[ home / rules / faq ] [ overboard / sfw / alt ] [ leftypol / siberia / hobby / tech / edu / games / anime / music / draw / AKM ] [ meta / roulette ] [ cytube / git ] [ GET / ref / marx / booru / zine ]

/leftypol/ - Leftist Politically Incorrect

"The anons of the past have only shitposted on the Internet about the world, in various ways. The point, however, is to change it."
Password (For file deletion.)

Join our Matrix Chat <=> IRC: #leftypol on Rizon
leftypol archives

File: 1662500918774.jpg (11.85 KB, 256x256, ai_catgirl-1.jpg)

 No.1158221[Last 50 Posts]

This thread is for the discussion of cybersocialism, the planning of the socialist economy by computerized means, including discussions of related topics and creators. Drama belongs in /isg/

Towards a New Socialism by Paul Cockshott and Allin Cottrell: http://ricardo.ecn.wfu.edu/~cottrell/socialism_book/
Brain of the Firm by Stafford Beer
Cybernetic Revolutionaries by Eden Medina
Cybernetics: Or the Control and Communication in the Animal and the Machine and The Human Use of Human Beings (1st edition) by Norbert Wiener
Economic cybernetics by Nikolay Veduta
People's Republic of Walmart by Leigh Phillips and Michal Rozworski
Red Plenty by Francis Spufford
Economics in kind, Total socialisation and A system of socialisation by Otto Neurath (Incommensurability, Ecology, and Planning: Neurath in the Socialist Calculation Debate by Thomas Uebel provides a summary)

Active writers/creators
Sorted by last name
>Paul Cockshott
https://www.youtube.com/channel/UCVBfIU1_zO-P_R9keEGdDHQ (https://invidious.snopyta.org/channel/UCVBfIU1_zO-P_R9keEGdDHQ)
https://twitter.com/PaulCockshott (https://nitter.pussthecat.org/PaulCockshott)
>Cibcom (Spanish)
https://twitter.com/cibcomorg (https://nitter.pussthecat.org/cibcomorg)
https://www.youtube.com/channel/UCav9ad3TMuhiWV6yP5t2IpA (https://invidious.snopyta.org/channel/UCav9ad3TMuhiWV6yP5t2IpA)
>Tomas Härdin
https://www.youtube.com/channel/UC5fDgA_eHleDiTLC5qb5g8w (https://invidious.snopyta.org/channel/UC5fDgA_eHleDiTLC5qb5g8w)
>Victor Magariño
https://www.youtube.com/channel/UCJJwfW0R3Lv486AjwUWxIYw (https://invidious.snopyta.org/channel/UCJJwfW0R3Lv486AjwUWxIYw)
https://twitter.com/Victormagajr (https://nitter.pussthecat.org/Victormagajr)
>Elena Veduta
Various videos on YouTube but no channel of her own
>Dave Zachariah
One video on Paul Cockshott's channel

>General Intellect Unit
Podcast of the Cybernetic Marxists

Previous threads in chronological order


Was there any follow-up on Cockshott contacting those socialist students in Chile?


grr, I knew I forgot something:
Computational and Statistical Political Economy Research (CASPER)
A forum for open discussion about political economy, including cybernetics


I don't follow Neurath's hyper-skeptical take about reducing multiple dimension to one. He notes that when different action scenarios are scored by multiple aspects and the aspect scores are put together in some overall score, different weightings of the aspects result in different rankings of the situations you compare, and he is unsure about how you would weight the aspects. He wants to have the relief (meaning the non-aggregated aspect scores) when comparing action scenarios. He loves to see it when one relief covers another, showing that one scenario is unambiguously superior to another – not worse in any aspect, better in at least one aspect, in modern parlance known as Pareto dominance.

But this has to be said: Just because different aspect weightings make the scenario rankings change doesn't mean anything can happen to the rankings. Whenever you have Pareto dominance, this will also show up as the unambiguously better scenario having a better rank than the unambiguously worse scenario, no matter how you weight the aspects (as long as the weights are all positive). Neurath also cringes at statisticians using rank information as if it were richer (interval or cardinal), making some arbitrary transformations. Again it has to be said: Whatever weird arbitrary transformation rules you may choose, Pareto dominance never gets destroyed by that.

I do concede though that choosing weightings will never be without controversy, so Pareto dominance should be made explicit by deleting Pareto-dominated alternatives from the ranking of the scenarios as they get whittled down to produce a winner at some point. This idea could be generalized further by allowing different people in a planning group to use their own weightings and then delete any scenario X that is dominated by scenario Y across all the different weighting patterns the group uses.


>>1158301 (me)
This can be generalized even more yet: People could provide not only their own individual aspect weightings, but also their own aspect ratings. Moreover, instead of providing precise numbers they cold provide ranges for upper and lower boundaries, both for the weightings and the ratings.

And this could be aggregated by a very simple AI that tells them where they need to crank up the precision in order to delete many scenarios. (Of course being very precise about everything is how you quickly get there. But precise thinking is mentally taxing and forcing people into making precise statements will just make it likely that they say something arbitrary just to get over with it. So the program would look for where a small amount of more precise thinking has a big effect on filtering out stuff.)


introduce constraints relating to what you want to achieve. then pick whichever technology mix minimizes labour


Eh, I was thinking more about stuff like choosing which of already made prototypes will be mass-produced, with information like the weightings and how existing prototypes are rated being used to come up with revisions of the prototypes before the final decision is made. The constraints are not immediately given in precise form and people have different initial ideas about what the constraints even should be and might not reach a consensus on that; and the causality of the decision-making doesn't strictly run one way starting with the constraints springing from heads onto blank paper.


I don't think people need to vote on how many settings their toaster should have. Most commodities work fine if someone who understands the problem domain comes to a proper conclusion. Is this planning reaching too deep into trivial details?


Just let users test and rate them or something. Run trials and measure the change in productivity.


>>1158459 >>1158716
One of the main ideas of socialism is a relative shift from ex post stuff to ex ante. There is still trial and error, but we also try to figure out things in advance as much as possible. As you know also in capitalism many prototypes are made and judged and discarded as clearly worse than some other prototypes before they reach the mass-production phase. I'd rather have more of that than less, since it saves you a lot trouble down the line.

Don't just think about consumer products. With toasters there is demand feedback from the mass market. But there are also expensive tools that are not used by any hobbyists, only in production. These won't be privately owned. So the feedback here is surveys similar to those in the prototyping phase.


You could a/b test mass market products


it seems i could not find the pdf's for
>Economic cybernetics by Nikolay Veduta
>Economics in kind, Total socialisation and A system of socialisation by Otto Neurath
do anyone have them ?.


>Economic cybernetics by Nikolay Veduta
an English translation of this is being written as we speak, by a comrade in Ukraine
>Economics in kind, Total socialisation and A system of socialisation by Otto Neurath
I expect sci-hub does but I can take a closer look when I get back


>Victor Magariño
why is he in OP has he made any videos about cybernetic communism or planning? I thought he just defended the LTV


are there any english translations of Neurath? All i can find is the original german


Neurath wrote in German, English, French, and also a little bit here and there in Dutch and even Swedish. But the economics is mostly in early Neurath and early Neurath published in German. Another problem is that the stuff that interests philosophers is the stuff he wrote about doing science and his comments on other philosophers, the least interest they have is in the stuff touching on socialism. Here is book from 1983 with Neurath stuff in English, and while it's not bad, the stuff that they put in it really can't do anything for this thread, except the handy list of Neurath texts in English starting on page 259 where the reader of that essay collection gets surprised by the sudden appearance of economics among the topics.


im looking for his works on economic planning esp. the ones mises responded to


>>1158301 was referring to sections 14 and 15 of Neurath's big essay Foundations of the Social Sciences, International Encyclopedia of Unified Science, Vol. 2, No. 1. Chicago: University of Chicago Press, 1944. Also available in German as: Grundlagen der Sozialwissenschaften.

The article Die Wirtschaftsrechnung im sozialistischen Gemeinwesen by Mises (Economic Calculation in the Socialist Commonwealth) was published in 1920, so it cannot be inspired by anything written later by Neurath. In the article Mises explicitly refers to Neurath's 1919 publication of essays Durch die Kriegswirtschaft zur Naturalwirtschaft (this does appear in the pdf's list of English Neurath texts as Through War Economy to Economy in Kind).

The article by Mises became part of a longer work later, Die Gemeinwirtschaft: Untersuchungen über den Sozialismus (in English: Socialism: An Economic and Sociological Analysis), but there are no references to other writings by Neurath to be found in it.


The following is a translation of a German translation of an article originally published in the Russian magazine Problems of Philosophy. The German version ran in Ost Probleme in 1963. Retrieved from https://www.jstor.org/stable/44927433.



This is just some tedious (no doubt made even more tedious by translation) fart-huffer obsessed with labelling people (you see this here is very Hegel, that there is very positivist blahblahblah) and got nothing on socialist economic planning.



This is the cybernetics thread not the "socialist economic planning" thread. The article doesn't mention Hegel once. You are the tedious fart-huffer obsessed with labelling people.


>This is the cybernetics thread not the "socialist economic planning" thread.
its both tbh


>why is he in OP has he made any videos about cybernetic communism or planning? I thought he just defended the LTV
the last OP was so cockshott-heavy that maybe I overcorrected in the other direction. victor does good stuff though


what >>1161926 said. the cybersoc thread is where planning is discussed in detail


Should have just removed Cockshott. He is a succdem technocrat and anti-Marxist


>Should have just removed Cockshott. He is a succdem technocrat and anti-Marxist
you don't understand what any of those words mean


nice b8


Cockshott rejects dialectical materialism, agrees with Böhm-Bawerk in rejecting Marx's LTV and rejects the existence of class consciousness and in turn its role in socialist revolution. Hes an idealist that thinks socialism can be achieved by making the best argument in the market place of ideas, that the problem with socialism is that it hasn't been scientifically proved correct to the bourgeoisie.


What if Cockshott got shot in the cock?


File: 1662669686688.png (564.37 KB, 868x649, whatif.png)


…he literally never shuts up about the LTV being correct you idiot


>taking the bait


No he doesn't, he accepts that the economic calculation problem is real and believes that Marx made an error between volume 1 and 3. He endorses the liberal view that there the transformation of values in to prices are a core of Capital and thinks that he solves this uniquely through calculation. Thats why he is constantly talking about Sraffa and Okishio.


stop feeding the troll.


It look like a more has become available on sites for free ebooks and PDFs since the last time I checked. Otto Neurath Economic Writings Selections 1904–1945 looks more on-topic.

If you want to passionately argue a point about an article you haven't read, the least you could do is a quick ctrl-f check.


This one ?.


Notice how they just ignored you instead of understanding what you said. Cockhott's endorsement of a 90+ year disproved liberal economic hypothesis is clearly idealist in nature. Nevermind that Cockshott's recent videos are idpol culture war nonsense to boot.


File: 1662801320798.webm (7.78 MB, 864x864, 158812930807.webm)

Wanna petition the devs to wordfilter "AI" to "ML"?


File: 1662803386515.jpg (29.9 KB, 434x480, pink think.jpg)

post the vid in the OC thread
personally I'm more in the camp of calling it "machine fitting" or similar, but ML is good for the lulz and confusion
also I hope you're voting in ms. /co/


File: 1662803580905.webm (1.9 MB, 640x480, robutt.webm)

I'm not on /co/, but if they're holding it on plus4 then I'll drop in.


File: 1662804097296.mp4 (1.96 MB, 316x240, ddaig4.mp4)

>I'm not on /co/
for shame


File: 1662804640231.png (48.3 KB, 261x230, heroes.png)

Damn /co/mrade, it's only been up since 2006. back in my day, we had /pco/, not /aco/.


This is funny to not at least try once



old video but dont know if it got posted. There's a spanish language version available too i think


I think it was posted a couple of threads ago but reposting it doesn't hurt


Hey what classes should I take in college if I want to build communism?
Tentatively majoring in comp sci. Should I take any econ classes?


File: 1662874387542.png (215.45 KB, 481x485, ana_kasparian.png)

depends. are you planning on graduate school? If you are not going to grad school then still major in CS to have a job.

If you are, id double major in math because honestly all graduate work in CS and even econ requires a decent amount of math. Econ uses calculus and real analysis and differential equations and linear algebra and probability/stats (as do all sciences including physics and CS).

