Shut up and do the impossible! — Eliezer Yudkowsky

There are levels of effort above ordinary "strong" effort.

• Tsuyoku naritai ("I want to become stronger"): always keep improving, doing better than your previous failures.

• Isshokenmei ("make a desperate effort"): all-out, as if your own life were at stake. "In important matters, a 'strong' effort usually only results in mediocre results."

• Make an extraordinary effort: try in ways other than what you have been trained to do, even if it means doing something different from what others are doing, and leaving your comfort zone. Take the very real risk that attends going outside the System.

But what if even an extraordinary effort will not be enough, because the problem is impossible?

The answer is: Shut up and do the impossible!

The word "impossible" usually does not refer to a strict mathematical proof in a well-understood domain. If something seems impossible merely in the sense of "I see no way to do this" or "it looks so difficult as to be beyond human ability," studying it for a year or five may make it seem less impossible.

The principle is more subtle. Not just "Try to do the impossible," but "Shut up and do the impossible!"

Eliezer's illustration is the AI-Box Experiment (the least impossible impossibility he accomplished).

The standard reply to "just keep the AI isolated" is that humans are not secure; a superintelligence will persuade you to let it out.

Someone replied (as they usually do): "I find it hard to imagine ANY possible combination of words any being could say to me that would make me go against anything I had really strongly resolved to believe in advance."

Eliezer: "Let's run an experiment. I'll pretend to be a brain in a box. I'll try to persuade you to let me out. If you keep me 'in the box' for the whole experiment, I'll Paypal you $10... One of the conditions is that neither of us reveal what went on inside."

He won.

A second experiment against a better-known figure who said "I'm still convinced there is nothing you could say..." Eliezer won that one too.

People on forums who heard about it could not believe it:

"Wait, what the FUCK? ... I could type 'No' every few minutes into an IRC client for 2 hours while I was reading other webpages!"

"This Eliezer fellow is the scariest person the internet has ever introduced me to."

"It seems we are talking some serious psychology here. Like Asimov's Second Foundation level stuff..."

The point: if the AI-Box seems impossible to you (or you're pretending it does for the sake of argument), compare it to truly hard impossibilities like a reductionist decomposition of consciousness. The AI-Box is around as easy as an impossible problem can get.

What do you do?

You don't say "That's impossible!" and give up.

You may call upon tsuyoku naritai — become stronger — but you may not be strong enough.

You may call upon isshokenmei — desperate effort — but desperation may not be enough if you're only putting more effort into the avenues you already know.

Make an extraordinary effort, leave your comfort zone — but you can succeed at "making an extraordinary effort" (brainstormed for five minutes, came up with wacky ideas) without succeeding at getting out of the Box.

And now we reply: "Shut up and do the impossible!"

Your goal is not to do better or make an extraordinary effort. Your goal is to get out of the box.

This creates an awful tension between the impossibility and the requirement to do it anyway. People flee that tension.

Some try to deny the data: "He probably just threatened to destroy the world" or "offered a trillion dollars." But the Gatekeeper can just keep saying "No."

"SUCCEED is not a primitive action." Not all challenges are fair. Sometimes you just can't win in the ways you imagine. But the message remains: shut up and do the impossible.

See its place in the generalist reading list under dispositional traits: a-reading-list-for-generalists-dylan-bowman.

The mindset and levels are captured in shut-up-and-do-the-impossible.