The Submission Economy: A Framework for Studying Algorithmic Selection and Recommendation

An algorithmic interpreter increasingly stands between an information producer and the scarce resource it is competing for, whether that resource is a recruiter's attention, an investor's meeting, an editor's decision, or a consumer AI system's eventual recommendation. This working paper proposes a framework for studying that phenomenon, which we call the Submission Economy, and treats it as a hypothesis to be tested rather than an established fact. We distinguish three phenomena often collapsed together: machine mediation, which changes how a submission is encountered; machine gatekeeping, which changes whether it reaches a human at all; and machine-mediated decision-making, in which an AI output directly produces or materially shapes a selection, with or without a human decision-maker in the loop. This distinction separates hiring-style gatekeeping from AIVO's own prior research into AI-mediated brand recommendation, which belongs to the third category. We present a five-stage model, encounter, interpretation, evaluation, routing, and outcome, defined in observable, loggable terms. We propose the counterfactual difference between an AI-mediated pathway and a matched human-only pathway, the AI effect, as the framework's central methodological test. We present a targeted, non-systematic evidence scan across seven cases, rather than a systematic review, and rate each on an operationally defined evidence-maturity scale. No domain in this scan currently reaches our highest evidence-maturity level; hiring is the strongest case and sits one level below it. We offer the framework and a set of falsifiable propositions as a research agenda, not as an established account of a new economic regime.