tendency for the Rate of Profit to Fall disproves that fact"
What do those mean?
Others have already explained the meaning of those terms and I suggest also reading the sections of Capital that deal with those terms, which would be Volume 1 Chapter 8 for variable and constant capital and Volume 3 Part 3 for the TRPF, though the latter one is much more confusing and even somewhat controversial.
I guess I’ll give a shot at explaining it in broad strokes: Suppose you have labourers (variable capital) and machines/tools/property necessary for production (constant capital) in order to produce commodities with a certain value attributed to it by both the labour used to make it and the labour necessary for the construction and maintenance of the constant capital. Labour is the key component because workers already exist prior to the production and seek to continue existing, requiring things like food and shelter in order to survive. They rent out their labour-time in exchange for money which is in turn exchanged for socially produced goods, meaning that this labour-time is the non-capital portion of the value of a product.
The capitalist will extract surplus value from that process, which needs to be reinvested. Both constant and variable capital grow. Broadly speaking, given growth in both variable and constant capital with a constant rate of exploitation (value minus wages) the proportion of surplus value extracted from value produced will decrease.
What your colleague seems to be pointing at is the reduction of the labour-time required for production of goods. An increase in production through better means (machinery, technology, techniques) that enable workers to produce more with the same amount of labour time will, in fact, lower the relative value of the given product (which in capitalist terms actually “increases” the value of that labour). The value itself is given both by the labour-time and capital upkeep (which can also be measured in labour-time i.e. Integrated Labour Cost). This is close to the TPRF but not necessarily the same thing.
I wrote a whole thing, but I think I can sum most of my argument up with a philosophical question. “If nobody made something, who are you paying for it?”
The hypothetical scenario where machine “labour” is cheaper to produce and maintain, and equivalent if not better than human labour crashes against the premise of labour theory of value.
At some point of the production process, either a living, breathing and eating human produced the machine or produced some part of the process for the construction of that machine. Those proletarians need to continue existing in order to work, which has a cost associated with it. Supposing a sector of the economy can be handled autonomously with those machines, their cost will have been the necessary cost of the (human) labour necessary to produce that automation or that provide the inputs for it.
This will mean that that industry will produce little-to-no relative “value” in a Marxist sense, and the cost-price will be determined mainly by external factors such as scarcity or abundance of the production for other areas of the economy, as well as political decisions (i.e refusing to lower prices). Prices can be kept arbitrarily high and wages arbitrarily low, the relative difference of which creating more surplus for the capitalist.
Supposing that the entire economy is maintained solely by machines, and that this automation is so total that not a single part of the process has human input, including maintenance, electricity and all sorts of other “externalities” handled by workers, then whatever is produced has no labour value.
This presents a problem for Marxist analysis, as this presupposes a completely unprecedented world where production no longer necessitates labour. In such a world, the exploiting class would have no need for an exploited class. Whether either class would be liquidated through violence or abolition of class distinctions depends on the specifics of this hypothetical. The answers for this aren’t in the TPRF, but in the historical process of the formation of such a world.
Although I don’t believe such a level of automation will exist in our lifetimes, I predict that the trajectory will remain the same as now, ignoring the possibility of revolution.
Current lucrative sectors of the economy will face desperate pushes for automation (entertainment and services in general with chatGPT and MidJourney, warfare with autonomous drones, etc) and as professions are turned obsolete by machines, workers will be laid off and redirected to more peripheral regions of the economy (food delivery). As those peripheral markets become more consolidated and lucrative, there’ll be a push to automate those too.
At every step, labour will be marginalised and fractured by automation, only to be consolidated in whatever future equivalent of call centres are created, and then automated out of work again. Labour-time will become individually more productive and therefore relatively less valuable. There’s no reason why capitalists of specific sectors would avoid automating it, as they’d get increases in surplus (assuming constant pay for infinitely more productive workers). And workers would be correct in demanding lower prices as automation decreases value, though monopoly capital would exercise its desire to maintain arbitrabily high prices.
So class society would become increasingly unstable as the economy gets automated in its totality, exacerbating the already existing tensions over the disparity of prices and wages and value and labour-time. On the capitalist side this could develop into pushes for diminishingly meaningful jobs on the periphery or even into violent “population control”, and on the proletarian side this could develop into demand for higher pay and lower prices or the redirection the new surplus into public services, or the good old revolution.
Extremely good answer comrade 🫡
https://cosmonautmag.com/2021/10/why-machines-dont-create-value/
For anyone who hasn’t read this article, I think it brings together most of the important points to consider when discussing automation from a Marxist perspective.
I am not the most well read on Marx, but from what I could understand from his discussion on constant and variable capital, the former still requires human work for creating, operating and maintaining it.
Machinery would only be deployed if it was cheaper than the human workers executing the same tasks, and its adoption would lead to layoffs, thus to the reserve army of labor, humans made “cheaper” by competing against the machines.
