When production is an O‑ring (multiplicative) technology, tasks are quality complements: automating one task alters the marginal value of others, can force discrete bundled adoption choices, and may increase earnings for workers who retain control of remaining bottleneck tasks. Simple linear task‑exposure indices therefore mismeasure displacement risk and policy should focus on bottleneck structure and time allocation.
— This reframes automation policy and labour forecasting: regulators, firms and retraining programs should target where automation changes the structure of bottlenecks, not average task vulnerability, because the social and distributional outcomes can be qualitatively different.
EditorDavid
2026.03.29
85% relevant
The article documents Maximo robots (AES) and Nvidia simulation work overcoming the installation bottleneck—installation rates of ~1 module per minute and near‑doubling of output—illustrating how targeted automation addresses a concrete construction chokepoint for renewable deployment.
EditorDavid
2026.03.08
85% relevant
AUAR's micro‑factory is a direct example of automation aimed at relieving a construction labour shortage: the company claims one micro‑factory can produce a house's framing in about a day (vs four weeks for a crew) and be ~30% cheaper. That suggests the industry bottleneck may move from on‑site framing labour to other constraints (panel transport, on‑site fitting, permitting, local trades), matching the existing idea that automation shifts rather than eliminates bottlenecks.
Tyler Cowen
2026.01.05
100% relevant
Gans & Goldfarb’s recent working‑paper model (reported by Tyler Cowen) formalizes an O‑ring multiplicative task technology, derives discrete/bundled adoption and the possibility of rising labour income for remaining tasks, and explicitly warns that linear exposure indices overstate displacement.