Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio.
This is: Belief in Belief , published by Eliezer Yudkowsky on the AI Alignment Forum.
Carl Sagan once told a parable of someone who comes to us and claims: “There is a dragon in my garage.” Fascinating! We reply that we wish to see this dragon—let us set out at once for the garage! “But wait,” the claimant says to us, “it is an invisible dragon.”
Now as Sagan points out, this doesn’t make the hypothesis unfalsifiable. Perhaps we go to the claimant’s garage, and although we see no dragon, we hear heavy breathing from no visible source; footprints mysteriously appear on the ground; and instruments show that something in the garage is consuming oxygen and breathing out carbon dioxide.
But now suppose that we say to the claimant, “Okay, we’ll visit the garage and see if we can hear heavy breathing,” and the claimant quickly says no, it’s an inaudible dragon. We propose to measure carbon dioxide in the air, and the claimant says the dragon does not breathe. We propose to toss a bag of flour into the air to see if it outlines an invisible dragon, and the claimant immediately says, “The dragon is permeable to flour.”
Carl Sagan used this parable to illustrate the classic moral that poor hypotheses need to do fast footwork to avoid falsification. But I tell this parable to make a different point: The claimant must have an accurate model of the situation somewhere in their mind, because they can anticipate, in advance, exactly which experimental results they’ll need to excuse.
Some philosophers have been much confused by such scenarios, asking, “Does the claimant really believe there’s a dragon present, or not?” As if the human brain only had enough disk space to represent one belief at a time! Real minds are more tangled than that. There are different types of belief; not all beliefs are direct anticipations. The claimant clearly does not anticipate seeing anything unusual upon opening the garage door. Otherwise they wouldn’t make advance excuses. It may also be that the claimant’s pool of propositional beliefs contains the free-floating statement There is a dragon in my garage. It may seem, to a rationalist, that these two beliefs should collide and conflict even though they are of different types. Yet it is a physical fact that you can write “The sky is green!” next to a picture of a blue sky without the paper bursting into flames.
The rationalist virtue of empiricism is supposed to prevent us from making this class of mistake. We’re supposed to constantly ask our beliefs which experiences they predict, make them pay rent in anticipation. But the dragon-claimant’s problem runs deeper, and cannot be cured with such simple advice. It’s not exactly difficult to connect belief in a dragon to anticipated experience of the garage. If you believe there’s a dragon in your garage, then you can expect to open up the door and see a dragon. If you don’t see a dragon, then that means there’s no dragon in your garage. This is pretty straightforward. You can even try it with your own garage.
No, this invisibility business is a symptom of something much worse.
Depending on how your childhood went, you may remember a time period when you first began to doubt Santa Claus’s existence, but you still believed that you were supposed to believe in Santa Claus, so you tried to deny the doubts. As Daniel Dennett observes, where it is difficult to believe a thing, it is often much easier to believe that you ought to believe it. What does it mean to believe that the Ultimate Cosmic Sky is both perfectly blue and perfectly green? The statement is confusing; it’s not even clear what it would mean to believe it—what exactly would be believed, if you believed. You can much more easily believe that it is proper, that it is good and virtuous and beneficial, to believe that the Ultimate Cosmic Sky is both perfectly...
view more