= 27d567bc2d48d9f40e9ccc20ea370b15)
chaosmosis comments on The Fabric of Real Things - 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 (305)
I think we must have different definitions of evidence.
In general, as all of reality is entangled with itself, it's pretty rare for something to be precisely zero evidence for any hypothesis -- it would indicate that P(H|E)=P(H|~E) which is by itself a very specific point in probability space--- it's much more common for something to be so weak evidence that for all practical purposes a person wouldn't know how to evaluate impact and/or would be wasting time to attempt to do so...
E.g. "The baby I'm thinking of was born on a rainy day. Is this evidence for or against the baby being named 'Alex' ?". A really intelligent and informed agent would be able to correlate the average raininess of various regions in the world with the likelihood of the child being named Alex in said region, and thus the raininess would be evidence for the baby's name...
In that case, the fact that I believe something different than wedrifid is evidence against his point.
Wedrifid's preferences aren't literally zero percent evidence for his view of the universe, but they're probably less than one in a googol. My overall point is pretty clear. The nitpicking is annoying, and it seems to me like it's being done because people don't want to change their beliefs.
Your argument doesn't really make sense. You say that although agents might not know how to evaluate the probability of an action's occurrence, the evidence is still unlikely to be perfectly balanced. But probabilities are subjective, and so the fact that you don't know how to evaluate a piece of evidence indicates that for all functional purposes the evidence is perfectly balanced.
Generally agreeing with your point -- it is nitpicking. Still it may be good practice to remember to say and think "that's not usable evidence" instead of "that's not evidence".
Yes, mine is along the lines of "The stuff that would cause correctly implemented Bayesian agents to update". Yours is one of the more socially defined kinds. You will note that the evidence is described as weak evidence, relative to the post that should resolve your confusion entirely. And weak evidence is what it is, no more and no less.
It's ridiculously weak. Even mentioning it is an inefficient use of time and cognitive resources.
To elaborate a little bit more, humans believe all sorts of stuff that is dumb. I don't have any particular reason to believe you're invulnerable to this. I think that your arguments about the implied invisible don't pay rent, and they complicate the way that I define things as real or unreal, so they're not worth believing in. Your attempt to use emotionally laden thought experiments instead of real arguments about epistemology furthers my skepticism about your beliefs. The fact that you insist on using your own preferences as evidence for their justification makes your credibility go even lower, because that evidence is so obviously weak that the only reason you would mention it is if you're extremely biased. Mentioning it is a waste of time; I get no predictive power out of the belief in my daughter who I will never see again.
You might as well argue for the weak existence of an afterlife by saying that you prefer the idea that your daughter still exists somewhere. I suppose that this should make me update my probability that there's an afterlife by about one in a googol, but it's also pretty clear that you're privileging a hypothesis that doesn't deserve it.
Eliezers, not mine. And they may not "Pay Rent" but they do, evidently, stand between you and the murdering of your siblings. They also constitute a strictly simpler model for reality than the one you advocate.
Again, not my attempt, it was multiple other people who were all patiently trying to explain the concepts in a way you might understand.
The arguments were real. They are what you rejected in the previous sentence. This may or may not mean you actually read them.
I wasn't, and I included careful disclaimers to that effect in both comments. I was merely making an incidental technical correction regarding misuse of the word 'evidence'. You made a point of separating your "intuitive judgement" from your abstract far-mode ideals. When people do this it isn't always the case that the abstract idealized reasoning is the correct part. I often find that people's intuitions have better judgement---and that is what I see occurring here. Your intuitions were correct and also happen to be the side of you that is safer to be around without risk of being murdered.
Fortunately, current engineering technology is such that your particular brand existence-denial does not pose an imminent threat. As has been mentioned, if we reached the stage where we were capable of significant interstellar travel this kind of thing does start to matter. If there were still people who believed that things magically disappeared once they were sufficiently far enough away from that person then such individuals would need to be restrained by force or otherwise prevented from taking actions that they sincerely believe would not be murder---in the same way that any other murder attempt is prevented if possible.