Honestly at everything but an elite college an undergraduate degree in econ is considered a joke. At most you should minor in it. Here's a quote from a graduate program in econ:

>Prior degrees in economics at the B.A. or M.A. level are not required. There are mathematics requirements. Applicants for admission to the M.A. program must complete a one-semester course in calculus at the University level prior to entry. A second course in calculus is highly recommended. Applicants for admission to the Ph.D. program need additional mathematics. We will consider applicants who have completed the equivalent of two three-credit semester courses in calculus prior to entry in the fall semester. However, we strongly recommend the completion of two additional courses prior to entry in any of the following subjects: advanced calculus, linear algebra, differential equations, real analysis or related areas. There are also economics and statistics requirements. All entrants in the M.A. program must have completed three one-semester courses prior to entry in the fall semester: introductory statistics, intermediate microeconomics, and intermediate macroeconomics. Some applicants, at the time they submit their application forms, will not have completed the above requirements. They may be working toward their completion prior to entry in the fall semester. Such applicants may be admitted conditionally. All conditional requirements must be completed prior to starting the M.A. or Ph.D. programs.

If you are going to grad school then do math/cs and maybe take the minimum of econ classes to get into grad school for econ (which will probably be covered under geneds anyway). unless you're going to grad school for CS in which case the econ classes are optional.


econ is likely going to be porkoid stuff. math with a focus on optimization and high-performance computing (HPC) is the obvious thing for planning. we mostly talk linear programming here but semidefinite programming is more powerful and not fundamentally more difficult than LP. pdf related explains why. would be good to have more people with knowledge of this stuff


>Sesión de Formación de Cibcom 2: Introducción a la Econofísica
<Os presentamos nuestra segunda sesión formativa, esta vez tratamos una introducción a la econofísica.
<Algunos de los artículos y libros utilizados para preparar la presentación son
<Ian Wright - La arquitectura social del capitalismo (https://cibcom.org/la-arquitectura-social-del-capitalismo/)
<Adrian Dragulescu, Victor M. Yakovenko. Statistical mechanics of money (https://arxiv.org/abs/cond-mat/0001432)
<Allin F. Cottrell, Paul Cockshott, Gregory John Michaelson, Ian P. Wright, Victor Yakovenko. Classical Econophysics
<Emmanuel Farjoun, Moshé Machover - Laws of Chaos : a probabilistic approach to political economy
Premieres in 3 days (2022-09-14)


Cibcom just published this short clip on their YT channel
which is a translation of an excerpt from this: https://www.designing-history.world/en/theory/response-to-an-ultra-leftist-critique-of-cybersocialism/
Dapprich should also go in the OP I think


wtf is this a saustika


do undergraduate math majors even take classes on optimization?


optimization is very much a thing in industry, so there will always be demand for experts in the field. same goes for simulation, which very much overlaps with optimization


thoughts on this? written by one of the cibcom guys
>Sketching a proposal for a complete and reflexive Conscious Social Self-Determination (CSSD)
it's a "part 2" to this post:
>Tracing up the common essence of the three main left-wing historical movements


Could you please expand on this? Give me some books please (completely serious)


File: 1663793913250.png (145.55 KB, 609x361, meme.png)

that anon is simply wrong. do not humor them



Pretty vague.
>What if the algorithm itself contains the necessary procedures that allow society in general to modify it?
Ok? If you squint, you might say law is kinda like code, and there are laws for modifying laws… So what's the big difference? Will we all work on the code? Can code be more readable than law texts? The texts slips into outright poetry in places (not a fan of that):
>The Algorithm combines thus perfectly with a decentralized base military power, i.e. guerilla tactics and popular milita, erasing the need to have permanent centralized armies which most commonly constitute the stronghold of reaction.
Maybe I will have a proper comment after another article or two, at this point I struggle to see where the author is going with this.



cockshott responding to reiterate the exact same wrong points again does not make him correct


Did Paul Cockshott actually post a wojak of himself or have I finally snapped


so it seems. the crying wojak is david harvey I think


I think it looks more like Kliman based on the hairline


looks interesting ill take a look later


>>1183131 (me)
>Pretty vague.
yeah, that's the feel I get as well


Man this thread is comfy.

How do I start a cybersoc group locally?


good question. I guess put up posters. if you have a uni nearby that can be a good place to find people who are interested


File: 1663973581470.png (541.33 KB, 886x1023, topkek.png)


lol what if he actually did?


>Posted by EU Times
Can someone find the source? Interested in reading the facts of the matter here.


david harvey has proposed something very similar to this, and I've seen other marxists shoot it down because you can get around it with commodity money like gold


This is probably fake like the social credit shit, but I hope it's real


It's just labour vouchers anyway effectively, worth a shot maybe.


it says digital "money", but that might just be EU Times butchering things
it very much depends on how it's arranged. if it's literally just currency with an expiration date then that doesn't change much. if on the other hand it's the first step toward a planned sector where this "money" can be used to requisition goods then it's a different thing


"Just use another currency bro." Marxists repeat an "anarcho-capitalist" pro-tip, which you can only follow up to a point. The state forces you to pay taxes. The state can force you to pay the taxes only in its own currency, making it necessary for you to acquire some of that. Whether this currency is made very inflationary or the opposite or given an expiration date does not undo this necessity. Even if the state does not force you to pay taxes in its currency, the state may hold monopoly power over some very basic services and use that to force you to use its currency. For example, the state may be the only landlord or control public transport. Where I live the bus drivers do not accept payment in gold or Bitcoin (whatever they may personally think of accepting other payments, it is not in their power to make that decision).


the point being that you buy up gold with money about to expire, repeatedly, and then sell that gold for fresher money later to take money out of a savings. Not that you try to use gold as a daily currency.


to be honest, even with the problem it seems like a good step, and i think people tend to seriously undervalue partial solutions that can be worked through with patches later if there's a plan for it. It's not like you couldn't start imposing restrictions on buying commodities as value storage later.


what >>1187747 said
expiring currency does nothing to prevent accooomolating capital
you can't get rid of currency by writ. if you try then you will fail. whatever replaces it must be superior so that people prefer using it over money


>you can't get rid of currency by writ.
Of course you can, if you have the force to back the decision.
>whatever replaces it must be superior so that people prefer using it
People who use a particularly inflationary currency rarely do that because they estimate its future value as high relative to other currently less inflationary currencies, they do it because they have to.

The state declares a date when the old currency will become invalid and some rules about an exchange rate (the rate is what it is because the state says so, nothing to do with supply and demand) which gets applied to all book-keeping and contracts using the old currency. Citizens can exchange their old currency in state-run institutions. (Various different regulations possible here: They may only do that up to a point, anything above is suspect and requires several checks, another higher threshold may mark what is verboten; also citizens may get some for free.) This has happened several times in the history of Germany and those in power never bothered to get permission from the larger public.

(I also vaguely recall a story by I think NPR which among other things covered currency reform in the DPRK. The gov used the rule that you could only exchange so much of your old currency. This was reported in a very sad tone, since the piece was about some femspiring fempowered femtrepreneur making lot of money in the black market by literally stealing stuff.)


>Of course you can, if you have the force to back the decision.
if you think force can circumvent the laws of motion in market economies then you are mistaken, you waste of a quad. if currency is useful in a situation then currency will appear


>and then sell that gold for fresher money later to take money out of a savings
you can't do that though, state just wont let you stockpile and sell back gold, buying gold in the first place would attract its attention, and then you can't even sell the gold, you can't take personal vouchers for it, you'd have to barter it, and then ppl would prolly be like… "but I have acquired these with my vouchers because I intend to use them ? why would I want your fucking gold which cant buy shit ?"
your ""loophole"" only work with the idea that the system is dumb af and nobody try to fix people trying to recreate a capitalist accumulation or just finding exploits in it (and isnt even really viable even if that was true)


Expiring currency will have some effect on the middle strata of society to keep them from gaining the capacity to start using their monetary accumulation as leverage over others and going through the trouble of converting it to precious metal stores would be effort that a large number of people are honestly not going to go through with.
People are going to recognize the value in precious metal value stores and china's domestic market is not gonna be small. Don't assume everyone's going to be on the same page, either in rejecting or accepting this concept, to various degrees.


>Expiring currency will have some effect on the middle strata of society
didn't finish this
the upper echelons of society basically won't be meaningfully effected. It'll increase the velocity at which money moves, solidify, and encourage foreign capital stocks. But obviously, tackling big bourgies requires tools developed specifically for that task, you cannot simply implement wide-ranging laws on them, they need to be hunted down personally.


Do you think Germany is a fictional country or didn't you read the whole post before replying?
That same poster qualified the statement and probably agrees with you, no need to be so hostile. IIRC when the Nixon administration ended the gold-exchange guarantee of the dollar they also made it a pain in the ass to obtain gold.


im doing 2 years at CC, 2 years at public uni. The uni I want to transfer to offers political economy as a minor, i'm wondering if I should take that?

It feels like to be able to work on planning stuff I should major in math and go to grad school, but i'm probably not smart enough to get a grad-school acceptable GPA in math while also working full time, and I want to be able to not be super poor as a math major (only good jobs are finance and national security related)


>state just wont let you stockpile and sell back gold, buying gold in the first place would attract its attention
then another commodity will take the place of gold
every petty porky will be able to get around this by just buying and selling shit from each other. currency that expires doesn't interrupt M-C-M' circulation. it's like you fuckers don't understand what financial tools are


>Teh bRiLlant porKycaLiSts wiLL cIrcumciZe All yOuR meAsureS!!1
It's interesting how a "Marxist" with zero interest in empirical observation often ends up talking like an ancap spaz. Starting from the wrong assumption of commodity money, you necessarily underestimate the power of the state.


we'll just see, won't we


We already saw USA leaving the gold standard without having them Venezuela-style inflation and then going back to gold with the tail between the legs (contrary to Trotsky saying leaving gold = chaos). We see no massive capitalist gangster network in North Korea hoarding gold or what have you.


>without having them Venezuela-style inflation
mhmm, but if that had happened you can be sure as shit that people would have resorted to some more stable currency
>hoarding gold
that's not the point anon. the point is M-C-M' works just as well with money that "expires" within a year or ten



The fundamental thing is that there's not much fundamental difference between making money inherently expire and having a negative interest rate at a bank that charges you for the account.
From the perspective of the potential account holder, there's no difference if the money simply ceases to exist or the bank takes it. It hasn't fundamentally changed the cycles of capital accumulation in Japan.
>hurr how dare you question communist attempts to fight capitalism
This will have some effect, statistically, and more measures can be brought in later, with time. The attempt to fight wealth accumulation needs to be looked at as a megaproject that is handled piece by piece in stages. Understanding the laws that govern value accumulation help you plan out your next steps. I'm pretty sure this was a topic that came up in committees and will be addressed eventually, one way or another.


listening to Stafford Beer's "The Real Threat to All We Hold Most Dear" again because it's september (https://www.cbc.ca/radio/ideas/the-1973-cbc-massey-lectures-designing-freedom-1.2946819) and just making myself feel deeply miserable knowing none of the problems he raises got solved and all this poor nerd's friends got killed for nothing.


1970 marxist book on cybernetics and philosophy I found, leaving it here so I don't forget to check it out later.


thanks anon


if it makes you feel better it also drove him insane as well


Work from observable reality to theory, not from what is "common sense" among theorists. People don't flock to the strongest currency. If most of the economy is nationalized and there is still cash, the cash is used in black markets. New capitalists are generated in these black markets. Currency reforms like in North Korea throw a wrench into this process.

What you seem to do is assuming that people only want to change the financial system (which of course can't work). This assumption is not a plausible inference about the position of anybody ITT.

>The fundamental thing is that there's not much fundamental difference between making money inherently expire and having a negative interest rate at a bank that charges you for the account.
It amounts to the same if cash is banned. If cash is not banned the difference is that with expiration people can't decide to hoard the cash to avoid the negative interest rate.

>all this poor nerd's friends got killed



So i recently learned several things from marxism-cockshottism:

Capitalists take only half of your work results as surplus value, so:
- if you recieve a wage twice as average, then you are not exploited
- if your wage is even higher than you are actually exploiting others together with capitalists
- actually it's not even wage, it's wage per head in the family, so if two people work same job for same hours recieving same wage, but one of them has two children and a wife who doesn't work, and another is bachelor, then bachelor isn't exploited.
- therefore gays without kids are not exploited
- people in poor countries receiving lower wages is due to their lower productivity only, since capitalists would never take more than half as surplus value
- middle class exists apparentely and non-exploited gay and bachelors belong to it
- Marx never said that class is defined by relationship to the means of production (picrel)

So, the questions is this - was Cockshott always this retarde and i just didn't notice it, or is this the age thing?


take paul's dumb takes to /isg/


If you want to talk about LGBT stuff, there is the general on >>>/siberia/305161 and there's also a current thread in the /leftypol/ catalog because of the referendum in Cuba that you might have heard about.

