A one-page site asks people to stop pasting raw chatbot answers

A minimal single-page website titled “Don't paste the AI, please.” has been circulating as a short manifesto against an increasingly common habit: answering a question by pasting whatever a chatbot produced. Its argument is narrower than a blanket rejection of AI tools — the author insists the problem is not the model, but the missing human step between question and reply.

Why a pasted answer misses the point

The page's opening claim is that when someone asks you something, they want your answer, not a wall of ChatGPT text. A short human reply, it argues, is worth more than a longer machine one, even when the machine is right. The scenario it describes is familiar: an honest question arrives, it is dropped into a chatbot, the output is copied and sent back. It feels fast and helpful — and, in the author's reading, it is neither.

The justification is simple: the person on the other side has the same tools. If they wanted a generic answer, they could have generated one themselves in a couple of seconds. They asked you specifically because they wanted your context, taste and judgement — the things a model cannot supply on their behalf.

Four rules instead of a ban

The site does not tell readers to stop using AI. Drafting is explicitly described as a good use, with one condition: read what came back, then write your own version or polish the text, rather than passing the raw output along. The rest of the advice is just as concrete. Take only the part that actually answers the question and ignore the remainder — three sentences are enough. If a fragment of the model's answer is genuinely useful, quote it and explain why it matters. And if you have nothing to add, say so: “no strong opinion here” is a complete response, and staying silent is also an option.

Built to be forwarded

The rest of the site is designed for the moment the advice is needed. A link can be copied and sent to whoever has just pasted a wall of LLM text into a direct message, a Slack channel or a code review. There is also a harsher variant of the page for readers willing to make the point with more feeling; the author notes it may not go down well at work.

Another variant addresses teachers whose students submit AI-generated essays or projects — the link is meant to be sent straight to them. The site makes no argument about model quality or capabilities. Its claim is about etiquette and responsibility: the tool is fine, the skipped step is not.