The mechanism is psychological as well as institutional. Observers who encounter only the finished work cannot infer the labor that produced it. They credit the visible producer — the named author, the celebrated scholar, in the AI case the AI itself — and fail to credit the invisible curator whose judgment determined what the visible producer did or did not do. The invisibility is structural: it is a feature of how finished work presents itself, not a contingent failure of observation.
In AI collaboration, the misattribution has a new dimension. Observers watching AI-assisted work often credit the AI with capabilities it does not possess, because they do not see the human judgment that directed the output, rejected the failed drafts, revised the inadequate responses, and iterated toward quality. The AI's apparent autonomy is a function of the invisible curation that surrounds it. Without the curation, the AI's raw output would display its limitations openly. With the curation, the output appears more capable than it is — and the credit flows to the model rather than to the curator.
The consequences for AI-era labor are not merely about credit. Organizations that treat AI-assisted work as automation rather than curated collaboration will structure their workflows, compensation, and professional development in ways that undervalue curatorial judgment. They will reward speed and volume — metrics that AI optimization naturally produces — rather than the evaluative depth that distinguishes excellent AI collaboration from merely competent AI use. The result will be organizations abundant in output but impoverished in judgment.
Blair's historical framework makes the resolution visible. The institutions that recognized and supported curatorial labor in previous eras produced intellectual achievements the historical record celebrates: scholarly editions, research libraries, critical review journals. The institutions that failed to support curatorial labor produced the abundant worthless output the historical record has forgotten. The AI era faces the same choice — and the invisibility problem is one of the main obstacles to making the choice wisely.
The concept is implicit throughout Blair's historical work on compilation, editing, indexing, and reference production. Its naming and extension to the AI context is an explicit application of her framework to contemporary conditions.
Labor hidden by success. The smoother the finished artifact, the more invisible the labor that produced it.
Misattribution to the visible. Observers credit the named author or the AI, not the invisible curator.
Institutional undercompensation. Invisible labor is structurally difficult to value, measure, or reward.
AI amplification. The invisibility is more intense in AI collaboration because the curator's interventions are not even visible as edits to a draft.
Solvable through institution design. Institutions can choose to recognize and support curatorial labor; the choice has historically made large differences in intellectual outcomes.