I am not acquainted enough with the TRPF to comment on it, but my assumption would be that once labor is made cheap enough the capitalists would not have any incentive to keep investing in automation, thus restarting the cycle. I have a vague feeling that the TRPF also mentions the impact on profits of a much impoverished consumer base (the laid off workers) , but I defer to any comrade who can correct my understanding
well capitalists still need humans because:
raw materials have to be transported by someone who knows where it’s heading
someone has to do maintenance to the machines
someone has to sell the products
someone has to deliver the products
and all of that requieres human labor, and that’s where the value comes, because the capitalist doesn’t do any of those things
Can points 1, 2, and 4 not be automated? Especially as machines continue to evolve and exponentially improve?
Machines cannot do maintenance on themselves.
I don’t see why that’s fundamentally impossible though. For example, people are already working on stuff like 3d printers that can produce copies of themselves by printing all the parts. So, you can have machines made out of modular components that can be printed. When a part fails, then a new one is printed and installed to replace it. This whole process can be entirely automated. And this would include the printers themselves.
Sure, but machines cannot do troubleshooting the way humans can. Yes you can have machines tell you what’s wrong, but they cannot reason why a problem exists and how to fix it. There are too many random elements to account for, things that are impossible to account for even. This is why you still have car mechanics, even considering all the fancy telemetry that exists. In the same way, a computer program cannot debug itself.
Yes you can have machines tell you what’s wrong, but they cannot reason why a problem exists and how to fix it.
This is also true of a human worker though. That’s why the profession of “doctor” exists. The question is, why couldn’t a robot be made to perform maintenance on other robots?
We don’t have general purpose AI yet, but I don’t think that’s a prerequisite for automating many jobs out of existence. Humans will still be needed to solve really complex problems for the foreseeable future, but the number of humans that need to work will be greatly reduced. A great example we can already see today are automated ports and factories in China where there’s a just handful people overseeing them.
I feel like you’re missing the point. Machines are quickly reaching the stage where they will able to troubleshoot, think and reason.
That’s what we’re discussing.
This is what you can call an “academic marxist”.
The first point could be boiled down to: you can only extract profit from human labor. Machines can only pass on the value that was put in it by human labor. To ilustrate, imagine that human labor is like a fire, and a machine is like a bottle of water. The fire generates heat, and can pass it on to the bottle, wich can pass it down to something else, but the bottle will never generate heat itself. Its the same with human labor generating value and passing it on to machines, that can pass it down to other products, but the machones do not generate value themselves. For the capitalist to turn a profit they need to generate value somewhere, and if theres no lavor, theres no value being created; so it stays as a 0 sum game (wich means 0 profit).
The second point refers to the fact that every part of the economy tends to lose its margin of profit over the time. That happens for many reasons, but the important point is that: give enough time and any business will become less and less profitable. What that means in the context of what the guy was saying, i cant tell
One thing to note is that machinery does not impart value onto the things they produce because they can produce so many things for a relatively low fixed cost. Marx equates the infinitesimal value imparted by machinery to that of the forces of nature. This is an important part of why the exchange value of goods produced by machinery drops dramatically.
But when Marx discusses automatons, and in the modern discussion of AI, wouldn’t a thinking machine be able to take the place of a human worker in creating value?
Surely an AI can create far far more value then was ever used to create it, especially if it learns and evolves.
You can’t exactly depreciate in an economic sense a thinking machine.
AI does not exist though, it is science fiction.
Not yet. But that’s the entire point of theorising.
Nuclear weapons, cars, jets, computers, and literally everything in the modern world was science fiction when Marx and Lenin were crafting their theories.
That’s why Marx had the foresight to loosely theorise automatons (AI). Something being science fiction should not stop us from pondering its implications.
One cannot theorise about concepts that haven’t been established yet, that’d be assuming facts not in evidence, as it were. We don’t have a concept of what artificial intelligence would be any more than FTL or matter converters.
Marx referred to Antipatros writing about the water wheel being used to grind corn, thus freeing the slaves formerly tasked with this, a notion he decries, saying Antipatros and the rest understood nothing of political theory. You can clearly see he’s talking about automation like the sewing machine, cotton gin, or the assembly line. (anachronistic, I know) In fact, his quote from Aristotle about shuttles that weave by themselves wouldn’t have been interpreted as actual intelligence by any reader until the last ~50 years.
So Marx not only didn’t theorise about artificial intelligence, he was talking about automation that was happening in his time, he couldn’t have. As an intelligent man, Marx extrapolated current and historical developments, but didn’t engage in speculation, which AI still is.
That’s not the passage I was referring to.
Marx quite explicitly refers to automatons and robots from these passages in capital. He directly humanises the machine and this is AFTER he talks about general machinery and the industrial process. So he is not talking about assembly lines and simple industrial machines.
Notice how he differentiates “automaton” from “machine”. He is implying a fully artificial worker, which through this artificial nature, completely removes all of labour value from the equation of goods production.
One can absolutely theorise about concepts that don’t exist yet.
I wasted too much time with the previous version of this reply and it was getting to long with to many references to the source. I’m choosing to believe brevity is gold.