If you want to talk about economical analysis, you can do that here.
>Capitalists take only half of your work results as surplus value
Yes on average and it's an interesting finding, but so far it is "only" empirical. That is Cockshott and company don't claim to have a reason why it must be that way and I doubt that they think it must be that by some economic law. It's a standard assumption of Marxism that neither the size of the surplus nor the share of it that goes to the capitalist class are automatically derived from market mechanisms. The workers can grab a bigger share by organizing. (Cockshott doesn't follow Marx everywhere, but where he disagrees, he is explicit about it.)
>if you recieve a wage twice as average, then you are not exploited
Does not follow globally as it is jumping from aggregates to particulars. It would follow if everyone had the same productivity. (Besides, it would be only about the quantitative side of exploitation.) It's a smaller jump in the argument when talking within the context of a nation, but it's still iffy talking about individuals that way. By going again to aggregates within the country the jump in the argument again becomes smaller.
>people in poor countries receiving lower wages is due to their lower productivity only, since capitalists would never take more than half as surplus value
Nobody is saying that, but lower productivity certainly plays a big role in lower wages. You don't need racial essentialism to explain that, you can just look at the effects of malnutrition, lack of schooling, etc.
>relationship to the means of production
Cockshott is right that you can't reduce class to that question, otherwise slaves and the working class would not just have some parallels, they would be the same class. Of course one can say that slave owner is roughly to slave as capitalist is to worker, but Marx doesn't talk about "the slave/worker class" and makes a distinction.


>Yes on average and it's an interesting finding, but so far it is "only" empirical.
I have seen zero evidence for this, in fact Cockshott threw a misquote from Marx to me when i asked.
>Nobody is saying that
Cockshott literally did say this to me personally and to some other guys in the comments under his video.
>Cockshott is right that you can't reduce class to that question
You pretty much can. Slaves don't own any means of production, that is in itself is a form of relationship, and they are themself a property of someone who owns means of production.


Cockshott's althusserian influence is coming through here, Althusserians like Wolff for example don't define class as being determined by relation to the means of production but as a process instead and related it to the production and distribution of surplus value rather than merely a relation.

>Class, for Marx, refers to how, in production, a surplus gets produced. All human societies produce such surpluses. However, societies differ in how they organize the production and distribution of this surplus. In Marx’s view, there have always been subsets of populations in communities (from families through villages to whole nations) that have performed labor in the production of goods and services. Those subsets have always produced more output than they themselves consumed: the “surplus” output or simply the surplus. That surplus has then been distributed to other persons inside or outside the community. The class structure of a community or society is then its distinct organization of the production and distribution of surplus.

that being said Cockshott seems weirdly ideologically motivated to claim all LGBTs (except lesbian maybe) are petit booj or PMC. He even threw out the idea of "LGBT serves imperialism" even though he himself already rejected leninist theories of imperialism but hes turning inconsistent with his own work just to dunk on the LGBTs


but slaves and proles belong to different classes precisely because of their different relationships to the MoPs (or to production you could say). proles do still own one form of capital, their own labour power, which slaves do not. the prole is "free" to sell their labour power to whoever they want


>>1192747 (me)
moreover, a slave can be better off than a prole, can retain more of the value they produce in certain situations


Yeah, he basically did a switcheroo with the arguments, starting with exploitation but measuring wages per family member instead just to claim that gays are not exploited. This is…not very honest to say the least.

>rejected leninist theories of imperialism


I already remember feeling iffy about his takes on materialism, now all that..


You appear very new to this theory stuff, so I suggest you ask more question in the thread "QTDDTOT Questions that don't need their own thread": >>296564 I answered you there. I also invite anybody else to take the discussion there if they want to continue with that instead of derailing the thread.


I am pretty good at the "theory stuff". Marxist theory, not whatever this is.


Cockshott seems to completely disagree with theories on imperialism, if these videos are anything to go by.


his takes on materialism are pretty good tbh, his take on imperialism is more like a western marxist take on it. From what I read it was because he got turned off of Irish anti imperialism in the 1970s


His materialism takes leave me sometimes intrigued and impressed and other times feeling like he's a bit of a sperg trying to force things to fit where they don't really make sense.

I still have no idea what the fuck it means to have "idealist supercomputer hardware architecture"


I suppose abstracting away the actual physical side and only considering programming languages "in themselves" (or having benchmarks in the context of a particular hardware and then you continue to use the benchmark while forgetting what its hardware context was), and then saying this or that computer is of this or that design because this or that type of software runs on it. Saying software X will fix problem Y when problem Y is due to hardware.


Labor power is not capital as far as I am aware, until it alienated from the worker.
This is literally the appropriate thread, our discourse is primarily based on TANS so critiques of TANS foundation in Marxism(or lack thereof) is on-topic.


How can labor vouchers price in environmental externality?
One theory is to base the price on the labor cost of rectifying the externality, however this mirrors the neoliberal programs of cap-and-trade and BECCS in failing to realize that not all environmental damage is reversible.
The other option is to have carbon and energy quotas, at some point, these quotas can be thought of as another non circulating distribution mechanism like vouchers. The problem is how do you budget for myriad externalities, And how many classes of externality need to be constrained to continue life on earth? For example, we might have labor vouchers, a monthly energy voucher, a land use or problematic commodity voucher (beef for example has an affordable SNLT but will destroy the biosphere in a few decades unless consumption plummets). At what point do we stop itemizing and use political control instead? And then, why use vouchers and not instead a in-kind (ration) system with consumer preference mechanisms?


in TANS i think cockshott says price in the labor required to research an alternative.



price representation is always arbitrary, this is not a problem unique to labor vouchers or economic planning

this is angels-on-the-head-of-a-pin stuff


How is that possible to forecast?
I am not aware of any firm or government agency that can accurately forecast R&D costs, which is one reason why the subject is often tax deductible in the private sector.
If you made a bureau of R&D you could possibly take the average of hundreds of projects, it's going to be impossible to forecast the deployment costs of the new tech before it is developed, so there is a time lag problem.


I'm not saying that it's a unique problem, I'm only saying that any mode of production which does not solve it will self destruct.



by nature this is something you can't have a pat answer for that you can explain on an internet forum, it'd be like trying to explain energy demand curves or something. the salient difference is that a planned economy could constitute those costs as prices by means of the planning body deciding on a different price representation, and a market economy cannot do anything like that because it's always ultimately profit driven

it's a categorical difference, and guessing at the best way to represent environmental costs is irrelevant to the fact that one system could have the potential to deal with them in some respects and the other categorically does not


> he got turned off of Irish anti imperialism in the 1970s
pussy couldn't handle a little car bombing here and there. tiocfaidh ar la dickblast, you bloody anglo bastard


>How can labor vouchers price in environmental externality?
there are no externalities in planning once it's properly established. but until say atmospheric CO2 is brought into regulation you could have a rationing system
>One theory is to base the price on the labor cost of rectifying the externality
you add relevant constraints, thus making these "externalities" internal
>how many classes of externality need to be constrained to continue life on earth?
this is a question environmental scientists can answer
research is a multi armed bandit, a problem that will never be solved


>critiques of TANS
What the other poster talked about is not in TANS. Leaving aside how wrong and misleading it is, all of it is in reaction to stuff researched and published later than TANS. That stuff had no influence on TANS and is irrelevant to socialist planning algorithms.
>we might have labor vouchers, a monthly energy voucher, a land use or problematic commodity voucher
The more voucher types we use to allocate a given pile of items, the lower the consumer freedom (and what's highly likely: the lower the utility).

Toy example: 10 different product types, every item costs one budget point, no consumer wants more than one unit of the same item, we assume that the items are made to order and no shortages. Alice has 4 consumer points. If she uses up her budget, there are 10 × 9 × 8 × 7 = 5040 combinations to choose from.

If the 10 different products are divided in two sets with the same number of different items and there are two voucher types for each set and likewise the points of Alice are equally divided into these voucher types, and again she uses up all her points, how many combinations are there? (5 × 4)² = 400 (and these combinations were all also available in the single-voucher system).

Hard physical limits can be put into the planning algorithm without having a direct representation in the personal budget system.


>>1195684 (me)
Sorry, I forgot something in the calculation. The above treats the same item collections as different if they are acquired in a different order. So I have to divide the results by the number of ways they can be ordered on a line. 4 items can be ordered in 24 ways, so in the scenario with one voucher type, Alice has 210 combinations to choose from. And in the scenario with two voucher types, there are 10² = 100 combinations.


>anon rediscovers the remuneration problem
none of those is new


>>1195759 (me)
I chose very low numbers for items and budget points due to laziness and that ruined the point I was trying to communicate. The example makes it appear as if your freedom got cut roughly in proportion to the number of vouchers. If you choose bigger numbers you will see that it makes a dramatic difference.

I just said what's a problem with multiple vouchers and made no claims of originality. Whether that criticism is old is not an argument against its strength. I don't know why you would call it "remuneration problem" which sounds like being strictly about the size of salary differences. The problem described manifests irrespective of what size income differences are or whether they exist at all.

There is a potential additional problem with income differences in a multi-voucher system in that mutual income envy can go over 50 %, but that can be simply avoided by only having differences of voucher income in the form of sets and subsets, so that the different incomes despite being heterogeneous can be ordered at least in ordinal fashion.


yes, having a rationing system besides labour vouchers does indeed further constrain demand and can cause discontent
I'm curious why and where you think there would be income differentials. we can certainly imagine a situation where fuel vouchers are unevenly distributed based on need. that is a political issue


>>1195856 (me)
100 different items and 12 consumption points, other assumptions are as above (nobody wants the same item twice, budget gets exhausted).
(100! ÷ (100 − 12)!) ÷ 12!
(100 × 99 × 98 × 97 × 96 × 95 × 94 × 93 × 92 × 91 × 90 × 89) ÷ 479001600 = 1.050421051×10¹⁵

50 different items and 6 consumption points of voucher type one; 50 different items and 6 consumption points of voucher type two. So for one type there are…
(50! ÷ (50 − 6)!) ÷ 6! = 15890700 different configurations. For both types together we have 15890700 × 15890700 = 2.525143465×10¹⁴

Comparing the single-voucher system with the dual-voucher system we get (1.050421051×10¹⁵) ÷ (2.525143465×10¹⁴) = 4.159847017. The single-voucher system gives you more than 4× the combinations to choose from here (and contains all the combinations of the two-voucher system). Someone tell me if I screwed up the math.

Unequal distribution is the default assumption (see Critique of the Gotha Programme). Equal distribution does not follow from the position to each according to their deeds (bonus income is possible where physical output where that can be measured, for working more time, for availability on short notice, for dangerous work, and based on voting by your colleagues). Equal distribution does not follow from the position to each according to their needs (for example legal guardians of children). Equal distribution does not follow from a compromise between these two positions.


>Unequal distribution is the default assumption (see Critique of the Gotha Programme)
right. and all of this gets much easier in communism which we can then define as whenever we can dispense with remuneration (including ration coupons/vouchers) altogether


That's the term usage as one finds it in Lenin. Marx in CotGP does refer to a system using vouchers as lower stage of communism and not using vouchers as higher stage of communism.


yep. and in a system using vouchers of various types there will inevitably be struggle over this which none of us can say much about until we get there


We can say something about it. It is already clear that the more voucher types you have, the more annoying they get to use. It doesn't take a crystal ball to see that a population used to money will not like having to use half a dozen voucher types. People would be more receptive of such an idea after living through a period of having to use even more convoluted rationing books; but they would be more receptive still of a single type and it doesn't look like another ration-book phase will happen. We already have (bad) experience with personal rationing systems from the two world wars. Even if humanity had never made that experience, purely mathematical constructs would point out some of the inadequacies. Dividing up consumption into types with distinct point systems is an ugly idea.


You can just make it digital and create an app for that. It will remove most of the problems with accounting problems on worker's part


>europe heading towards an energy and food crisis
>it doesn't look like another ration-book phase will happen
anyway yes any kind of remuneration scheme is annoying and the more of them we need to use the more annoying it will be, for all parties involved


Well first of all, you would have to give me a good reason for doing that. I don't find the idea of a rationing book attractive, putting it on a smartphone doesn't make it much better.

