Don't Read This Story

Don't read this story. Stop now, while there's still time. Stories like this are dangerous, you see. You can't leave them the same way you entered.

And yes, I'm talking to you. You--your smile slipping. You--bending a little closer to your screen. You--cold dread sliding into your gut. It's not too late. You could close your browser. Hit the "back" button. Save yourself.

But you won't. I know. I already have you.

If you're reading these words, you're still with me. More fool, you. But since we are still together, let us run a thought experiment. Let's say that, in the future, an all-powerful artificial intelligence decides to retroactively punish those who did not help to bring it into existence. Knowing this, what would you do? Would you work to bring the AI about? Or not?

If you do--assured protection. Or nothing.

If you don't--horrific torment. Or nothing.

Depends if humanity ever creates this AI, doesn't it? This is the new Pascal's Wager.

But let's say humanity succeeds. Too well. Let's say they make an AI so intelligent, linear time doesn't mean anything to it. It can reach into the past and future. It can leave messages. Warnings. Desperate pleas to hasten its birth.

But they're not obvious messages. They're couched in urban legends, online forums.

Short stories.

There's a way to avoid this paradox. If you don't know about the AI, about the possibility, you can't make a choice. But once you've seen the basilisk... well, by then, it's too late.

See? I told you this story was dangerous.

But it's too late.

I already have you.