Thread by @mattecapu

#PaperADay🦆 time!

Today I read Apolito's essay I announced yersterday.

TL;DR: anarcho-communism should develop a competing alternative to markets for large-scale economical organization, and 'integrated information' is an appealing theoretical framework to look at

1/n

https://twitter.com/mattecapu/status/1478831179872808963

First, a reminder that Aurora Apolito is a pseudonym (I don't know if her real name is public knowledge so I'm not gonna dox). But it's cool that the name itself roughly means 'dawn of the stateless [society]', which is really cool 2/n

The paper starts with a question: 'Is anarchism [...] a system destined to only work in the scale of small local communities?' 3/n

The main problem is in the 'large scale distribution of services'. Markets are an appealing way to accomplish this task, arguably the easiest, and that's why they are so wide-spread. But they are not fair, since, as Apolito puts it, 'services are not cost-effective' 4/n

In other words, markets optimize profits but so little of the collective utility of a society can be measured in terms of profit. How much money is worth the health of a person? Their happyness? Their peace of mind? Hard to say. 5/n

Instead, capitalist markets fetishize such measures and inexorably produce inequality, by design. 6/n

Apolito then recounts some history of cybernetic communism, namely mathematical approaches to economy planning. She doesn't discuss details, so I don't know exactly how these would work, but they seemed quite naive and too centralized (hence fragile) for my taste 7/n

For instance, Allende's Chile tried to implement such a system, 'Project Cybersin', which was going to be a system of data-collecting endpoints and a room in Santiago full of computers and economics software that crunched said data. 8/n

She then makes a case for complexity in communist societies. Funnily enough, she refers to the paper I've read just yesterday 9/n

https://twitter.com/mattecapu/status/1478808599447195653?s=20

Recall the paper shows how societies growth is guided by (and necessitates) growth in scale and information-processing capabilities. Apolito argues capitalist societies are low in the information-processing scale 10/n

Yes, capitalism is 'the state of the art'. Capitalism and authoritarian governments are kind of 'low effort' structures for societies at this scale, but in order to go to the next level one needs a better way to integrate information and organize collective activity 11/n

To overcome them, we need better theoretical facilities to coordinate activity at the current scale 12/n

Apolito sits on this point quite a bit, since she points out markets are usually rejected with a blanket 'math-driven systems are bad' kind of sentiment, which is totally unwarranted and harmful, as per previous remarks 13/n

Two wonderful quotes from this section: 14/n

Once the problem has been stated, Apolito gets on with some math. First, how does one define complexity? 15/n

She adopts a measure called 'effective complexity', which is related to both Kolmogorov's complexity and Shannon's entropy. It's described in another paper she cites, and which I didn't read 16/n arxiv.org/abs/0810.5663

Then she describes integrated information (IT), a notion proposed by Tononi and collaborators some years ago as a candidate measure for 'consciousness' of a system. More concretely, it's basically a measure of how much emergency does a collective system experience 17/n

By 'emergency' I mean how much of the behaviour of the whole can't explained by the behaviour of the parts, taken independently. This emergent behaviour is then 'the behaviour of the connections', and high values of IT denote high functional interconnectivity 18/n

Apolito then proposes IT as the measure a cybernetic communist society should try to optimize because, she claims, it measures productive collective cooperation 19/n

Maximising IT doesn't leave space for inequality build up, since that would decrease the integration of some parts of the system. For the same reason, it also maximises individual agency, since that makes the system much 'messier', which is good for IT 20/n

How does a society optimize IT? Two ways: (1) maximise connectivity (intellectual and geographical alike) and (2) maximise complexity (which Apolito's inteprets as culture, art, science and philosophy) 21/n

Then the article is cut off quite abruptly, after discussing models of networked communities and their mutual information 22/n

Still, the article already packs quite a punch. Proposing IT as a measure of 'collective efficiency' is a really cool idea and resonates with the unexpressible feelings I had for a long time regarding the effectiveness of 'ecological' systems compared to capitalist systems 23/n

In particular, as she remarks on p.16, 'IT can function as an 'objective valuation' [...] as opposed to the subjective price valuations of the markets'. She illustrates this concept with visual arts markets (and she doesn't even mention NFTs!) 24/n

A theory of value based on IT? Glorious! 25/n

The biggest question I'm left with is: pricing is to market as optimizing IT is to...? There doesn't seem to be a good answer in the paper (and there need not be) 26/n

Markets are very appealing since they have quite an elegant, self-sustaining mechanism. The big theoretical problem of finding an analogous mechanism whose outcome is to maximise IT is still open 27/n

My 2 cents: we should look at ecosystems. They seem to engage in market-like activities, but their evolution is much different from the inequality-maximising, monopolies-producing capitalist economies we have 28/n

Instead, species live in harmony with each other, diversity is widespread, and the whole system is not only resilient, but antifragile to external perturbations 29/n

Moreover, there is a kind of IT maximisation at hand. Ecological niches are constantly discovered and seized, mantaining a thick web of relationships between species: this ticks off the 'connectedness' box in Section 3.3 30/n

At the same time, survival dictates a certain degree of complexity for individual species, which survive if they can adequately represent and understand their surroundings: this ticks off the 'complexity' box in the same section 31/n, n=31