## A New Industrial Revolution?

## Source details

**Canonical URL:** [A New Industrial Revolution?](https://www.imf.org/-/media/files/publications/fandd/article/2025/12/kishtainy.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/fandd/article/2025/12/kishtainy.pdf.md)
- [Structured JSON version](/-/media/files/publications/fandd/article/2025/12/kishtainy.pdf.json)

---

### General purpose technology and historical parallels
- Steam power, electricity, and ICT are identified as past general purpose technologies; AI is likely a distinct general purpose technology and a "new method of invention".
- Key historical markers cited:
  - First Industrial Revolution began in Britain in the late 18th century.
  - Second Industrial Revolution began in the late 19th century (electricity).
  - Third Industrial Revolution began in the late 20th century (ICT).
- Economic implication: general purpose technology has broad application across many fields and can raise labor productivity, demand for labor, and living standards over centuries.
- Quoted historical observation: machines led to lives "blighted" for many workers in early factories; cloth production benefited from inventions such as the spinning jenny in the 1760s, while mechanized weaving later "annexed by the power-loom."

### Speed and magnitude of productivity effects
- Historical evidence on timing of effects:
  - Nicholas Crafts found gains from steam came only after 1830 because steam-powered sectors initially constituted only a small fraction of the economy.
  - Robert Solow’s observation: "you can see the computer age everywhere but in the productivity statistics"—the "Solow paradox".
- Comparative assessment:
  - ICT delivered productivity gains faster and larger than steam did historically.
  - AI's current contribution to productivity has so far been "modest" and has been labeled a "productivity paradox" by some observers.
  - If ICT is a guide, AI’s productivity impact may be felt faster than earlier general purpose technologies, though perhaps not matching the most optimistic projections.

### Distributional effects, labor, and wages
- Short-run and medium-run complications:
  - The Industrial Revolution’s early decades showed rising output per worker but stagnating real wages until wages rose "in line with productivity" only after the middle of the 19th century.
  - Engels’ observations: some workers (e.g., hand-loom weavers) suffered collapsing wages and 18-hour workdays when mechanization advanced.
- Theoretical effects modeled by Acemoglu and Restrepo:
  - Displacement effect: automation of tasks reduces labor share in national income and can decouple wages from productivity.
  - Reinstatement effect: technology creates new tasks in which humans have comparative advantage, generating new employment (examples across 19th–20th centuries include engineers, telephone operators, machine technicians, software designers).
- Recent trends and concerns:
  - Since the late 20th century, real wages in many leading economies have been flat.
  - Many ICT and AI innovations have targeted automation rather than task-creation, exacerbating stagnant labor demand, slow wage growth, and rising inequality.
  - There is an argument for pursuing "labor-reinstating AI" (examples: education and health), where AI tools could increase demand for teachers and doctors by enabling individually tailored learning and treatment programs.

### Machine singularity, limits, and policy options
- Singularity debate:
  - Technologists discuss AI reaching "singularity" where machines could improve and invent themselves, potentially eliminating labor reinstatement.
  - William Nordhaus’s empirical tests suggest most conditions for such singularity are "far from being met" because much of the economy is physical rather than informational (examples: poaching eggs, cutting hair, soothing toddlers).
- Policy leverage:
  - A key difference from earlier revolutions: policymakers now have tools to influence the economy and technology’s path.
  - Innovation has significant market failures; currently choices about AI are largely left to corporations with limited concern for broader economic impacts.
  - Normative recommendation: governments and regulators can and should guide technological development to ensure economic benefits are broadly shared, contingent on political will.

*Source: Niall Kishtainy, "A New Industrial Revolution?", F&D, December 2025*

---


_Source: https://www.imf.org/-/media/files/publications/fandd/article/2025/12/kishtainy.pdf_