The plank in brief

Automation should amplify what people can do, not replace democracy. Under the old model, capital chooses first and workers get consulted later, if at all. Surveillance is invisible, people get deskilled without being asked, and the boss turns into an algorithm. In the commonwealth, workers and communities govern AI where it matters most. Each sector has a technology council of workers, professionals, consumers, public institutions, and technical experts who decide together on high-impact automation. High-risk deployments need an impact assessment first. Monitoring, staffing, deskilling, safety, and the productivity gains themselves are all on the bargaining table.

Some lines don’t move. No firing by algorithm. No unreviewable black-box performance scoring. AI is not infallible, and a person has to stand behind any decision that changes someone’s life. When automation raises productivity, the gains are split between higher pay, shorter hours, lower prices, and the social wealth fund. When it materially eliminates jobs, the workers who lost them get a negotiated share of the productive asset or its royalties. Public AI research goes where markets underinvest: care, accessibility, ecological repair, education, translation, and the dangerous work nobody should have to do. Done this way, AI handles the scheduling and the lifting so caregivers can focus on people, cuts admin so teachers can teach, and takes the deadliest jobs so workers come home safe. AI beside the worker, not above the worker.

Objections and replies

None posted yet. If you have a strong objection to this plank, send it to me. The strongest ones go up here with my reply.

Changes

• November 2026: first published here.