Past Automation and Future A.I.: How Weak Links Tame the Growth Explosion
A May 2026 working paper that uses task-based growth accounting and an endogenous automation model to study past automation and the pace of future AI-led growth.
Authors or creators: Charles I. Jones, Christopher Tonetti
Checked: 2026-07-29
Source note: Primary working paper, version 0.5. The quantitative forecast is model-dependent and may change in later revisions.
Claims connected to this source
- Automation replaces slowly improving human task inputs with faster-improving capital.
- Complementarity among tasks creates weak links that slow aggregate gains.
- Growth can eventually accelerate sharply even when the first decades look similar to the historical path.
Open original · Stanford Graduate School of Business and NBER