Why people should help design the AI that affects them

Technical teams are often skilled at turning a defined problem into a working system. The difficult part is that public problems are rarely defined by technical teams alone. Residents, workers, service providers, and communities understand constraints that are invisible in a dataset.

Participatory design gives those perspectives a place early enough to change the brief. It can surface harms, reveal missing measures of success, and make the trade-offs of an AI system discussable before they become infrastructure.

I see this as a complement to engineering, not a rejection of it. Better framing produces better systems—and makes it more likely that a system earns trust over time.