~/wiki
Récent+ Ajouter

Jevons' Paradox

Mis à jour le 2026-06-11Confiance : high
jevons-paradoxeconomic-theoryefficiency-paradoxsoftware-generationdemand-scalingai-developmentresource-consumption

Economic principle stating that as technological improvements increase the efficiency of resource use, the rate of consumption of that resource tends to increase rather than decrease. Originally observed with coal efficiency improvements in the 19th century, now highly relevant to AI-generated software development.

Core Principle

When a technological advancement makes a resource more efficient to use, demand for that resource often increases because:

  • Lower cost per unit makes the resource more accessible
  • New use cases become economically viable
  • Users expand their consumption to take advantage of efficiency gains
  • Previously constrained applications become feasible

Application to Software Development

andrej-karpathy identifies Jevons' Paradox as a key dynamic in AI-assisted software development:

Traditional Software Development

  • High time/cost barriers limited software creation
  • Careful prioritization of development efforts
  • Focus on multi-purpose, reusable solutions
  • Conservative approach to custom tool creation

AI-Generated Software Era

As AI makes software creation nearly effortless ("working software increasingly comes out on a tap"):

Demand Explosion: Rather than reducing software needs, efficiency gains dramatically increase software consumption:

  • Custom explainers and visualizers for every concept
  • Project-specific dashboards and monitoring tools
  • Bespoke single-use applications become viable
  • 10X expansion of test suites and validation tools
  • Automated optimization and refactoring tools
  • Custom interfaces for every research project

Mental Model Transformation: "Free your mind" - moving from scarcity-based thinking to abundance-based possibilities.

Historical Context

Original Jevons' Paradox (1865)

  • William Stanley Jevons observed that improved coal efficiency increased total coal consumption
  • Better steam engines made coal more valuable, driving increased demand
  • Efficiency gains enabled new industries and applications

Modern Examples

  • More fuel-efficient cars leading to increased driving
  • Cheaper data storage resulting in exponential data creation
  • Faster internet enabling bandwidth-intensive applications

Implications for AI Development

Strategic Planning

  • Prepare for explosive growth in software creation and deployment
  • Infrastructure scaling to handle increased AI-generated workloads
  • Quality control systems for high-volume software production

Development Practices

  • Shift from "build once, use many times" to "build many, use specifically"
  • Embrace rapid prototyping and disposable software
  • Focus on problem definition rather than implementation constraints

Economic Impact

  • Potential disruption of traditional software markets
  • New business models based on software abundance
  • Skills shift from implementation to orchestration and design

See also