Most AI education teaches people how to use tools. AI Without Bosses asks who owns and governs the systems behind those tools.
The course, built for the co-op movement and others interested in solidarity enterprises, moves beyond familiar AI debates about prompts, productivity and safety. It looks at AI’s supply chains, data monopolies, infrastructure, labor conditions and ecological costs, then asks how cooperative and solidarity-economy models might create alternatives.
That framing matters because AI is not only a technology adoption problem. AI is an ownership stack: data, compute, data centers, labor, energy, models, platforms and governance. If that stack remains concentrated, the productivity gains will remain concentrated too.
The course also brings in Majority World perspectives, which are too often missing from AI policy. Much of the debate is dominated by U.S. and European firms and regulators, even though the effects of AI infrastructure, labor and data extraction are global. Training data, content moderation, mineral supply chains, energy use and digital labor all connect AI to places far beyond Silicon Valley.
The phrase “AI without bosses” names the power question directly. Who gives orders in an AI-mediated economy? Who decides what systems are built? Who benefits from automation? Who is made legible to machines, and who can contest the result? These are not only technical questions. They are ownership questions.
Cooperative AI alternatives remain early. Shared data, public digital infrastructure, cooperative data centers and democratic oversight all sound promising, but each requires governance capacity, capital and technical expertise. A course will not build the infrastructure by itself. But it can help create a community of people who understand the problem at the right level.
The connection to broader solidarity-economy work is important. It suggests that the conversation is moving from critique to institution-building. The co-op movement is not simply saying dominant AI is extractive. It is beginning to ask what alternative systems would require: ownership forms, governance rules, financing, education and technical collaboration.
Education becomes a form of pre-institutional work. Before cooperative AI infrastructure can be built, people need a shared vocabulary for data, compute, labor, governance and ownership across the whole system. Without that shared language, alternatives remain scattered.
The risk is that cooperative AI remains too small relative to the scale of corporate AI. The opportunity is that concentrated systems create demand for alternatives. As AI becomes more embedded in work, communication and public life, the need for democratic infrastructure will become more obvious.
If AI is reorganizing production, information and power, then shared ownership cannot stay at the edge. It has to move into the stack. That means asking not only how workers use AI, but how workers and communities might own the systems that shape their future. The political stakes are clear. If AI becomes a universal layer of economic life, then democratic ownership cannot remain limited to small businesses and co-ops on the margins. It has to reach the infrastructure itself. Education is therefore not peripheral. It is how movements prepare people to build the institutions that concentrated AI will not build for them.