This seems like an analysis of a [relatively] good-faith algorithmic society, or at least assumes the social network as a sufficiently illuminating archetype.
"The algorithmic society sees and serves the engaged" is probably not generic enough in two independent senses:
* There's nothing in "algorithms" as such that precludes them from seeing the non-engaged (or even the attempting-to-hide).
* There's nothing in "algorithms" as such that forces them to serve any user, engaged or not (Enshittification Theory, Cory's very Hegelian update of Marx, would say that ultimately they only serve themselves in a purely extractive "nice reified social capital you have here, shame if you were to leave it behind when you switch platforms" sort of way).
A more general look at an algorithmic society might be one where more of what we call resource allocation decisions (in a very general sense of resource) are made in a case-by-case basis by software of arbitrary algorithmic complexity. That has no implications on visibility or preferences; based on our general experience with large software systems, my bet is that at a first order of approximation:
* It increases potential transparency, but in practice makes individual decisions opaque.
* It increases the precision of control by whoever controls the algorithms, but increases the speed and size of side effects and mistakes.
* It increases the potential for finding nontrivial equilibria, but also increases the benefits of adversarial behavior at margins of speed or scale.
Hypothetical historical non-rethorical question: If the freeish market/limited democracy revolution happened initially not in large established countries but in less powerful ones that were unable to prevent the internal development but ended up benefiting from it (e.g. the Low Countries/England vs the Hapsburg or even China), then perhaps an algorithmic society oriented around optimized social decision making and not, say, surveillance, might be more likely to happen in compact countries, rich in relative terms but small in an absolute sense? E.g., I'd imagine a hive mind Switzerland, Luxembourg, or Taiwan first, rather than a continuous progression from larger countries - the bad equilibrium of algorithmic surveillance and oligopoly spamming/commerce/advertising platforms is too tempting.
This seems like an analysis of a [relatively] good-faith algorithmic society, or at least assumes the social network as a sufficiently illuminating archetype.
"The algorithmic society sees and serves the engaged" is probably not generic enough in two independent senses:
* There's nothing in "algorithms" as such that precludes them from seeing the non-engaged (or even the attempting-to-hide).
* There's nothing in "algorithms" as such that forces them to serve any user, engaged or not (Enshittification Theory, Cory's very Hegelian update of Marx, would say that ultimately they only serve themselves in a purely extractive "nice reified social capital you have here, shame if you were to leave it behind when you switch platforms" sort of way).
A more general look at an algorithmic society might be one where more of what we call resource allocation decisions (in a very general sense of resource) are made in a case-by-case basis by software of arbitrary algorithmic complexity. That has no implications on visibility or preferences; based on our general experience with large software systems, my bet is that at a first order of approximation:
* It increases potential transparency, but in practice makes individual decisions opaque.
* It increases the precision of control by whoever controls the algorithms, but increases the speed and size of side effects and mistakes.
* It increases the potential for finding nontrivial equilibria, but also increases the benefits of adversarial behavior at margins of speed or scale.
Hypothetical historical non-rethorical question: If the freeish market/limited democracy revolution happened initially not in large established countries but in less powerful ones that were unable to prevent the internal development but ended up benefiting from it (e.g. the Low Countries/England vs the Hapsburg or even China), then perhaps an algorithmic society oriented around optimized social decision making and not, say, surveillance, might be more likely to happen in compact countries, rich in relative terms but small in an absolute sense? E.g., I'd imagine a hive mind Switzerland, Luxembourg, or Taiwan first, rather than a continuous progression from larger countries - the bad equilibrium of algorithmic surveillance and oligopoly spamming/commerce/advertising platforms is too tempting.
Thanks! A great deal to think about in here...
Brad
Overall blog question: Do you want people pointing out typos, or are you just annoyed by those comments? I've found bloggers in both categories.