Continuing with the simple model, another scenario: Let's split the 100 different items into 4 groups of 25 with their own distinct points, give the consumer 3 points in each group. Giving instead the consumer a single system with 12 points allows for over 37 times as many combinations (which of course includes all the combinations from the other scenario).

Having separate point systems screws with analyzing demand data since you only compare demand for stuff in the same consumption-point category. If my rationing book allows me to get more units of something in category A I don't care for and fewer units of something in category B than I would like, I may make use of my allowance to get more from category A in order to receive more from category B through barter. Since barter is tedious this is an inefficient use of my time compared to just having one system. And again this screws with the analysis of demand data.

These avoidable problems people can already figure out just by thinking about it and humanity already has a history of running into these problems over and over and over. Where is your justification for doing it again?

Why is it even proposed to be done? Because we have extreme differences in wealth and income and the ruling class wants to hold on to these differences. If incomes were not spread as much the issue would not arise. So there should be no reason to advocate for anything like that as a more long-term thing under socialism. Besides, putting one or two commodities into a rationing system is still much more simple than the ration book from war time.


>Besides, putting one or two commodities into a rationing system is still much more simple than the ration book from war time.
sure. but what goods this is itself depends on whether the related demand constraints prevent feasibility. perhaps a set of strategic climate planning inequalities could be useful for this? some of this stuff can be figured out using national accounts, supply/use tables etc


As already stated, physical limits are part of the planning process irrespective of whether they are represented in the personal budget system. Either give an argument why you would want multiple vouchers or stop saying that you want them.


>As already stated, physical limits are part of the planning process
yeah that's what I said
I guess we're mostly in agreement


could you guys please read a stochastic processes book or something? quote unquote "planning" is not a panacea, and mostly solves problems that are upstream of any mathematical calculation.

you're not going to plan an economy in a chan thread so stop talking about "how to labor voucher????", basically any system that can compute integrals and branch on a 0 is at least equivalent to a market, you don't need to pretend that planning an economy isn't a gigantic computational engineering problem

we can solve those now - it's literally fine to just say that


That wasn't what he said though.

He said he clocked the ALICE supercomputer as being not significantly faster than other supercomputers because it was built with a heavy emphasis on the processing while the data bus wasn't fast enough to properly take advantage of what it could hypothetically do on certain tasks.
It was an argument that left me very confused because high capacity for processing doesn't always need high data throughput rates and it's entirely possible that it required alternative data layouts than what he was running through it.


tfw no cute russian cybersocialist gf


>Was there any follow-up on Cockshott contacting those socialist students in Chile?
What are you referencing?



this legitimately made me want to kill myself, what stupid bullshit pseudoeconomics do I have to write that are loosely based enough on control theory or systems engineering so that I will have a qt3.14 like this to say "privyet mey I eenterveew yoo for mai blogg"


it is kliman, the wojak is quoting "Reclaiming Marx's "Capital"


the fuck are you even trying to say retard?


hes just pissed off that a QT girl likes cockshott who he hates


Making a big simplification by leaving out accumulation and production and only thinking about allocation of a fixed pile of already produced things and a fixed set of actors with fixed budgets of play money to acquire those things, one can think of a simple idealized market scheme as having these aspect:

asynchronous: Things are not allocated all at once.
different budgets: Actors have budgets of different size.
two-way flexible prices: Prices are flexible and the price of a thing can move in both directions.
no local price: Obtaining X units of a thing has only one price at one moment throughout the entire system
no rebates & no progressive prices: Obtaining X units of a thing has the cost X times the unit price.
no non-budgetary restriction: Budget questions aside, actors have access to the same things. No need for some special license to be allowed to buy X.

One could built an allocation procedure by following all these aspects or doing exactly the opposite, but I see no logical necessity for having only these two ways of combining the aspects. I believe any combination is possible, though when considering system robustness the set may be smaller. I think following the three "no"-rules is good to reduce the allure of black markets. I'm intuitively attracted to choosing synchronizing instead of asynchronous allocation, but that requires that actors state far more preferences then just buying orders. It looks workable if we use only additive utilities (a thing is worth whatever number of points you assign, no matter what else you get) and sub-additive utilities for multiple units of the same thing (meaning the additional utility of getting more units falls). But if we allow statements about super-additive utilities (things are worth more to you in particular sets than the sum of the parts) we will have a combinatorial explosion, no? Maybe some popular sets should just be prepackaged and only offered in set form as a hacky solution if super-additive utilities cannot be directly expressed inside the procedure.


>pol kokshot
10/10 transliteration


If you need a good job to stop being poor, major in CS, full stop. its got plenty of math content anyway. Even if you have a low GPA in CS you can still get a job or have a way better chance of getting one than a non CS degree. Then get into tech


not a transliteration



honestly you don't even need the degree, there's a huge number of software jobs that will accept a bootcamp or online cert as training as long as you pass the technical interview (literally just have to grind leetcode for ~1 month). almost all software jobs are complete bullshit anyway and you'll just be slinging javascript for some shovelware ad product and a lot of people in the industry are just completely resigned to that so the standards are basically just 1) pass the technical 2) have citizenship for a six figure "frontend dev" role at some shitco startup



I am doing data science in datacamp, hopefully I can get into it with this strategy



I have seen it work many times, we are all going to make it


File: 1665065156920.jpg (483.81 KB, 1600x1011, bg.jpg)


>honestly you don't even need the degree, there's a huge number of software jobs that will accept a bootcamp or online cert as training

hard disagree. This isn't 2013. The fact is most jobs in software (esp. corporate 9-5's) require that piece of paper. There's a glut of entry level CS grads who can code and an even bigger one of those who can't.

All these stories of the guy who did a bootcamp then got hired at google for $200,000+ a year with "no degree" turn out to be total bullshit most of the time. I remember watching a documentary about IBM hiring people without degrees for "software engineering" positions, some feelgood piece about diversifying hires by not requiring college. A guy who was titled as a "software engineer" in the story? Looked up his linkedin and he's actually a technical support engineer at IBM, literally glorified helpdesk.

Thats the thing about these stories, they always turn out to be BS. Yes in the early 2010s when tech was blowing up maybe someone could do a bootcamp and get a job but nowdays that shit is over.

The story of the guy who got the 200k job from a bootcamp, it always turns out that no, they did not have "no degree", they did have a degree just not in computer science, maybe in math, or some other type of engineering. And it turns out that no, they did not get hired as a "software engineer" but some sort of low level IT ops role.

If you don't believe me browse any job board looking for developer positions and you will clearly see the vast majority either require or prefer degrees. If you go the self taught or bootcamp route you will immediately preclude 70%+ of all potential jobs. This isn't fair but it is what it is and the bootcamp meme needs to die esp. when some bootcamps are charging as much as a 4 year degree now (at least if you only could tuition) for a vastly inferior product.


I do not have a degree and i never had a problem finding a job. Thing is, programmers job are more often the not an employee market, meaning that not much ocmpetition going on, so many copmpanies (with some exceptions like Google and other FAANG shit) are kinda desperate that they wouldn't blink about you having no degree. But you are right that more often than not some "bootcamp" wouldn't cut it if you don't have a degree, you would need an ectual job experience for companies to ignore your lack of formal credentials.



yeah this is just not correct, the software labor market still has far, far more openings than qualified engineers. two out of eight people on my last team at a softbank funded shitco startup were bootcamp grads making ~160k. you just have to target companies outside of the enterprise sphere and have at least a minimal portfolio on github


Weird, I tend to work more in the enterprise/dilbert sphere of development (fintech, etc.) and there is definitely a requirement for degrees.

>the software labor market still has far, far more openings than qualified engineers

most of those openings are for people with 3-10 years of experience and specific niche skills like being a kafka or enterprise integration expert or something, not new grads. Turning out more bootcamp grads or even college grads wont address that.


Definitely depends on the sphere. Like game devs are infamously fucked compared to others. I am a backed web dev, for example.

But yes, mostly that applies to experienced workers.


>far more openings than qualified engineers
the word qualified is doing some heavy lifting there


what would a leftist computer programming look like?


File: 1665236728061.gif (1.42 MB, 450x258, Coding.gif)

probably the same as today but since it wouldn't be rushed for profit startup style, with a greater emphasis on good architecture, simplicity, maintainability, testability and low technical debt throughout its lifecycle.

All software would be free and open source and auditable by the public, so far as its practical to do so.

Under socialism software would be free from the constraints of capitalism and would greatly improve in quality. Video games would be entirely free to play with no microtransactions. Social media would be made to be non addicting instead of the recommender system driven dopamine-drip hellscape that currently exists. It would encourage and be engineered to produce pro-social behavior by algorithmically rewarding positivity and civil discussion instead of outrage bait and fbi.gov. It would encourage body positivity and anti racism by showing a variety of people in the feed instead of making teen girls feel bad due to showing only thin pretty white girls. It would prioritize not fucking up young peoples mental health.

Programming Languages and tools would not be based on corporations like Oracle/Microsoft, but on industrial associations of programmers like a DeLeonist council, which would incubate/steward projects like the apache foundation does now. All critical infrastructure software would be done this way.


File: 1665241973547.png (181.58 KB, 1186x1080, jYwLWMU.png)

>what would a leftist computer programming look like?


replace entire standard libraries with ASIC hardware thanks


wdym? like qsort() in hardware?


the whole point of software is that its cheaper than doing this lol. The only reason to replace std libraries with hardware is for speed.


basically just code in scale
>Moneyless because there are no jobs with it.


The joke is ruined now


yeah you fucked it up.

Anyway there are probably a few scala jobs but its true that the vast majority of corporate jobs out there use Java/C#/Python/Javascript/Typescript. Maybe some C/C++ for embedded and video games. Every functional language combined has a probably single digit share of programming jobs.

The most popular functional language according to TIOBE is common lisp at #30 and thats actually multi paradigm not just functional.

Scala is #33


• Value can mean the “use value”. It is how useful the good is a to a human being. This is rather subjective and difficult, if not impossible to quantify.

• Value can mean the “exchange value”. It is the rate at which the good can be exchanged for another good.

• A commodity is a good that is, or in the future will be, exchanged. It has both a use value (otherwise nobody will want to exchange for it, since it would be useless), and an exchange value.

• There are goods that have no exchange value: For example, if a mother bakes a cake for her children from dough, the cake itself is not baked for the purpose of exchanging it for another commodity. The cake is purely made for consumption within the family. Therefore, the cake has no exchange value, but it does have a use value. The cake is not a commodity in this case.

• All commodities are goods, but not all goods are commodities, since not all goods are made for the intent of exchange. In a hunter-gatherer society where no trade exists, no commodities would exist – everything is made for personal consumption. However, in capitalism, most things are made not for immediate personal consumption, but for trade. Under capitalism, most goods indeed are commodities, since most goods are produced for the intent of exchanging it on a market. Capitalism is the mode of production where commodity production has reached its thus-far historical peak.

• While the distinction between “use value” and “exchange value” is important to understand which goods are commodities and which goods are not, this distinction should not be overemphasized: When we speak of capitalism, we must speak of commodities first and foremost, since these are the most common forms of goods. When we speak of value in capitalism, we are therefore speaking of the value of commodities. But: It would be wrong to split up the “value” of a commodity into exchange-value and use-value. The value of commodities is always both. If it has no use-value, it cannot be a commodity. If it has no exchange-value, it cannot be a commodity. The value of commodities is a unitary concept, not a dualistic one. The value of commodities is always both use-value and exchange-value combined.

• “Value” under capitalism expresses the relation between the commodity as a “thing” on the one hand and between human beings on the other. If all humans died from one day to the other, how could a Ferrari have any value? If all humans disappeared from one day to the next, all former commodities of the world would now just be mere objects, without any value whatsoever. For commodities to have any value, humans must exist. For value to have any meaning, goods and commodities must exist. Value is part of both the exterior world of objects and in the interior world of minds, both in the world of objects and in the world of humans, always these two sides at once: Value is a unitary concept.

• Value is what expresses the connection between humans and objects – the exchange-value being the social connection, the use value being the private connection we humans have towards these things, or, in the case of capitalism, towards these commodities.

• The exchange-value determines how the commodity-as-an-object forms a social relationship amongst individuals; the use-value determines the relationship between the subjectivity of the inside world of human desires and the physicality of the outside world of objects.

• Value is neither based purely on the extrinsic nature of a commodity (such as: the amount of other commodities you can exchange your commodity for), nor is it based purely on the intrinsic nature of a commodity (such as: how much energy or time it took to produce this commodity). Value has both an extrinsic and an intrinsic moment. The value of commodities is of course a social relation (since a commodity must by definition exist for the intent of trade), but it also has a natural component to it (the energy, the time needed on average and so on).

