I don't think it's (necessarily) equally dumb. Maybe if comparing LLM output to a book chosen at random. But I would feel much safer taking a passage from Knuth at face value than a comparable LLM passage on algorithms.