CUDOS

In the AI Story

The communalism norm is under greatest pressure in contemporary AI, where the most capable frontier models remain proprietary despite being trained on communally produced data. The open-source movement represents genuine commitment to the norm—knowledge should be shared—but operates within a commercial environment that rewards proprietary control. Scholars have documented this tension as a normative crisis: the ideal of communal knowledge coexists uncomfortably with the reality of privatized capability, and the gap between ideal and practice is widening rather than narrowing.

Organized skepticism faces its own structural challenge in the AI age. The norm demands critical scrutiny before acceptance; the competitive environment compresses evaluation timelines below the threshold at which adequate scrutiny is possible. A 2025 paper invoking Merton warned against 'people being so fooled by the hype that they do become slaves, not to incredibly intelligent computers but to stupid computers that are taken to be unquestionable authorities.' The warning captures the inversion: a community that should be institutionally committed to doubt is, under competitive pressure, institutionally committed to credulity.

Some scholars have proposed that the AI community operates under an inverted normative structure—DECAY rather than CUDOS. Differentialism (claims evaluated by claimant status), Egoism (individual advantage over communal benefit), Capitalism (commercial interest subordinating knowledge production), Advocacy (organized promotion of predetermined conclusions). The DECAY framework may overstate the inversion—genuine commitments to safety and openness exist—but it captures real pressure: the platform economy's incentive structure conflicts with Mertonian ideals in ways that the scientific community's structure did not.

Origin

John Ziman introduced the CUDOS acronym in his work on the social structure of science, drawing directly on Merton's 1942 essay 'The Normative Structure of Science.' The acronym reordered Merton's list (originally universalism, communism, disinterestedness, organized skepticism) to produce a pronounceable word, sacrificing Merton's logical sequence for memorial convenience. Merton himself did not use the acronym, but it has become the standard way scholars reference his framework.

The norms were articulated during World War II, when science's relationship to state power was being transformed and the question of what distinguished science from propaganda or ideology was urgent. Merton's framework was partly descriptive (this is the ethos scientific communities aspire to) and partly prescriptive (this is what science must be to remain science rather than becoming something else). The dual character—simultaneously describing ideals and defending them—has produced decades of debate about whether the norms are empirical claims or moral commitments.

Key Ideas

Communalism. Knowledge as communal property to be shared openly—violated by proprietary models, trade secrets, and privatization of publicly funded research.

Universalism. Claims evaluated by impersonal criteria independent of claimant characteristics—violated when institutional prestige or demographic identity shapes evaluation.

Disinterestedness. Structures minimizing personal interest in outcomes—violated when evaluators have financial stakes in what they evaluate.

Organized Skepticism. Institutionalized critical scrutiny of all claims—violated when competitive pressure compresses evaluation below the threshold of adequate testing.

Norms as Dams. The institutional standards that redirect knowledge production toward cumulative understanding rather than private advantage—currently being contested in real time in AI development.