AI agents routinely skip content when reading documentation, grabbing just enough to generate an answer. The entities with the most technical leverage to fix this have the weakest financial incentive to try. The technical writers who care the most about the content are being systematically removed. So who's left to make sure the AI gets it right?
