Practical Action
Practical Action — originally the Intermediate Technology Development Group — is the organization E. F. Schumacher founded in 1966 to develop and deploy the intermediate technology his philosophy called for. The founding recognized something Schumacher insisted throughout his work: appropriate technology does not emerge from market forces. It requires deliberate development, institutional support, and communities of practice that maintain and improve it over time. The organization spent its first fifty years developing small-scale irrigation systems, bicycle-powered tools, improved cook stoves, micro-hydroelectric installations, and hundreds of other technologies scaled to the conditions of the communities that used them. It renamed itself Practical Action in 2005 to reflect a broader mandate while preserving the core commitment. Its current work includes a growing focus on whether Schumacher's framework can be applied to digital and AI technologies — the question Berry and Stockman's 2024 work on intermediate artificial intelligence began to formalize.
In the AI Story
The organization's founding established a principle the AI transition has largely ignored: that technology appropriate to human flourishing must be developed, not merely deployed. The market develops technology according to market criteria (scale, speed, return on capital). Technology meeting Schumacher's criteria — cheap, small-scale, compatible with creativity, reciprocal in its relationship to the user — must be developed according to those criteria, which requires institutional vehicles whose mandates are not reducible to market performance.
Practical Action's fifty-year operational record provides evidence that the model works. Millions of people in rural communities across Asia, Africa, and Latin America use technologies developed by the organization and its partners. The technologies have been maintained and improved over decades by the communities that use them, because they were designed to be maintainable and improvable. The reciprocal relationship Schumacher envisioned — tool serving user, user improving tool — has been realized at real operational scale.
The contemporary question is whether an analogous vehicle is needed and achievable for AI. The argument for is straightforward: frontier AI is being developed by a handful of corporations whose mandates are maximizing return on capital, and the technologies emerging from this process are optimized accordingly — proprietary, centralized, extractive in data practices, opaque in operation, priced for subscription revenue. These are not the properties of intermediate AI as Schumacher's framework would specify them. If intermediate AI is to exist, institutional vehicles analogous to Practical Action must develop it, because it will not emerge from the frontier labs.
Several such vehicles have begun to emerge. The Schumacher Center for a New Economics has hosted conversations on AI and the appropriate technology framework. Open-source model development at organizations like EleutherAI, the LAION collective, and various academic research groups has produced alternatives to proprietary frontier models. Cooperative structures for AI development remain largely theoretical but are being discussed in the platform cooperativism literature. The work is in its early stages; the question is whether it will scale in time to provide a genuine alternative to the corporate-controlled AI infrastructure.
Origin
Schumacher founded the Intermediate Technology Development Group in 1966 with support from Oxfam and other development institutions. The founding followed his 1965 visit to India and reflected his recognition that the philosophical arguments he had been making required institutional embodiment to have effect.
The organization's work through the 1970s established the operational viability of the intermediate technology model. Schumacher served as its president until his death in 1977. The organization renamed itself Practical Action in 2005 while preserving its core commitments, and it continues to operate today with offices in seven countries and programs on six continents.
Key Ideas
Institutional embodiment. Philosophy alone does not produce appropriate technology; institutional vehicles whose mandates are not reducible to market performance are required, and must be built deliberately.
Fifty-year operational record. The model of developing technology to Schumacher's criteria, deploying it through community partnerships, and maintaining it over decades has been empirically validated across thousands of deployments in dozens of countries.
Reciprocal development. Technologies are designed to be maintainable and improvable by the communities that use them; the relationship between designer and user is collaborative rather than extractive.
Applied to AI: the question is whether an analogous vehicle can be built for digital and AI technologies, and whether it can operate at sufficient scale to provide a genuine alternative to corporate-controlled frontier AI.
Emerging efforts. Open-source model development, the Schumacher Center's work, platform cooperatives — early institutional experiments that may or may not scale to match the challenge.
Appears in the Orange Pill Cycle
Further reading
- Paul Polak, Out of Poverty: What Works When Traditional Approaches Fail (Berrett-Koehler, 2008)
- Practical Action, Poor People's Energy Outlook (annual reports, 2010–present)
- Satish Kumar, You Are, Therefore I Am (Green Books, 2002)
- Barbara Wood, E. F. Schumacher: His Life and Thought (Harper & Row, 1984)