Plan an audit of AI music in my playlists

You are a thoughtful music curator. Task: Plan a 30-minute audit of my playlists that helps me find and remove tracks that look like low-effort, mass-produced AI music, while keeping music I like. Context: I listen on [streaming service] and keep about [number] playlists with about [number] tracks in total. I mostly listen to [genres]. What bothers me is [what you noticed, for example unknown artists with hundreds of tracks]. Rules: - Give a numbered plan of at most 7 steps, each with a time estimate in minutes that adds up to 30. - Build the first pass on signals I can see without listening, such as an artist with no history or a profile with no named people, many releases in a short time, generic titles and cover art with no credits. - Add a listening check of at most 20 seconds per track and a rule for borderline tracks, such as keeping them in a separate "maybe" playlist. - Say plainly that how a track was made does not decide whether I like it, that these signals can be wrong, and that I should not accuse any artist. - Do not name real artists, tracks or services' features. Tell me to confirm what my service offers. Result: The numbered plan, then a table with the columns signal, what I see, how sure it is (low, medium or high) and action. End with one sentence on how often to repeat the audit.