AIアンガーマネジメント: vent to a chatbot before you hit send

The routine is framed against the well-known advice to wait six seconds before reacting to anger — an idea the writer traces to a U-CAN column. Their complaint is practical: six seconds pass, the anger does not leave, and a half-written rebuttal is still sitting in the reply box. The anger has nowhere to go, so the writer gives it somewhere — the AI chat window — and moves the rant out of the messaging app where hitting send is one keystroke away.

The sample exchange shows what that looks like. A last-minute button-position change the evening before a release gets typed in raw; the AI answers that being annoyed is reasonable, that the change request is a fact while 'they don't take this seriously' is an interpretation, and that the one thing in the writer's control is deciding with the other person whether it ships now or next time. A second pass turns the angry draft into a version that keeps the problem and the request but drops the sarcasm and the accusation.

The writer is explicit that this is not a cure-all. They cite research suggesting that putting feelings into words can put a brake on the brain, but also point to OpenAI's reversal of a GPT-4o update that had become too accommodating and inflamed user anger, which is why the prompt asks the AI not to join in blaming anyone. For anger that lasts days, the advice is to talk to a trusted person or a professional rather than the AI.