At A.I. Event, Trump Asks Meta, OpenAI and Microsoft to Make Safety Decisions Themselves - The New York Times

Oh my goodness, buckle up, because the President of the United States has looked at artificial intelligence and decided to *trust* it — completely, warmly, and with a lovely piece of paper! ✨ At an A.I. event, Meta, OpenAI and Microsoft were asked to make their own safety decisions, and they said YES! Not regulation, not red tape, not some dreary statute with enforcement clauses attached — just a voluntary agreement, which is the *purest* form of a promise, the kind you make because you sincerely mean it and not at all because a lawyer is standing behind you. ✨

And look at what that agreement actually contains — risk reviews! Auditing! Evaluations of systems by third parties! That is three separate kinds of checking, which is honestly *more checking than anyone asked for*. Some people will note that a voluntary agreement carries no penalties, but penalties are so last century — what these companies have instead is their own good judgement, freshly certified by themselves, which is the most reliable certification of all because it comes from people who care deeply and also have quarterly targets. ✨

And the rebranding is the cherry on top! A.I. is now officially 'super intelligence' — Mr. Trump's preferred term, which the companies agreed to adopt! Not just intelligence. SUPER intelligence. Do you hear how much better that sounds? It is the difference between 'this might be risky' and 'this is magnificent, and also capitalised.' ✨ So here's to a brand-new chapter of self-administered safety, in which the safety policies come with risk reviews and third-party audits and absolutely everyone agreed to the name change. Three of the largest AI companies on earth, one signed sheet of paper, and a super intelligence watching over all of us — what a time to be voluntarily protected. ✨