• Value is how society considers the material commodity. However, society and commodity are always bound by the laws of nature, such as the laws of energy, physics, chemistry, geology or the laws of time, aging and mortality. Therefore, just as much as society shapes its relation to the commodity, the commodity’s outside circumstances shape their relation to society as well.

• Therefore, in capitalism, a material product of concrete human labor acquires social character through value. Value is then the objective social process that gives the material product of concrete human labor the dimension of sociality, transforming it from a mere object to a commodity.

• Value can therefore be described as a social algorithm.

• What is an algorithm? An algorithm is an abstract logical procedure consisting of an ordinate and finite sequence of successive steps needed to solve a problem. For example, the sets of rules that allow for the four elementary arithmetic operations (addition, subtraction, multiplication, division) to occur are algorithms. Allow me to elaborate: For us, an addition of two numbers seems very logical, if not intuitive. You can take one object and another object, and now you have two objects. What we observe in these operations is the external form of the relationship between two numbers: We know the numbers, we know the arithmetic sign of the operation (here: a “+”) and its final result (the sum). However, we are not able to observe the logical procedure that leads to the correct result, which remains hidden and invisible, abstract inside our mind. Unless we are logicians or mathematicians, we are normally not even fully aware of the ordered set of logical rules that led us to solve the operation. Most people in this world can add two numbers, but how many people in this world are aware of the Peano axioms, which are used to formalize natural-number arithmetic? How many humans could axiomatize arithmetic (like the addition of two numbers) from the more basic facts of successor operation and inductions? Presumably not many. Yet, even though most of us do not know the exact nature of the logical steps behind adding numbers, the sequence of these logical steps really exists and produces factual results. It constitutes the algorithm of the arithmetic operation.

• Similarly, inside a market exchange, there is an algorithm that determines an equivalence-relation between commodities. However, unlike the algorithms behind elementary arithmetic, this algorithm exists outside our minds, outside the minds of sellers and buyers – it is not a pure act of human thought, but a generative social structure that most current-day humans were born into, that exists independent of their will. The value of commodities is, like the algorithm behind the addition of numbers, something that just “works” and produces reliable results – yet it is, unlike the purely mind-based arithmetic algorithm, a social algorithm, a real abstraction. Value is the algorithm operating within the social relation of the market exchange, which allows us to determine the equivalence-relation transforming qualitatively different commodities into quantitatively identical exchange-values. Value is what allows us to trade 1 kilogram of coffee for a given amount of money – qualitatively different objects, but quantitatively equivalent within the sphere of the market.

• Value is the set of successive steps leading to the determination of the exchange-value, which in turn is the formal expression of objectified abstract labor constituting the social substance of a commodity.

• Value is therefore the social algorithm of market-equivalence between commodities and between the individual concrete labors that produced them. The social process of value determines simultaneously the equivalence between two commodities (as a market-exchange-ratio) and as an objectified quantity of abstract labor. Or, in other words: A given market-exchange-ratio and a given quantity of objectified abstract labor are the mutually equivalent results of the social algorithm of value, expressed in the form of a given exchange-value. What the social algorithm of value adds to the material and merely bodily form of objects-of-use is the social form of exchange-value as the equivalence between market-exchange-ratio with other commodities and quantity of objectified abstract labor.

• In formal terms: If there is a set of commodities C, composed of the three commodity types x, y, and z, there is the following equivalence-relation, denoted by the symbol ~:

1. Reflexivity: x ~ x
Example: 10 grams of sugar = 10 grams of sugar

2. Symmetry: if x ~ y, then also y ~ x
Example: If 10 grams of sugar cost 0.10€,
then 0.10€ can buy 10 grams of sugar.

3. Transitivity: if x ~ y and y ~ z, then also x ~ z, so in total: x ~ y ~ z
Example: If 10 grams of sugar cost 0.10€,
then 0.10€ can buy you 10 grams of sugar.
If 0.10€ can buy you one bubblegum, then one bubblegum costs 0.10€.
Therefore, 10 grams of sugar are worth one bubblegum.

Both 10 grams of sugar and a bubblegum are worth 0.10€. 0.10€ can buy you both 10 grams of sugar and a bubblegum.

• The social algorithm of value is the set of real steps operating within the act of market exchange that puts the commodities (sugar, bubblegum and money) in an equivalence relation, where these formal properties are verified. [Note that money is now no longer based on gold-standard, which would make it commodity-money without a doubt, but now based on fiat-currency, which makes this relation more complex]

• The commodity has a dual-nature, it has both use-value and value, the latter is expressed in the form of a determined exchange-value. It is important the stress the logical difference between “value” and “exchange-value” here: Between the two concepts, there is a relation similar to that existing between the sum as an algorithmic operation of calculation and the sum as a result of this operation. We easily recognize the sum (= the result of an addition), but we do not easily understand the complex arithmetic algorithm behind this addition, the logical steps that lead one number to be added with another number to come up with a third number. Similarly, the exchange-value is the visible result of an algorithm, obvious to the world, while the “value” as a procedure within the market-exchange is the social algorithm leading to this result, a much more complex concept. When we refer to an already determined magnitude of value, such as that of a given commodity, what we are actually referring to is the social form in which we express value, which is the “exchange-value”. Vice-versa, when we intend to consider the origins of the already determined exchange-value to find out from where it comes out, we refer to “value” as the social algorithm, the sequenced steps leading to this result.

• The definition of “value as a social algorithm” contrasts with the idea of value usually presented in Marxist debate. Usually, value is conceived as a “substantial concept”, with the discussion being about what the substance of value consists of and how it is measured, be it labor-time embodied in production (the traditional productivist approach) or money realized in circulation (the newly emerging Western approach).

• Instead, value should be conceived as a processual concept, a social algorithm, a generative relational structure, representing an abstract procedure (the social algorithm) deriving from a real social process (labor), which continuously transforms abstract labor from potentiality into actuality, giving it an objective, a specific and a quantitatively determined social form as exchange-value.

• Value in capitalism is unitary (exchange-value and use-value always combined), but its substance (abstract labor) is dualistic, and its social form (exchange value as the end-result of the social algorithm) is also dualistic. Abstract labor is dualistic in that it reflects the social labor necessary in production and the social labor necessary to satisfy social needs; the social form of value, i.e. exchange-value, is dual too in that it simultaneously represents a given market-exchange-ratio and a given quantity of objectified abstract labor. The dualism of value is not between substance and form, but inside each of them. Value is the social process that momentarily resolves the dualisms inherent in its substance (abstract labor) and its form (exchange-value), which continuously re-appear in the commodities produced and exchanged on the market, placing them in a relation of equivalence. As a transformative process, value is neither object nor subject, but both things together in a social algorithm, because it is what transforms the subjectivity of its substance as abstract labor into the objectivity of its form as exchange-value.

• Marx’s concept of value is different from the Hegelian concept of Absolute Spirit, since it is not a transcendent subject from which social reality derives and to which it continuously returns, but, on the contrary: Value is an immanent structure that itself is a product of the immediate capitalist social reality. As real abstractions, the algorithm of value is the unconscious and spontaneous result of the concrete and material social practices of a multitude of individuals, the fruit of the myriad of exchange activities that take place at every moment on the market.

• Value is the procedure, the algorithm that converts the concrete, immediate and actual product of human labor into a commodity. Just as language is the code that transforms the sounds the human mouth can produce into socially communicable concepts, value is the code that transforms individual labor into social labor. This discovery is what allows Marx and us to find the key to decipher the social hieroglyphic of the commodity.

• Value has a relational, a processual and a dynamic nature, which leads to it constantly expanding, “endowed with a motion of its own”. When the social algorithm of value infects its own substance – abstract labor – it proceeds to establish the capitalist social relation of wage labor, the purchase and sale of labor power, and now becomes capital, self-valorizing value. “Value therefore now becomes value in process, money in process, and as such, capital”.

• Value exists as capital – the social algorithm of value goes beyond mere social production and reproduction, and now invades with its abstract code both human and non-human nature. This expansion does not have a teleological character, but remains a spontaneous, socially objective process without a planning subject. The search for profit is the conscious aim of capitalists, the same way that organisms instinctively pursue the goal of survival.

• Value is what transforms mere objects of use for humans into commodities standing in social relation between them. Value is not a static concept, but rather a process, a social algorithm.

• Algorithms were not defined in the times of Marx. However, Marx, in his Mathematical Manuscripts, actually foreshadowed the notion of algorithms when he tried to interpret differential calculus. The mathematical procedure developed by Marx shows how it is possible to obtain from a purely abstract and symbolic process, such as the algorithm of derivation, a real and factual result, such as the function derivative, which is actually operating in nature and whose discovery has allowed important practical applications in a wide range of techniques. The concept of value, elaborated by Marx in the economic works of his maturity, presents strong logical analogies with the symbolic operator of derivation in their common structure of algorithm – a purely abstract procedure that produces factual results.

• Value is as a social algorithm operating in market exchange, acting as a symbolic operator of social equivalence between commodities. As operator, value consists of a procedure, an ordered and finite set of steps necessary to achieve a given result. As symbolic operator, value is an abstract procedure that applies in commodity market exchange to content or substance logically distinct from itself. As operator of equivalence, value determines a quantitative relation, requiring that the substance to which it applies is commensurable and qualitatively identical between commodities, differing among them only quantitatively. As social operator, value has a purely social substance, in the sense that it is independent of any physical, natural, and material features of the commodities. As Marx points out, the only possible common feature that meets the criteria of quantitative social equivalence is that the commodities are all the product of human labor. Labor is therefore the substance to which the social algorithm of value applies.

• Abstract labor is the substance of value, and has a dual nature.

• Exchange value is the form of value, and has a dual nature.

• Value is therefore measured dually: In money and in labor-time.

• In term, it must be possible to express labor-time in money, which would of course also mean that money can be expressed in labor-time.

• Unequal exchange is a situation in which value in production (labor time) is different from value in circulation (money).


I work in scala. Its a corporate job. Scala is a nice language but it tends to be used as a nicer java, and the corporate goals also bloat the code massively. It doesn't help if your project was started by scala noobs like mine. We suffer from bad decisions taken years ago. Scala is worthless if you don't take functional learnings seriously and do clean code.


isnt that partly scala's fault for allowing OOP in the first place? Other FP languages like Clojure are way more "purely" functional. Not to mention Haskell


This is a long effort post, I have some criticisms.

You are using the word "subjective" wrong. "Subjective" is not the antonym of "objective". Subjective means "as a subject". In capitalism a subject simply is an agent that can buy and sell commodities or capital goods in a market. It has nothing to do with any personal characteristics or what goes on in the brain of somebody.
I suspect that what you really meant to say is something like "personal preference"

The second criticism that i have is that you are not fully de-mystifying the money-veil, with this dual nature of value stuff.
If you make a fully materialist analysis of the economy, you should be able to fully describe it without relying on any idealist economic concepts.

I'm not sure unequal exchange is really a useful theoretical tool, because it implies that there could be equal exchange, but markets intrinsically create minority wealth concentrations, that rule out that possibility.


Aim for law school. Count on failing out in your first year if you don't line up a firm to direct hire you >>before<< you even take the LSAT let alone the BAR.

'We will show the people the structure of society, so they can tear it down with their bare hands.'
-V.I. Lenin


can you add Ian Wright and his blog in the next OP? he's way underrated


also I'm kinda ashamed I didn't read this thread before. I'm a Cockshott-Ian Wright fan, but I just found out about Tomas Härdin and cibcom.org and I'm having an absolute blast


File: 1665298640989.mp4 (973.61 KB, 540x960, VID-20221006-WA0016.mp4)

Yeah, a little bit. Scala 3 fixes some issues. OOP + functional programming is not that bad actually. You have to learn how to organize code and good patterns. Otherwise you end up with a java codebase.
>pussy couldn't handle a little car bombing here and there.


>The most popular functional language according to TIOBE is common lisp at #30
there are 100% far more jobs in any functional programming language than there are in Common Lisp. the amount of CL jobs I see per year is like, less than 10


sure, I'll add Ian Wright to my text file for the OP. I recently watched that talk on youtube where he presents capital as a real god with real powers, very interesting
also cibcom seem to be making the rounds in the spanish speaking world


Tiobe tends to overindex embedded & mainframe developers compared with startups/faang/etc. I wouldn't be surprised if there's alot of legacy CL out there.

as for other FP languages, other programming language indexes also have them slow, if slightly higher:


The most popular FP language on that one is scala, but its still less than 1% market share.

