Categories
Uncategorized

Sweet, an opportunity to use my Epistemology tag!

Slate Star Codex:

“I’m increasingly uncertain that confirmation bias can be separated from normal reasoning.

Suppose that one of my friends says she saw a coyote walk by her house I know there are coyotes in the hills outside Berkeley, so I am not too surprised; I believe her.

Now suppose that same friend says she saw a polar bear walk by her house. I assume she is mistaken, lying, or hallucinating.

Is this confirmation bias? It sure sounds like it. When someone says something that confirms my preexisting beliefs (eg ‘coyotes live in this area, but not polar bears’), I believe it. If that same person provides the same evidence for something that challenges my preexisting beliefs, I reject it.”

No, you’re not wrong to do this; you’re using your beliefs for their proper purpose: making judgments about the world. The whole reason you have a belief that polar bears are extremely rare or non-existent in Berkeley is so that if you think you see a polar bear, you’ll look again more carefully, or that if your friend says “Polar bear!” you’ll consider that she might be playing a joke on you, etc.

The point of having beliefs is not just to have them. It’s to use them to guide yourself through the world. You use them to, e.g. make judgments about how likely it is that your friend is lying or playing a joke on you, etc.

Furthermore, it’s a known fact that people sometimes joke, lie, are mistaken, etc. What entitles you to dismiss that fact? If you believe your friend, you’re abandoning your well-founded belief that people sometimes say false things AND your well-founded belief that there are no polar bears in Berkeley. That’s a weird decision to make.

If you disbelieve your friend, you are retaining your well-founded beliefs that people sometimes say false things and that there are no polar bears in Berkeley. That seems sensible, given the monstrously large number of times humans are observed to say false things, and the large number of times you’ve failed to observe any polar bears in Berkeley.

If I said I saw gnomes dancing on my roof, what would you actually do? Slightly raise your probability that there are gnomes, or significantly raise your probability that I’m a jokester?

2 replies on “Sweet, an opportunity to use my Epistemology tag!”

It’s a Bayesian thing. You have a set of priors about the probability of a coyote or the probability of a Polar bear in her area. Then there’s an estimate of how reliable the observer is—is she one of the people currently making hard liquor fly off the shelves during this pandemic?
There’s nothing logically impossible about a polar bear wandering the streets. Maybe he escaped from a zoo. One lion around Orlando has escaped like 3 times from the park where he’s nominally an attraction. But he’s pretty improbable.
Priors matter tremendously, and people with radically different priors often can’t even understand each other.

For instance, someone like myself who believes that God exists and meddles with dice rolls in great events would look at a ridiculous event in history and see God’s hand. (For instance, if you wrote the story of Cortez conquering the Aztecs as a fictional book in some world where it didn’t actually happen, people would cry bullshit, its just too improbable, but nevertheless it happened). But if you have a prior of God either not existing or never meddling, you’d never conclude that regardless of how improbable it was.

Like

“It’s a Bayesian thing.”
Exactly.
“You have a set of priors about the probability of a coyote or the probability of a Polar bear in her area. Then there’s an estimate of how reliable the observer is…”
Yeah. Another example regarding reliable reporters is literal reporters. I know they’re psychopathic liars and left-wing propaganda agents, so I simply discount anything they say that swings to the political left. It’s much more probable that they’re lying than that they’re telling the truth.

Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s