Sam Altman: This is how we can make AI safe for everyone · HackerLangs