The automation dividend answers a specific failure: AI can produce abundance while simultaneously destroying the mechanism — paid work — through which most people used to reach abundance. If the new capacity stays concentrated where it is generated, total output rises while access falls, which is productive capacity up and ecological capacity flat. The dividend is the framework’s insistence that when the ecology generates enormous new capacity, that capacity cannot remain entirely trapped in the node that captured it.
The framework is plural about how the return is captured (public equity stakes, levies on automation rents, social wealth funds) and specific about what it funds: the floor credits first, then the commission funds. In the transition, the public share is secured as the price of public inputs while that leverage lasts. Its limit remains: a dividend must be extracted or distributed from whoever owns the models, compute, and output, and standing alone does not compel that. How it can be compelled is writing circulation into the rules. That is why this project sits at the seam between job displacement and ownership concentration — the dividend names what should happen, and the power to make it happen belongs to the harder scenario next door.