Abstract. Hiring decisions rest on an inference problem: employers must predict on-the-job capability from imperfect proxies, of which the academic credential is the most widely used. This paper synthesises six decades of work across experiential learning (Kolb), authentic assessment (Wiggins), constructive alignment (Biggs LinkedIn Economic Graph), and labour-market signalling (Spence) to advance a single conceptual claim: a completed, provenance-verified piece of work is a lower-noise indicator of job-readiness than a grade, because it functions as direct evidence of capability rather than as a costly proxy correlated with it. The paper’s contribution is to read this claim through signalling theory and to identify its binding condition — verification integrity — which it shows to comprise two distinct problems of very different maturity: record integrity (largely solved) and production integrity (largely unsolved). It argues that, in an environment of widely available generative AI, the value of demonstrated work increasingly resides in evidence of authentic authorship rather than in the artefact or in a tamper-proof record of it. This is a synthesis of existing theory; it advances a testable proposition rather than a measured outcome, and reports no empirical results. Keywords: authentic assessment; experiential learning; constructive alignment; signalling theory; skills-based hiring; verifiable credentials; portfolio assessment; assessment integrity; generative AI; job-readiness.
A. Acosta (Mon,) studied this question.