one of the first things you learn when you start writing mathematical proofs is the principle of explosion: from a single false premise you can derive any conclusion, true or false. interesting to reflect on how this has shaped my thinking and orientation towards truth
Conversation
the principle of explosion gives falsehoods this viral infectious quality - make a single mistake in a proof and nothing you write from that point on can be trusted. this is why cranks think they have a 1-page proof of the riemann hypothesis etc, all it takes is one mistake
3
91
it’s kind of amazing from this pov how well humans can function while believing things that are wildly false. there’s a whole thread of rationality that consists of being offended by this observation. “muh dutch book arguments” etc.
2
83
mostly i don't think it's too hard to account for this though. most people just don't believe things in anything like the way mathematicians believe mathematical statements, for many reasons. beliefs in the wild are highly contextual and/or for signaling
Quote Tweet
one of the most unpleasant things i learned about myself from my years among the rationalists was that i basically had two completely separate sets of "beliefs," one of which i could actually act on and one of which was almost purely for signaling games
Show this thread
1
67
i do think though that the principle of explosion correctly points towards the consequences of holding a false belief sufficiently *rigidly* and *globally*. if there's a thing you've decided is always good or always bad, and you're committed to ignoring exceptions...
2
5
68
...then the principle of explosion suggests, correctly imo, that there's no upper limit to how insane your beliefs or your behavior can get, starting from that false, rigid, global premise. the more universally and rigidly you apply your pet falsehood the worse things get
3
10
86
ime this sort of thing happens because the false, rigid, global belief is psychologically load-bearing in some way, it would be psychologically devastating somehow to admit that there were any exceptions whatsoever, it would break some kind of faith you have in something
2
5
54
That's the psychological feeling. I think the underlying computation just gets converted to a stuck 1 or 0 bit which is way way more energy efficient and therefore is a local minimum that it is hard to get out of.
1
1
6
I wish I had an evocative term for this. Something like the false compression of 'god did it' which leads you away from the messiness of trying to find a real compression.
1
5
Replying to
the neural annealing post from QRI talks about "dead neurons" which i liked as an evocative precise metaphor
qualiaresearchinstitute.org/blog/neural-an