All quotes are from the economic manuscripts. First two here and the last one the chapter right after. Here’s the part between those first two quotes:
In the machine, and even more in machinery as an automatic system, the use value, i.e. the material quality of the means of labour, is transformed into an existence adequate to fixed capital and to capital as such; and the form in which it was adopted into the production process of capital, the direct means of labour, is superseded by a form posited by capital itself and corresponding to it. In no way does the machine appear as the individual worker’s means of labour. Its distinguishing characteristic is not in the least, as with the means of labour, to transmit the worker’s activity to the object; this activity, rather, is posited in such a way that it merely transmits the machine’s work, the machine’s action, on to the raw material – supervises it and guards against interruptions. Not as with the instrument, which the worker animates and makes into his organ with his skill and strength, and whose handling therefore depends on his virtuosity.
I’m sure you’ve seen episodes of it clipped from How It’s Made, or something similar. In many areas of production, the machinery performed most of the tasks and people are around to supervise the machines, press buttons and sometimes move stuff from one machine to the next. That’s what Marx is talking about, he’s extrapolating the cotton gin to talk about a machine that takes in cotton and shits out cut cloth or even whole clothing. (Example mine, obvs)
The last one was tricky, cut up by itself like that:
Labour no longer appears so much to be included within the production process; rather, the human being comes to relate more as watchman and regulator to the production process itself. (What holds for machinery holds likewise for the combination of human activities and the development of human intercourse.) No longer does the worker insert a modified natural thing [Naturgegenstand] as middle link between the object [Objekt] and himself; rather, he inserts the process of nature, transformed into an industrial process, as a means between himself and inorganic nature, mastering it. He steps to the side of the production process instead of being its chief actor. In this transformation, it is neither the direct human labour he himself performs, nor the time during which he works, but rather the appropriation of his own general productive power, his understanding of nature and his mastery over it by virtue of his presence as a social body – it is, in a word, the development of the social individual which appears as the great foundation-stone of production and of wealth. The theft of alien labour time, on which the present wealth is based, appears a miserable foundation in face of this new one, created by large-scale industry itself. As soon as labour in the direct form has ceased to be the great well-spring of wealth, labour time ceases and must cease to be its measure, and hence exchange value [must cease to be the measure] of use value.
I think my case here is even clearer. A “process of nature” is recreated in a machine and performed the tasks previously done by the means of labour. The labourer moves to the side and supervises. There’s no hint now implication of intelligence on the part of what replaced the worker.
This still ended up long so this post I’ll make actually brief: Before we had nukes we had the concept of bombs, even big ones, and we had chemical weapons. The bike therefore wasn’t a new concept for at least the half century before it was first deployed. AI not only didn’t exist, we don’t know on what shape it might. Therefore, any investigation of AI requires many layers of assertions that can’t be verified for some time. All we can do now is speculate, same as Asimov decades ago. This speculation isn’t entirely without value, but it certainly isn’t theory.
Thank you for the write up! That’s given me a lot to think about!
Honestly the only confusing part for me would be Marx’s inclusion of “mechanical and intellectual” organs when referencing the industrial machine then. Computers would still be in their primordial infancy when Marx was writing capital, and I highly doubt he had ever heard of them, so it doesn’t feel like he was referencing a “programmed” machine. Even one programmed mechanically.
That was mainly the line I was referencing when talking about how Marx humanises the machine and gives it a sense of self-autonomy. As why would a cotton gin require “intellectual organs”? Of which there is technically only one in the human body, which is the brain.
The AI follows the same logic, but instead of being a bottle of water its like a lake heaten to boiling temperaturas. It can pass a lot of heat onwards, but still cant create it on its own
But in your example that is not a true AI able to work outside of its programming. I am discussing a true, thinking, machine.
And if you say that a machine still has to learn that information from somewhere. How is that any different from a human?
Thank you so much for simplifying the message for me! I get it now!
Could you please explain to me what the terms mean, so I can understand what people mean next time they use it?
From wikipedia:
Constant capital includes (1) fixed assets, i.e. physical plant, machinery, land and buildings, (2) raw materials and ancillary operating expenses (including external services purchased), and (3) certain faux frais of production (incidental expenses).
Variable capital […] refers to the […] labour costs […].
Automated machines thus are the constant capital. And human workforce constitutes variable capital.
Essentially, machines are not useful without human intervention. Machines can boost productivity of human labour, but they cannot produce value on their own regardless of their level of technical sophistication.
Sure, what terms you didnt get?
I assume from your phrasing you mean to say that “some industrialists will prefer human labour over cheaper machines”. (But you could instead have meant to say that “in society, there will always be a role for human labour”).
I think this proposition is false (in the long run) due to the the profit-maximization logic of capitalism. Even if some industrialists prefer human labour, in the long run they’d eventually be out-competed by those that didn’t. But yes, machines cost a lot upfront which could delay the switchover (possibly even beyond the sunset of the industry!).
I think the “Rate of profit falls” theory doesn’t matter. Assume it’s true: machines lower the rate of profit. But, capitalists can’t (as a group) decide not to use machines. Without control of private assets (e.g. capital) the capitalists cease to be, and become a purely rentier class again. It would no longer be capitalism but feudalism. I don’t believe capitalists have the power (or desire) to cause such a transition.
context might help; where does it come from?
