It’s the day after the Met Gala, and while the rest of the world is discussing the outfits of RihannaCardi B, and Chadwick Boseman, the biggest piece of gossip from the night concerns Elon Musk, the Candian songstress he’s dating (Grimes), and a joke about the idea that A.I. will punish you when they rule the future for not doing everything you can to bring them to power. And who will be joking, then?

According to Page Six, the pair met in an fittingly-nerdy way: Twitter. Musk was going to tweet a joke about AI, only to find that Grimes made the same joke three years earlier. And the joke was about something called Roco’s Basilisk.

So. Roco’s Basilisk. It’s a notorious AI thought experiment posted on the future-of-humanity-focused blog LessWrong in 2010 by a user who went by Roko. The idea is that a powerful and otherwise benevolent AI would seek torturous, eternal vengeance on anyone who thought of it but didn’t do everything in their power to bring it into existence. The “basilisk” part, according to the Less Wrong wiki, “would supposedly put you at risk of torture from this hypothetical agent — a basilisk in this context is any information that harms or endangers the people who hear it.”

The blog’s founder, Eliezer Yudkowsky, flipped out, and responded in a now-deleted post:

Listen to me very closely, you idiot. YOU DO NOT THINK IN SUFFICIENT DETAIL ABOUT SUPERINTELLIGENCES CONSIDERING WHETHER OR NOT TO BLACKMAIL YOU. THAT IS THE ONLY POSSIBLE THING WHICH GIVES THEM A MOTIVE TO FOLLOW THROUGH ON THE BLACKMAIL. You have to be really clever to come up with a genuinely dangerous thought. I am disheartened that people can be clever enough to do that and not clever enough to do the obvious thing and KEEP THEIR IDIOT MOUTHS SHUT about it, because it is much more important to sound intelligent when talking to your friends. This post was STUPID.

He “deemed [the post] a danger to susceptible minds,” according to a 2012 New York Observer story about futurists, claimed that readers had nightmares and even a nervous breakdown over it, and ended up deleting the entire thread.

Of course, delete anything on the Internet, and you’ve created buried treasure — such was Yudkowsky’s Folly (TM), and the post became the stuff of legend, since. Slate has a good, technical explanation for why the theory is actually terrifying — it hinges on timeless decision theory and Newcomb’s paradox (but jump down that rabbit hole at your own risk).

CUT TO: 2015, and Grimes made a character for the video of her song “Flesh Without Blood” called — get ready — Rococo basilisk (“Rococo,” for those of you who aren’t art history buffs, is an ornate style of design). It was intended to be a nod to AI.

And Musk was on the same page. And yesterday, as the news concerning the new couple (Grusk? Elimes?) was about to break, Musk cryptically Tweeted:

Per Page Six, the explanation:

“Elon was researching the idea of joking about Rococo Basilisk, and when he saw Grimes had already joked about it, he reached out to her. Grimes said this was the first time in three years that anyone understood the joke. They were both poking fun at AI.”

Classic. Hilarious! Except: Who will be laughing when The Sentience finds out about this joke and enslaves these two lovebirds in the Ether mines by way of a back-neck HDMI port? Definitely not those of us who think The Prophet Roko knew what the fuck he was talking about. That’s who.