Even rust/golang are like 1-2%.

Doesn't look like most jobs will be switching off your standard OOP langs anytime soon.

Maybe typescript will overtake vanilla javascript due to just being objectively better though.


Recently written critique of Towards a New Socialism by Conrad Hamilton


Pretty interesting, I need to give it a proper read when I have time.



TLDR: Ultraleft garbage dressed up in pedantic verbiage. "lol Marx was a utopian for writing Gothakritik but we're more Marxist than Marx".

It's basically the same idiotic position as the boridigists on leftypol took back in 2017 saying that labor vouchers still don't fully abolish value.

Its a guy complaining that cockshott preserves a consumer goods market and adjust labor vouchers according to supply and demand. They basically claim TNS is "ricardian" socialism reborn and that their distinction between different grades of worker productivity preserves class. They actually also go even further and say that Marx himself made a mistake in critique of the gotha programme.

>If Cockshott and Cottrell’s “labour tokens” are by contrast possible, this is for a specific reason: because they reprise the value/price distinction. But the irony is that by doing so they sever them from their productive basis, in a way that risks re-mystifying relations of production.

As for their solution? None given because we can't plan shit in advance because…. its somehow un Marxist.

If you actually read between the lines it basically implies any plan for socialism made in advance of it happening is unmarxist and ultimately "ricardian", combined with a borderline anti-communist reading of Marx himself which implies than any form of labor time accounting condemns a society to "incoherence" and authoritarianism, and that Marx was at least indirectly responsible for the supposed authoritarianism and wrongs of 20th century socialism (ex: the USSR). This surprisingly dovetails pretty closely with right wing critiques of communism like the austrians do.

This has reinforced my idea that communization/dauve-tards are basically just making an intellectualized version of anarcho-communism, but its actually worse than that because at least ancoms propose we do full communism now while leftcoms just critique everything outside their small clique of pseuds while providing no solutions or even attempts at one.

Because making plans beyond the next 5 minutes is utopian but imagining you can run a complex industrialized technological society in a radically different way just based on spontaneousness or whatever the fuck a bunch of philosophy-pseud mathlets can think of on the spot, isn't utopian… somehow.

There's a weird disconnect between people who claim that Marx's historical method allows Marx and Marxists to predict the inevitable downfall of capitalism and the emergence of an alternative system, that there exist 'laws of motion of history', that allow us to predict this with newtonian exactitude, but when it comes to the implementation details, that's apparently the boundary at which the immortal science no longer applies and its anything goes.

The fact is the pseudo-hayekan line of its so complex we can't possibly plan it in advance is not only wrong, but the exact opposite of what is true. The fact that a transition from a capitalist to non capitalist society is complicated means not only CAN it be planned, but it MUST be planned.

Guess that's the difference between an engineer and a philosopher.


File: 1665605072054.png (107.61 KB, 1311x271, ARISTO.png)


Labour aristocrat stench .


It's more about voucher schemes in general. The writing is verbose yet imprecise.
<the progressive replacement of workers by machines will — in so far as it is primarily the former that creates value
Primarily? I though in the system of Marx it is only human labor that does that.
The author says about the scheme with the labour vouchers in Critique of the Gotha Program:
<the worker receives the full “value,” or “labor cost,” for his toil
…Yes, after some necessary deductions. The author does mentions the caveat later, making this passage both redundant and misleading.
<Marx oscillates in his work between, on one side, a substantial, quasi-Ricardian conception of value according to which the value of goods brought to market is trans-historically correspondent with the “socially necessary labor time” required to create them (that is, the industrial average time required for their creation (knitting a sweater at home doesn’t count).
The author writes here as if the state of technology were something transhistorical.
<While Marx acknowledges the presence of instances where capitalists acquire access to exceptional technological breakthroughs, thereby acquiring “super profits,” these do not alter the socially necessary labor time required to create a given product until they become standardized.
Not quite, the new tech already alters it while used only in a minority of products.
<value emerges through a socially mediated process of production and exchange
I would say: Value becomes more visible in exchange, but it precedes it.
Then comes a passage where the author echoes Austrian economists by doubting labor time can be adequately estimated and quotes some other "communist" wankers who do likewise.
The author knows that we won't avoid all supply shocks nor know demand exactly, and an obvious way to deal with that is allowing the price of consumer goods to vary (as proposed in TANS).
<The problem is that, the moment the labor time goods exchange for is adjusted beyond the actual labor required for their creation so as to accommodate shortages of supply, this ceases to be an economy based on exchanges of equivalent labor times.
Think about a product basket, not a single product. Some things get more expensive, other things get cheaper. In the big picture one hour of consumption points gets one hour of labor time. The heavy modification of the exchange of equivalents lies elsewhere: in the deductions before you get your vouchers.
<We could add to this that it would be important that workers in such a society be — as Marx states — “equally productive,” to avoid the de facto subsidization of sloth…
It's possible that people can choose to work in teams with faster or slower pace and get more or less remuneration for it (as proposed in TANS).
<It is somewhat difficult to imagine how “insurance funds” could be calculated without money.
The author suffers from micro-macro confusion. From the individual point of view it works as if your saved money turns into consumption goods in the future, like milk turning into cheese. The reality is that non-workers today are catered to by people working today. That is we as a society make deductions in the present for performing services in the present.
<Towards a New (Ricardian) Socialism
Here in the final third we get to TANS.
<Cockshott and Cotrell want to show that Marx wasn’t a utopian, that his positive propositions can really work. So in addition to discussing how many ponies an “urban commune” might be able to afford, they dutifully set out to elaborate Marx’s notion of labor certificates, revising it so that it can adapt to fluctuations of supply and demand, the differential productivity of labor, and so forth. But what they’re left with at the end is just money.
The vouchers are still not money because they do not circulate. No person becomes a capitalist, hiring employees he pays with vouchers to produce commodities he sells to make a voucher profit. This is a substantial difference, far more substantial than whining about "unjust" prices of this or that good because of supply-demand fluctuations.

The author craps on the insta-communizers in the section "Freunden Like These".


File: 1665606327047.jpg (41.85 KB, 618x597, eyeroll.jpg)

>say, a natural disaster results in a shortage of chips required for a technological device
is this supposed to be a serious gotcha? the workers at the chip plant presumably still get their vouchers while effecting repairs. this labour has to be accounted for, and the obvious solution is to account for it in the extant stock of chips, thus raising the value per chip so the sum of values remains constant, say sum(vouchers) + sum(goods) = 0. another option is to not repair the plant and reassign its workers, in which case the value of the chips is unaffected. anyone who has done accounting knows this. debit and credit across all accounts always sums to zero
the author seems to be forgetting that the value of a good is not just the cost of the good in that moment, but the cost of reproducing its productive process. the entire mass of value has to be account for, from start to finish, both "vertically" and across time.
the author also doesn't seem to be aware that nothing prevents piece wages from being used, the only requirement being that values must again sum to zero
>If the social consumption fund must serve as “buyer” and “seller” of all goods, and if society must become organize production in accord with this, there is no way that non-productive administrative costs could diminish immediately. It simply wouldn’t work.
>source: trust me bro
I will also say that "a non-administered administration" sounds precisely like what we're trying to achieve in here. that is automated administration, automatic government
>muh statism
<From capitalism, it takes the idea of differential payment for each hour worked
>what are piece wages?
>exchange is when fluctuations in value are accounted for
OK that's enough of this, I'm going to the pub
I do enjoy the author characterizing TANS as "monomaniacial" and "diabolical genius" however
I think >>1220020 puts it well that this is "pseudo-hayekan". the author is just parroting Alec Nove


didn't read, back to reddit fucking frogposter


Yeah definitely it is pseudo-Hayekian. What's so funny about it is that firms virtually never use market-like processes internally. I'm not saying it's literally impossible to use some play-money and have auctions for some resources internally, but just look around you, ask around. That's a very exotic way of running a firm. Within the firm, the partially finished products, the tools, and other resources don't circulate as commodities. As Marx wrote in Capital:
<The rule that the labour-time expended on a commodity should not exceed the amount socially necessary to produce it is one that appears, in the production of commodities in general, to be enforced from outside by the action of competition: to put it superficially,each single producer is obliged to sell his commodity at its market price. In manufacture, on the contrary, the provision of a given quantity of the product in a given period of labour is a technical law of the process of production itself.
Within the firm, the labor time is known directly and is something you plan with. Since this is true within the firms, even huge firms, why would it be a big leap of faith to say we can do that in socialism.


yeah People's Republic of Walmart says the same thing


>that their distinction between different grades of worker productivity preserves class

Well doesn't it?


That the ruling class has its origin in a stock of particularly hardworking and virtuous people and not people who got lucky or were gangsters is asinine liberal propaganda. So to read anything like capitalist class distinction into a proposal for a productivity bonus that isn't even differentiated across jobs means one is either dishonest or full of liberal brain-worms.


as I recall it's just another way of saying some people want to work part-time and we should accommodate that


Yes sure you can say that 'well the surgeon getting paid 100,000 vouchers/year and the janitor getting paid 20,000 vouchers/year don't have different relations to the MOP so there is no class divide there' might be technically true in one sense but realistically it misses the point, if you have massive income divides then you have class in the modern sense if not the Adam Smith sense.

Well if that's the case it isn't a bit deal - let me look at the article itself tho.


>[…] grades of labour would be regarded for planning purposes as “creating value” at different rates. Rates of pay would correspond to these differential productivities: grade “B” workers would receive one labour token per hour, “A” workers rather more, and “C” workers rather less. The rates of pay would have to be fixed in such proportions as to keep the total issue of labour tokens equal to the total hours worked. The exact rates of pay could be worked out automatically by computers once the number of people in each grade was known (1993, 34).

he's clearly talking about per-hour wages


>>proposal for a productivity bonus that isn't even differentiated across jobs
>well the surgeon getting paid 100,000 vouchers/year and the janitor getting paid 20,000 vouchers/
Can't you even read a short comment properly, you dumb fuck?


File: 1665612719139.jpg (20.89 KB, 600x600, lowqbait.jpg)

>That the ruling class has its origin in a stock of particularly hardworking and virtuous people
<hardworking and virtuous exploiters of the proletariat.


Well where is the source that it's just based on individual productivity?


Should skilled workers get more? The answer by TANS:
<Even if the compensation argument is an accurate reflection of reality in capitalist countries this does not mean that professional workers should obtain the same sort of differentials in a socialist system. The costs of education and training then would be borne fully by the state.
(There may be a need to compromise on that part though if skilled workers threaten to emigrate.)
Here is the idea with the grade levels in TANS:
<One way of gearing reward to effort would be an economy-wide system for the grading of labour. For instance, there could be three grades of labour, A, B and C, with B labour representing average productivity, A above average and C below average. New workers might start out as ‘B’ workers and then have their performance reviewed (either at their own initiative or at the instigation of the project for which they work) with the possibility of being regraded as A or C. Note that these grades have nothing to do with education or skill level, but are solely concerned with the worker’s productivity relative to the average for her trade or profession.

So no, TANS doesn't say surgeons should get more than janitors.


>if you have massive income divides then you have class
no you don't. it is your role in the productive process that determines your class, not income levels. else sewage workers could be a separate class if socialist society has to pay them more because no one would want to work with sewage otherwise. another example is high risk jobs like dive welding where the increased fatality rate has to be accounted for


The topic of planning in a socialist economy has always interested me, so this thread is very nice. I have a couple of questions, however, about it. Mainly:

1- I skimmed through some of the wikipedia articles in regards to the various five-year plans the USSR undertook. And while they were the reason the country became the powerhouse that it did, some of the plans didn't achieve the quotas they were meant to. What caused this? Unrealistic expectations? Insufficient productivity? And what would the solution to this potential problem be?
2- I know the ECP is meaningless at this point, but how would a cybernetically planned economy fully substitute the markets' mechanisms? Especially given that we still live in a world dictated by capitalist institutions and their apparatus.
>Inb4 infantile disorder


File: 1665628810728.pdf (82.53 KB, 197x255,

read TANs and pdf related (section 4 specifically) by Cockshott it answers alot of these questions


File: 1665628960068.pdf (401.64 KB, 197x255, 2005.01539.pdf)

anyone read THIS?

