LLMs are second to none in generating somewhat believable BS, while being mediocre to absolutely abysmal in executing actual tasks. Remind you of anything?