I find this especially-concerning because LLMs will often serve up plausibly detailed incorrect answers, a failure-mode which requires the very same human skill

I find this especially-concerning because LLMs will often serve up plausibly detailed incorrect answers, a failure-mode which requires the very same human skills and habits which the product is busy destroying. (Or at least suppressing.) Kind of like those viruses which directly attack your immune-system.