One AI system can summarize a contract, analyze data, write code, interpret an image, and develop a marketing plan. But doing many things does not automatically make a technology general purpose.
The determining question is can a technology become an improving foundation upon which many industries build further innovations?
If AI is a specialized tool, organizations can purchase it for a defined task and measure the result. If AI is a general-purpose technology like electricity or computing its effects will extend beyond individual tools. It could reshape workflows, occupations, organizations, markets, infrastructure, and institutions.
Technology ought to support responsible human stewardship, creativity, service, and flourishing without displacing human dignity or accountability.
AI is potentially general purpose because it addresses capabilities found throughout human work: language, perception, prediction, creation, reasoning, and action. Its significance lies not simply in imitating conversation but in becoming an adaptable capability within many activities.
Several characteristics support classifying AI as an emerging general-purpose technology.
Pervasiveness: AI applies across accounting, finance, marketing, software, education, healthcare, science, media, government, and robotics.
Adaptability: The same base model can be configured through prompting, retrieval, fine-tuning, tools, organizational data, and workflow rules. It can become a tutor, analyst, coding assistant, customer-service copilot, or research aid without being rebuilt from scratch. Natural language further reduces the barrier to adaptation because users can describe goals, constraints, examples, and feedback without traditional programming.
Multimodality:Multimodality allows AI to connect activities previously handled by separate systems. AI increasingly processes text, speech, images, video, software, structured data, and sensor information. This matters because real work rarely exists in one format.
Continuous Improvement: As capability per dollar rises, applications that were previously too expensive or unreliable become possible. AI is improving through better architectures, training methods, inference-time reasoning, tool use, smaller models, and more efficient deployment.
AI enables others to create new products and processes: agents, personalized services, scientific-discovery systems, adaptive robots, natural-language software development, and new business models.
Like electricity and computing, AI produces substantial value only when combined with skills, data, workflow redesign, integration, trust, and governance.
Its value emerges from a sociotechnical system:
Technology + task + human + data + process + incentive + institution + consequence.
Organizations can use AI to improve speed, access, consistency, experimentation, and decision support. But they must begin with meaningful outcomes, redesign workflows, test reliability, protect data, and preserve human responsibility for consequential judgments.
AI may reshape how society produces knowledge, organizes work, and distributes economic power. But general purpose does not mean universally beneficial or inevitably transformative.
AI adoption can outpace reliable implementation, and falling costs can coexist with growing infrastructure, energy, inequality, and governance challenges.
Ask whether AI demonstrates:
- application across many sectors;
- adaptation to significantly different uses;
- continuing technical and economic improvement;
- integration across multiple forms of information;
- downstream innovation;
- dependence on complementary organizational change.
AI shows substantial evidence on all the six dimensions above. This supports classifying it as an emerging general-purpose technology, although its long-term productivity and institutional effects are still developing.
We should therefore approach AI not merely as a product to install, but as a capability around which work must be thoughtfully redesigned.
Leave a Reply