The question is not what AI can do.It is what it should decide.

↳I have spent over a decade designing products where getting it wrong has real consequences: on someone's health, their money, the energy a country runs on. That work taught me something the AI conversation tends to skip: the problem is rarely technical. It is a problem of judgement. What we delegate to the machine, what we do not, and where the loss of human judgement starts to corrode trust. PRAXIS is the framework I use to answer that question.