quick monolith
Many ChatGPT UI bugs are repeatable and well understood by users, but hard to escalate individually, so they often linger despite costing time daily.
Idea: allow ChatGPT to collect structured, opt-in UX/bug signals, aggregate them across multiple verified users, and only surface candidates for internal triage once a confidence threshold is reached — with a required human verification step before anything becomes a real ticket.
This wouldn’t auto-create issues, just highlight high-signal, time-costly UX problems (including common repro patterns and user-discovered workarounds).
Curious if others see value in this approach.