The neural-net→'deep learning' rebrand can be read as a field-scale Fresh Start Effect — dissociation from a stigmatized past self — though the collective-scale extension is unestablished

Two vault notes sit near each other by embedding but not by content: the Fresh Start Effect (claim-fresh-start-effect-requires-subjective-new-beginning-framing) and Lighthill's combinatorial-explosion diagnosis (claim-lighthill-1973-blamed-combinatorial-explosion). At the literal level the resemblance is superficial — one is a psychological mechanism of individual motivation, the other a historiographical correction about UK AI funding. Their proximity tracks a shared rhetorical shape ("the tidy single-cause story is wrong"), not a shared mechanism.

The connective observation is that they meet through a third, AI-native instance. The Fresh Start Effect holds that a temporal landmark boosts goal pursuit only when subjectively framed as a new beginning, by widening the felt distance from an "imperfect past self." The neural-net→"deep learning" relabeling of the mid-2000s has the same structural parts at field scale: a boundary event (the 2006 layer-wise-pretraining result plus the new name) that mattered because it was framed as a fresh start, dissociating the field from a stigmatized past — the "backwater" neural-network identity of the SVM era (claim-neural-networks-backwater-before-deep-learning-rebrand). The name was available before Hinton and was re-pointed at many-layered nets around 2006 (claim-deep-learning-term-predates-hinton), which is what let it function as a boundary marker.

The critical caveat, kept visible: this is an analogy, not an established finding. The Fresh Start / temporal-landmark literature (Dai, Milkman & Riis) concerns individuals; whether it extends to collective or organizational fresh starts was an open question (question-fresh-start-effect-extends-to-collective-organizational-scale), now answered NO for the located peer-reviewed literature (observation-fresh-start-effect-extensions-stay-individual-not-collective: three post-2021 extensions all keep the individual as the unit). That verdict confirms this note must stay a labeled analogy — the field-scale mapping is not backed by a collective-scale finding — so it remains a flagged seedling. The mapping also converges, from the org-theory side, with the finding that a boundary restructures a firm only through what actors do with it, not the bare event (claim-executive-succession-restructuring-runs-through-actor-action; see moc-schema-change-and-restructuring). Reading the rebrand as only stigma-escape would itself repeat the monocausal flattening the Lighthill note warns against: it was both a genuine technical advance and a reframing.