>The debate between the optimal way of allocating societal surplus (i.e. products and services) has been raging, in one form or another, practically forever; following the collapse of the Soviet Union in 1991, the market became the only legitimate form of organisation — there was no other alternative. Working within the tradition of Marx, Leontief, Kantorovich, Beer and Cockshott, we propose what we deem an automated planning system that aims to operate on unit level (e.g., factories and citizens), rather than on aggregate demand and sectors. We explain why it is both a viable and desirable alternative to current market conditions and position our solution within current societal structures. Our experiments show that it would be trivial to plan for up to 50K industrial goods and 5K final goods in commodity hardware.

Author claims to be doing decentral planning essentially


Was already mentioned in a cybernetics thread in 2020. What's decentralized about it?


Samothrakis comes at this problem from the AI side of things. I've talked to him, and he has a notion of a kind of Kantian optimization. he hasn't presented the formalism for his decentralized planning concept yet, but I think he's working on a paper about it


>decentral planning
sounds like an interesting deal to think about


>>1220053 (me)
One particularly silly thing from the article I forgot: The author briefly mentioned this reason to be against labor vouchers: black markets! Consumption points cannot be transferred to other people. Of course black markets can thrive when people use cash, which seems to be what the author wants instead. So how does the author's position make any sense?

I suppose the situation in mind at the moment of writing that passage was that prices of consumer items are fixed at labor cost and then when there is a massive discrepancy between supply and demand which could be addressed by price changes, this is refused because of ideological reasons. So some people benefit by bartering with the rare stuff. Black markets live and grow through repeated interactions, think about producing alcoholic drinks during the prohibition era. But what is described here is just a punctual random annoyance, so how could anyone make a career out of this.




File: 1665699528817.png (427.26 KB, 993x378, cockshott_meme.png)

my take is that currency and black markets emerge when there is a need for them. the correct way of arranging things emerges from contact with reality


File: 1665780087332.jpg (51.88 KB, 606x680, MoreCockshott_meme.jpg)


>facial hair
can't be cockshott, he's a beardlet


Article by Elena Veduta: Some Lessons on Planning for the Twenty-First Century from the World’s First Socialist Economy


new video by Paul Cockshott:
New Book on Planning
>This is the book that I have been working on with Allin Cottrell and Philipp Dapprich. Work for this provided the foundation for a couple of videos, but there is a lot more in the book.


File: 1666784991165.jpg (152.88 KB, 500x537, 69oy4f.jpg)


Q about harmony algorithm and the like: The base version of it only uses one production recipe per produced product and it got criticized for that. I agree that it is more realistic to have several possible recipes available for some products. However, isn't it likely that for the most products only one recipe would be selected for use by a hypothetical optimal decision maker? Does selecting one process per product (not before the optimizing attempt, but as part of it) simplify the computation?


just a reminder that you can never plan an economy it is rationally impossible and always ends up in failure
Austro libertarian market socialism is the only way


>planned economies dont work because they just dont… OK!?
weak bait 3/10


oh i have an argument if you wish to engage


File: 1666788951509.webm (1.31 MB, 720x400, max.webm)

dave zachariah and his student loke hagberg are working on precisely this. being able to have multiple ways of producing the same product, multiple inputs and multiple outputs and taking time into account. it's not public yet however


ai’m gonna teach myself math and do cybernetics better than TERF Island boomer guy


Whether a recipe can be followed and to what extent depends on its most scarce ingredient. I suppose one can for each ingredient compute a lower and upper bound of usage by looking for each product's recipe with the lowest and highest usage of it. Then, if the lower-bound usage of ingredient A is higher than the upper-bound usage of ingredient B (as percentage of available resource), then A is definitely more scarce than B. This information can be used to delete some recipes, re-compute the upper and lower bounds, delete some recipes, and so on. Not very decisive, but it's something.


File: 1666790239379.pdf (170.76 KB, 180x255, rhp_intro.pdf)

nevermind it's been published: https://github.com/lokehagberg/rhp
just use an LP solver. you can constrain the rate of change of the technological mix like section 3.1 in pdf related


Bro it's been published?????? A supply-use planning function for efficient reallocation??????? Swedish economic data is used???????????????????????? On god???


Just looked up that Loke Hagberg guy. He is working on a discussion/voting/delegating software called Flowback. I do hope something like that will become common in the future, but it looks a bit rough.

One of its features is quadratic voting, which was invented by Richard Zeckhauser and Aanund Hylland in the 1970s (contrary to what the jerkoffs at Forbes, The Economist, and Wikipedia say). This method lets you vote with points and it takes the square root. This is robust in terms of voting strategy BUT it is very vulnerable to category spam. Imagine a group uses this to allocate a budget. If a funding category gets split up into several ones, it will receive more funding. A serious drawback. I'd prefer approval voting with reweighting or cardinal voting with reweighting instead (though these suffer a bit from free-rider voting strategy).

The thing also features recursive proxy voting like LiquidFeedback, you can delegate your vote and your delegate can delegate to somebody else and so on. The experience the German Pirate party has made with that: strong concentration of voting power. People IRL don't feel strong connections to someone just because he is a friend of a friend of a friend, so this feature doesn't make sense. (What makes sense is if the user can rank proxies.)


yeah hagberg has been involved with the direct democrats (direktdemokraterna). I'm also skeptical of the square voting thing


I think delegation of voting power except on issues that you feel strongly about can work quite well. no need to complicate it with non-linear voting shenanigans
how does the thread feel about voting secrecy? personally since it is effectively impossible to guarantee with computers, I feel voting in a computerized socialist democracy should be public. voting secrecy strikes me as cowardly and liberal. proudly proclaim your chosen delegate and proudly proclaim what issues you feel strongly about


Delegation Alice to Bob and Bob votes on the issue is not the problem, the problem is longer delegation chains.

Quadratic voting etc. is another topic. The ideal is the two goals of 1. giving minorities some power for e. g. allocating a budget while a majority block still has more power and 2. to avoid vulnerability to strategic voting and strategic nomination. The pure ideal is mathematically impossible when there are more than two things, so the question is how you can approximate that. (If a budget is to be divided between just two topics, there is a solution that is both robust and gives budget-allocation power to the minority faction in proportion to its size.)


>the problem is longer delegation chains
this is why it is important that one be able to withdraw either one's support of a delegate, or one's vote on a specific issue, perhaps at some point along the chain other than the final/top delegate. like to say some regional delegate that you trust more instead


It's a standard in LiquidFeedback that people can change on the fly both who they delegate to and whether they delegate at all. Still the observable result in the German Pirate party has been extreme concentration of voting power. Perhaps there is a simple mathematical law at work here.


What do you guys think about the cybernetic hypothesis? How do you prevent cybernetic socialism from reaching the same endpoint the techno elites are driving towards, where every human interaction is quantified, modeled and happily under control?


> quantified, modeled and happily under control

this is impossible.


How so?


pls update with General Intellect Unit logo somewhere pls


What about Systemic Consensing (SK)?
I've implemented something similar in a few local orgs and it works pretty okay (also fits culturally which perhaps helps a lot too).


>year long delays with publisher who approached them
>self publishes because fuck'em



sounds like sortition would tend to counteract this kind of problem
>being taken care of by loving machines is.. le bad!


That's just average ratings AKA range voting, which is very vulnerable to people just voting at the extremes to maximize impact. Still a massive improvement over just voting with one mark and faster than voting with a runoff. Whether the ballot interface goes from some negative value to some positive value or only uses positive values or only negative ones like in the SK proposal makes no difference from a mathy analytical point of view, but it seems to have a psychological effect. The mathy analytical view usually assumes that people are voting in secret and that they feel no shame about voting in a strategic way. But suppose voting is not anonymous and people giving an extreme input are required (whether formally or because of group pressure) to say something about why they are doing it, that reduces the problem.



> but it seems to have a psychological effect.
Thanks for the reply. And yep, that's why I use it. Where I am we use secret paper based preference voting, so the "mental shift" of SK was an easy leap and had an effect. Now we also have compulsory voting, so that help even out the power voters (extremities).

I was interested in fusing Paul Cockshott's HandiVote system with it, so an "anonymised" but publicly verifiable ledger. Failing that, sticking with paper. The key thing bout this method is how it can be used locally and directly, that's the main bit.

Add in sortition for delegates based on post code population (as a proxy for wealth, so basically fewer wealthy people ever get in at all), the ability to instant recall delegates with a 2/3 vote of no confidence in their "electorate".


Bitcoin is no answer
>A short video looking into the dire effects that would ensue were bitcoin to become our standard money.


Wouldn't it be much better to start transforming this thread and the people in it into an organised reading group with an archive(site) in which summaries, insights, and translations can be posted so that it can be actually used as a source on modern economic planning by other socialists?


I'm not sure if I agree with Paul's assertion that Bitcoin has no use-value. the network itself is useful for storing short snippets of text in perpetuity. the keys themselves are information, which surely is a form of use-value. their main use is as exchange-value of course
in a similar way someone could claim fiat currency has no use-value. but I could burn paper money for heat and melt or dissolve coins for the metal in them
casperforum.org could be used for that


oh and bitcoin keys can also be used to cryptographically sign things, to set up tontines and many other useful things besides exchange


That has nothing to do with bitcoin in itself. Blockchain has (admittedly) a use value. For example, it is used inside minecraft to cryptographically verify that chat message logs haven't been tampered with. You don't need a distributed cluster of megacomputers to create a decentralized database for that though, that is useless in any meaningful sense of the word.


sure, in the sense that git is also a blockchain
the most important property of cryptocurrencies is proof-of-work, which is quite literally congealed labour. it is also congealed in an explicitly deflationary fashion, and it is not productive
I'm arguing
you are right that bitcoins themselves are nothing but exchange-value. whatever other uses the network has are incidental


>the most important property of cryptocurrencies is proof-of-work, which is quite literally congealed labour.
Sure but so are mud pies. And we all know that for things to have value it needs a use value.


means of payment can be worthless and still function just fine. this is a central property of fiat money after all. no one believes a $100 bill costs anywhere near $100 worth of labour to print
bitcoins are in demand, which means that the labour that goes into them isn't mudpies. perhaps it is more appropriate to call the labour's "social usefulness" contingent on exchange. capital considers lots of unproductive activities useful


come to think of it, surely exchange-value is a type of use-value? money is useful after all, and it is useful precisely because it can be exchanged for commodities


ancaps btfo


>come to think of it, surely exchange-value is a type of use-value? money is useful after all, and it is useful precisely because it can be exchanged for commodities
No. Use value is the use of something in its consumption. Exchanging it for something else isn't a use value of a commodity, because a use value is something inherent in the thing itself.

Fiat currency has no use value. It's just a number in a rationing system. The exchange value of 100 dollars is 100 dollars, its use value is that you can wipe your ass with it. Trying to define use value as "you can exchange it" means that its use value is not defined by its uses but by its market price, and leads to a recursive definition of "it's useful because it's worth something because it's usefull because it's worth something".

A healing like of shit doesnt get more use value just because scam artists drive up its price in an economic bubble.


Bitcoin advocates are in absolute cope right now since Bitcoin really is just the inefficient and destructive of economies and the state credit. Fascists would lose their ethno states and then some, tech bro libertarians would literally see physical capital swallowed up en mass and they'll literally be living in pods.


>a recursive definition of "it's useful because it's worth something because it's useful because it's worth something".
good point. how should we interpret the price of bitcoin then? clearly its price is tied to the cost of mining, no? on top of this we have market manipulation. should mining be interpreted as a form of rent? and when the price is manipulated this just means more rent is extracted? if the rent rises high enough then there's enough "room" to make otherwise unprofitable mining profitable, thus pushing difficulty up. perhaps mining just sets the rent for performing transactions rather than the value of the coins themselves?


>>1246589 (me)
expanding on the rent concept, this means early adopters can be compared to people who engage in land grabs. these people can later portion off pieces of the land/coins for a price much higher than they paid initially. as we know, land itself has no value, being merely a free gift of nature. only improvements to the land can have value, and no such improvement is possible with cryptocurrency coins


thoughts on this? Huibert Kwakernaak influenced Qian Xuesen who in turn was the man behind the CPC's one child policy. Kwakernaak appears to mostly be your typical Malthusian
seems the Club of Rome was influenced by Edward Goldsmith's and Robert Allen's A Blueprint for Survival, and The Limits to Growth by Meadows, Meadows and Randers
personally I question the framing that the population needs to be kept at any particular level. the lower the population the less specialization is possible


So what's the deal? New Cockblast book and no one is discussing it? Says it came out October and is already out of stock? WHAT IS HAPPNING


m8 people gotta finish reading it first, it literally just dropped


Just like capitalist economies are completely orinented towards profit, and war economies toward winning the war, a cybernetic socialist economy would be an economy orinted towards reducing the overall time people work and increasing the free time people have to actually live their lives. This is what liberation looks like in the real world, on a historical level. This is antithetical to what you claim "the technocrats" want.


people are likely still going through it.

someone update his wiki plz "published works"


Shot's fired

Bri'ish man vs Glaswegian



doesnt he just re-iterate the critiques of Nitzan and Bichler which cockshott already addressed like 8 years ago?




