All people know stuff about stuff, will talk (have already talked) to AI about that stuff that they know, and will figure out that AI is lying about stuff they

All people know stuff about stuff, will talk (have already talked) to AI about that stuff that they know, and will figure out that AI is lying about stuff they know. Maybe this can reduce the number of people who believe everything they read to a less dangerous level.