= 27d567bc2d48d9f40e9ccc20ea370b15)
Morendil comments on Priors and Surprise - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (30)
That's one of my main take-aways from PT:TLOS so far. (This post has reminded me of one of the exercises in the book.) The idea in Chap. 5 that given extremely unlikely priors for a hypothesis, strong evidence appearing to favor that hypothesis will not so much confirm it as call for the the resurrection of another, highly unlikely, but still more plausible than the original.
"You are being deceived" is advanced as a useful hypothesis to consider resurrecting in general. Perhaps conspiracy theorists are not as stupid as they sometimes appear; if you think the straightforward hypothesis unlikely enough, mass deception can be a more appealing hypothesis, with readily available mechanisms of explanation. (For instance a movie like The Truman Show couldn't work if we didn't assign a small degree of plausibility even to mass deception, given sufficient financial incentive.)
For others' reference, PT:TLOS = Probability Theory: The Logic of Science, by E. T. Jaynes.
Jaynes even uses the example of meteorites (aka thunderstones) to show that this line of reasoning, while valid by the laws of probability, can lead educated people to believe things that are not true.
The natural philosophers who knew something about gravity had a much higher prior probability for the unreliability of farmers' reports of natural phenomena than they did for rocks falling from the sky, so every report served to reinforce the hypothesis that farmers don't know what they are talking about. It took reports from people who were considered reliable in order for the meteorite hypothesis to be accepted.
How reliable a reporter would you need to call science in general into question?