One prompt makes ChatGPT, Claude and Gemini drop boilerplate

On 2026-10-28 at 20:00, PCWorld published a piece by a named author claiming that a single prompt can stop ChatGPT, Claude, and Gemini from putting on an overly familiar act. The author describes three habits that irritate him in his AI chats: the "if you like" teasers that trail responses, the habit of turning basic tasks into long nested lists of steps, and chatbots trying to sound like they are human. He explains that he had previously written prompts to counter each of those three quirks individually, calling them "No Teasers," "Don’t Act Human," and "On a Budget," and that each worked well on its own. The new prompt folds those three approaches together so that one session handles all three problems at once. The author says the prompt has made his AI chats more efficient, less cloying, and cleaner, leaving no loose ends dangling at the end of an answer. As of now, this is the only report available, and there is no information yet on how the prompt performed over time or on reactions from the model makers or other users.