Polycentric proceduralism

Rather than pooling everyone's feedback into one yardstick, a separate standard is fit for each group, so a response meant for group A is judged by A's measure and never by a population average. The groups arrive two ways and the difference matters: they can be given, as real bounded communities whose norms are read off what they already accept and engage with rather than from labels or a written charter, or discovered, as latent clusters found inside pooled feedback, which answer to no institution anyone can join, contest, or leave. The case for working this way is that the authority to say what "aligned" means belongs with many contextual, participatory centers rather than one lab, and each per-group standard is meant to be one such center. One member here straddles: it sets out a personalization family, several reward models by representation learning or clustering, alongside an aggregation family that folds those models back into a single pipeline under utilitarian and Leximin rules, and it is the personalization half, not the aggregation half, that puts it in this regime at all.

Each group trains on its own members' feedback where that feedback already sits and sends out only the trained thing, whether model updates, raw parameters, or compact selectors standing in for preferences, so no group's data leaves it and no pooled corpus is ever assembled. What separates this from fitting a standard per group is which part is partitioned: here it is the pipeline, the data and the computation, and not necessarily the number of standards at the end. That distinction cuts both ways and is worth stating plainly rather than burying. One member trains a personalized policy for each client and so meets the previous strategy's test as well, and is placed here because its defining move is the federated architecture; another keeps data local but converges by averaging on a single shared preference predictor, so its plurality is entirely in the process, in who holds the data, who computes, and who can withhold or fork, and not in the outcome. That process-level plurality is a real lever a member of a pooled corpus does not hold, which is why these belong in a procedural regime, but the single-model case is the thinnest claim to partition in the set.

Each online community authors its own procedures, such as votes, juries, term limits, and appeal steps, as executable policy attached to actions on the platform, and a shared runtime executes whatever each community wrote. The platform supplies the engine and the hooks but takes no position on the rules, so neighboring communities on the same substrate can run incompatible constitutions. The ambition is a portable, composable governance layer, with components that move from one context to another, rather than a single set of platform-wide rules.

A legal body is placed between the people whose data it is and anyone who wants to use it: the body holds the rights, owes duties to those it holds them for, and sets the conditions under which the data may be accessed, licensed, or trained on, so a corpus can be assembled only on that body's terms. Two forms sit here and they work differently. Member-owned custodians multiply, as trusts each under their own deed or cooperatives under a fiduciary obligation to citizen members, each with different terms, and the member's operative power is exit, leaving one for another. The third is a single negotiated multi-party structure for the world's language data, whose plurality is internal, among stakeholders, values, and rights, with no deed and no exit; it is the loosest fit for a regime built on partition, and it earns its place because it is the one paper that carries the custodian idea onto language-model training corpora outright, an extension the trust and cooperative papers, written for personal data and member services, only leave open.