Thread · 11 tweets · 11 Apr 2023

Indeed! Social structures are technology, and the naive way in which we do technology devalues them. Some notes: 1) this matches a broader pattern; 2) the converse is also true; and 3) Some people are working on this! A short 🧵 with thoughts.

(1) On the broader pattern: We have treated "nature" as external to humankind, as if we wouldn't be part of it, which in turn gave us leave to exploit it. We are now doing the same with sociality: it's external to our engineering and subject to extraction.

Much of what falls under "privacy" issues is really looking at social behaviour: most of that processing is only meaningful on groups. And the interventions that this powers, to influence behaviour, is aligning people and simplifying groups, in effect extracting complexity.

Details on algorithmic monocultures, LLMs, etc. are for a longer post but the principle is that the legibility/categorisation process extracts complexity as a resource, exploits the complexity → simplicity gradient. We did that with "natural resources"; how's that working out?

(2) The converse is true: any technology is that sets rules of any kind for human interactions, even indirectly, is a social structure (or, more precisely, an institution). Tech can be usefully studied using for instance Ostrom's ADICO institutional analysis framework.

A lot of us have been making computer things that mediate between people without considering the institutional dimension. That's basically doing half the job of a technologist. It's a shame because we can understand these systems much better as institutions.

In tech circles there are only three institutional forms: (benevolent) dictatorship, some vague kind of voting, or punt to the state. That's a lot of untapped potential for improvement right there. (Some of us will be talking about this at 2023.ipfs-thing.io.)

A point on which I disagree with @mattecapu is on natural vs engineered sociality. Natural systems can be purposive (see Beer's POSIWID: the Purpose Of a System Is What It Does). I think it's fruitful to look at all sociality as solving governance issues: berjon.com/internet-trans…

3) People are working on this! Lots of great stuff in @metagov_project & from @Ostrom_Workshop, to name just 2 among many. Again, I think that calling upon ecology, governance, and tech at once to form a crisis discipline as covered in berjon.com/stewardship/ is the way forward.

I'm not yet at a level at which I can usefully build upon arxiv.org/abs/2108.05318 or eg. suggestions that the capabilities approach (core to rethinking tech) is resource-theoretic — but there's definitely a lot going on in the space.

My concerns are more about whether we'll understand this fast enough to design effective interventions, if we can change enough of the culture of tech, and how we'll tackle an entrenched big tech as the new big oil. Anyway — back to work!