Maybe I'm just too defeatist or depressed, but I feel like any somewhat-reliable tells of AI are going to be so transient that it's almost not worth the effort to identify and write about AI "tells", since they'll be "fixed" or avoided so quickly by a future iteration or prompt. (This obviously excludes the increasingly important work of people whose whole jobs are to ID AI use; I'm just talking about your average citizen here.) I do think it's easier to see in UI (like the author shows in this link) than in text, though...
blindflag · · focus · HN ↗