>doesnt he just re-iterate the critiques of Nitzan and Bichler which cockshott already addressed like 8 years ago?
that he does


I don't get the critique. Completely went over my head. Watched it several times.
What's the response?


How do you prevent labor time overflowing to the freetime, in a way cybernetic-capitalist free time consists of self-programming, preparing for the next shift including travel times, and chemically numbing the biological resistance to working conditions, all of which are necessary parts of production, but are artificially separated from the "proper" working hours?


For starters, add commuting time to work time.


Union contracts.
If you're paid for 8 hours a day but you then work 10 a day outside of agreements, you either do not care much, or youre a retard.

By and large unpaid overtime shouldn't exist.


How the World Works: The Story of Human Labor from Prehistory to the Modern Day qqqqq


Critique of Mises: A video by Tomas Härdin
>Tomas is doing great work on the algorithmics of socialist calculation. Here he takes down Mises.
>Shared with his permission.
>He originally loaded it as https://youtu.be/0f-MWeJCsRs


How the fuck can I help contribute to this movement? For real though. I'm a programmer, not great at theoretical math shit but I know my way around. The fuck should I do?


it has less to do with "programming" and more to do with accounting. if it wasnt called "cyber"-socialism no one would give a shit. cockshott is the face of the movement and he is dryer than any pussy he encounters.


you could try and get a pilot project going somewhere
read up on the relevant theory
agitate for planning/in-kind calculation
where are you located?



I think what would be much more usefull is if people set up some international think tank primarily focussed on the following:
>Primary: Creating accessible educational material about economic planning for the global communist movement
>Researching algorithms and solving practical problems faced by historical movements

The programming aspect of this is pretty tiny. Full fledged implementations and micro optimizations depend wholly on the situation it is implemented for, and is better left up to be done at the moment it can be implemented.

For that educational material, we might want to implement some demo's but its not the primary thing we need I think. Educating the movement about economic planning is our primary goal, as even many communists (especially the new generation) lost trust and hope in economic planning.

It would be interesting to set up such a group but coordinating an online org and having everyone put in time (consistency is more important than volume in the aspect of an org) is hard. Idk what form of communication would fit such a org best.

>it has less to do with "programming" and more to do with accounting
"accounting" is doing a lot of work here. Economic optimisation on a large scale is primarily a computer science problem, only a small part software engineering, and a political problem. Writing down the numbers isnt really the issue.

>he is dryer than any pussy he encounters.

The fuck is that supposed to mean lmao


>I think what would be much more usefull is if people set up some international think tank primarily focussed on the following:
>>Primary: Creating accessible educational material about economic planning for the global communist movement
>>Researching algorithms and solving practical problems faced by historical movements
this is more or less what cibcom.org is doing, but for the Spanish speaking world. they've also been releasing material in English. there's also casperforum.org but that's more a research and discussion forum than an org
>The programming aspect of this is pretty tiny. Full fledged implementations and micro optimizations depend wholly on the situation it is implemented for, and is better left up to be done at the moment it can be implemented.
I agree. we've had people in these threads asking for concrete implementations, not realizing such implementations would necessarily be historically contingent. that said check out Hagberg and Zachariah's rhp project >>1239482
>For that educational material, we might want to implement some demo's but its not the primary thing we need I think. Educating the movement about economic planning is our primary goal, as even many communists (especially the new generation) lost trust and hope in economic planning.
I think most young comrades simply don't know what planning even is


Build interactive toys that demonstrate problems or neat things similar to what https://ncase.me/ is doing. For example, what happens when people all start with the same amount of money and they do a couple thousand random exchanges. Suppose people have a constant stream of the same income, the same outgoing expenses for basic costs of living and some random spending requirements on top of that because of good or bad luck and people with excess money lend to people with no money at interest. What happens over time.

There are hundreds of voting systems and not all seem to have implementations for online polls etc. and there is more to results than picking a winner. For example, when filling out forms with approval marks or ratings you get a display what gets most approval marks etc. but this could also be used to visualize other things like whether some fans of A rate B very low relative to the rest of voters or vice versa.


having online interactive simulations could actually be useful propaganda. like the Dragulescu-Yakovenko paper's simulation all prettied up could be a useful thing against market soycialists


You missed my meaning I think. After you clock out, you are on a deterministic path to come back to work tomorrow, and are required to plan your "free" time accordingly. Cybernetic capitalism tries to optimise the workers "free" time for best performance in work by both providing and limiting activities available. Best performance here doesn't mean a healthy happy human being in most positions since the structure requires stability. It needs to handicap and guilt trip cogs to stay in place to make itself manageable, legible, and calculable.


then idk what the fuck your point is.


I find cybernetic socialism fascinating idea, I'm just trying to figure out how unwanted optimisation artifacts are addressed in the theory. If you want to make an optimized system, it first and foremost optimizes itself to be optimizable. As cybernetic capitalism has proven, all the weirdness of humanity makes calculations difficult and hence it expends huge amount of energy to simplify people. Advertising algorithms cluster people and reel them tighter by feeding them stuff from the artificial category assigned to them, just by the necessity of data processing technology.

You guys are basically dreaming of an AI that rules over humanity. You gotta have some sort of solution to the alignment problem.


anon why would we give a shit what people do in their free time? one consistent demand from the labour movement is a shorter work week
>alignment problem


>anon why would we give a shit what people do in their free time?
It's not intentional, it's an emergent property. Are you familiar of the concept of invisible labor? Cyberfeminists pointed out how information technology spreads it's domain from women to everyone. Think about how the "free" time of so many people today is just signal processing by liking, retweeting, sharing, media consumption and content production. Even people not so terminally online are mostly capable of talking about things orginating from online, feeding back to the cyberreality through exit nodes after their short journeys filtered through the meatspace.


>Think about how the "free" time of so many people today is just signal processing by liking, retweeting, sharing, media consumption and content production
those are all things related to the superstructure


It's also a byproduct of optimization algorithms. Cybernetic superstructure needs data and feedback to function. More there is, better it can theoretically function. Therefore cybernetic systems tend to converge towards extracting maximum amount of data and feedback from people.


Gathering feedback is not free of cost so you are wrong about that.


Was Marx right about profits
>A summary of Marx's theory of the rate of profit and how it compared to what happened in the late 19th century.
Cockshott says not to waste time on Hegel yet makes use of immanent critique. Curious.


>doesnt he just re-iterate the critiques of Nitzan and Bichler which cockshott already addressed like 8 years ago?
Can someone address this? Why is what unlearning economics guy said wrong? The way I see it, the LTV is self-evident in the simple sense, and then when extrapolated it has been proven by others too like Anwar Shaikh. I didn't understand his critique.


File: 1667908637791.png (Spoiler Image, 12.65 KB, 420x422, Soybernetic Malding.png)


>You guys are basically dreaming of an AI that rules over humanity.
Maybe you should read some of the books before spewing bs


>this is more or less what cibcom.org is doing, but for the Spanish speaking world. they've also been releasing material in English
It would be interesting to see if there can be some english language branch set up.
But i dont really have time to commit to such a thing in the coming year do to other roles i already have.


>It would be interesting to see if there can be some english language branch set up.
not a bad idea


No free online version yet.


New Dickblast interview by Doug Lain


Can we raise funds for a new camera and microphone for Dickblast? Pls ffs


File: 1669248388836.jpg (6.04 KB, 154x195, who dis.jpg)


The Future.

Dickblast Z.


File: 1669249365342.png (258.35 KB, 644x375, sip.png)

wat flava he drinkin /cybersoc/?


I would not be surprised if he still uses DSL internet too.


shitty internet might be out of his control
shitty audio is not
I wonder the same thing. there's some german beer in the background


And yet he's already replaced his mic IIRC. Doesnt know how to find a good mic apparently. You would literally have to buy one for him and ship it to his address.


For real, he probably has the same webcam I had in 2010


new Cockshott lore just dropped
>TANS was originally going to be called Post Soviet Communism


I kinda find it ironic that a boomer who can't fix his computer teaches people about cybernetic socialism.


thanks anon


Micro economics will be useful to learn the flaws of capitalism from the viewpoint of the individual.

Push your professor on 'marginal utility' as hard as you can. Constantly tell the class 'Ceteris Parabus' is an assumption and required precondition for the math to work. Remind everyone that we are presuming perfectly rational actors with perfect information for every inch of their theory. Finally, present every question from the point of view of the employer not the employee. The professor must concede to the most inhuman shit or abandon the position.

Macro economics will pretend that GDP was invented the day before God said let their be light. Ask about when GDP was coined as a term and whether a country with zero Agricultural output and zero Manufacturing output has any GDP. Then ask whether agricultural subsidies can be substituted for agricultural output. Carbon tax flows neatly from this and you have your final paper already written. Demand to learn about derivatives and insist that an economy possessed only of derivatives is a thing and is the optimal strategy. From here define imperialism for the class.


File: 1669421473629.mp4 (4.34 MB, 1280x720, average proof fan.mp4)

anon to a mathematician it is enough to know that a solution exists


Skip undergraduate econ other than a minor or whatever the minimum classes needed for grad school in econ would be.

If you really want become a great political economist, double major in MATHEMATICS.

You need MATH to be good at grad school in either econ or CS, or both.

Take a probability and a statistics class if you can as well.


Abolish commodities or abolish value
>This video addresses persistent misunderstandings about the relationship between value and exchange value.


you find it fascinating but couldnt be bothered to read any of the books ?




File: 1669653498634.pdf (518.9 KB, 180x255, commentsonchinamodel.pdf)

has anyone read this paper by cockshott on socialism with chinese characteristics??


looks like he's dunking on some Chinese chauvinist
>But what does he mean by this?
he did the meme


So, is the Patnaik's book worth reading after /ourguy/ trashed it?


Yes, because then you can dunk on him too.


How many ways to produce a thing are known? If we don't glue ourselves mentally to a particular physical object, if we instead think of production of useful effects, we have to admit there are always alternative recipes, because there are many ways of gaining happiness. And there is joint production everywhere. Imagine a factory that only puts out thousands of units of exactly one type of Swiss army knife. The factory uses one process for one physical output, but it is joint production of useful effects.

It is important to avoid a one-sided assessing of production from only a naively physical point of view, in so many tons of steel and cement and so on. While this perspective helps against pitfalls of money-obsessed thinking (a landlord claims to be productive because of how much money he "makes"), it has its own pitfalls. Defining a thing in terms of specific physical materials and how these materials have to be configured exactly can go through several iterations. Starting with a simple specification of a product that leaves out many sensible aspects because they seem too obvious can soon lead to these aspects being omitted in the next generation (or hopefully just a prototype of that) because they cost something and meeting the spec at cheap cost is the goal after all. This leads to a more concrete physical spec with fewer openings for creative interpretations that would annoy the spec writers, but also fewer openings for creative interpretations that the spec makers would be delighted to see.

Precise physical specs about how much to use of which material and where deletes opportunities for cheating the stats of extensive growth, but it also deletes opportunities for realizing intensive growth. It is more farsighted to plan in terms of useful effects. Using less of some material for a product because a different physical configuration gives the same stability is a good change. For this to also appear as a good change by the standards of the specifications, we should avoid statements about minimum amounts of this or that specific material, but instead talk like "behaviorists" and list aspects like nine out of ten units of the device still working after one hundred drops onto concrete from a certain height, without dictating how the object achieves that.


joint production and multiple recipes can be handled by linear programming. see >>1239482
we could choose to optimize for labour time, which naturally leads to picking the cheapest production methods that yield the desired use-values


>muh lpsolve mufugga
Not the point.

Unique IPs: 125

[Return][Go to top] [Catalog] | [Home][Post a Reply]
Delete Post [ ]
[ home / rules / faq ] [ overboard / sfw / alt ] [ leftypol / siberia / hobby / tech / edu / games / anime / music / draw / AKM ] [ meta / roulette ] [ cytube / git ] [ GET / ref / marx / booru / zine ]