Intent Cannot Be Fully Specified Up Front: Why Specifications for Open-Ended Tasks Cannot Be Completed Before Exploration
Examines why intent cannot be fully specified in advance when generative AI is used for open-ended tasks—how concrete outputs shape judgment, why local feedback cannot be promoted directly to global rules, and how exploration creates new problem dimensions, reorganizes intent, and can ultimately invalidate the overall framework—while also considering the same structure in human communication and joint activity.