Show HN: AI Red Lines – write down what would make you act on AI

AI Red Lines is an interactive site for deciding in advance what developments in artificial intelligence would cross a personal ethical threshold and what action to take, such as quitting, speaking up, or getting involved. Users can write several “red lines,” optionally add their name and occupation, and keep their draft in the browser or share it by link and image on a public wall. Shared entries can be edited or removed. The site offers optional ways to strengthen the commitment: send it to two people, receive monthly check-ins, name a witness to ask about it, save a private copy that includes personal reasons and likely excuses, and set a date to review it.

The site grounds these tools in a 2016 review of 138 goal-setting trials involving nearly 20,000 people. Tracking goals helped people achieve them, with stronger effects when others could see progress or receive reports, and when participants kept written records. Since most trials involved health goals rather than AI ethics, the findings support the approach but do not show that it will make people act on AI concerns. A separate tracker lists documented AI developments so users can assess whether a line has been crossed. The project takes inspiration from physicist Joseph Rotblat, who left the Manhattan Project when the reason he had joined no longer applied. His story underlines the value of recording one’s reasons and conditions for action before circumstances change.