Hein de Haan
2.8K posts
Hein de Haan
@Heighn
Sentientist and writer, interested in morality, decision & game theory, history, and all things math and science.
The Netherlands Katılım Şubat 2011
753 Takip Edilen1.4K Takipçiler
@marc__adler @DeivonDrago That does not get you time travel though, right?
English
First of all, this is wrong.
If Einstein was right about relativity (which he was), we can travel into the relative future. That is the entire basis of the twin paradox.
Now, do we have the propulsion systems to go that fast over that much time? No, but that’s a tech issue.
Danny Jones@JonesDanny
"If Einstein was right about relativity, we will never have time travel." Physicist Dr. Sylvester James Gates explains why time travel is impossible based on our current understanding of the universe:
English
English
@Heighn @lmldias @robertskmiles @peach2k2 But why should it matters? If it's a random guess, the accuracy is 50%. If it's using some correlation factor, it should be higher. With an accuracy of only 51%, one box is already preferred (higher expected value). But not only that, the actual premise is that it's accurate.
English
@corsaren @_AashishReddy "No matter what DT you use, it is always possible to construct a problem set up where being rational is punished, so that’s not really a flaw"
Is it? Is there one for FDT?
English
@_AashishReddy 1. Two-boxing is a dominant strategy, and 2. No retrocausality means that CDT is essentially correct (barring Kantian arguments for FDT)
No matter what DT you use, it is always possible to construct a problem set up where being rational is punished, so that’s not really a flaw
English
I do accept that, strictly speaking, it is rational to two-box and the design of Newcomb’s is such that it punishes rational actors.
But also, you should just one-box. There is no set of norms that cannot be undone by a demon hellbent on punishing you for following them.
English
Hot take: "Two-boxing is rational, but you should one-box" is a contradiction.
corsaren@corsaren
I do accept that, strictly speaking, it is rational to two-box and the design of Newcomb’s is such that it punishes rational actors. But also, you should just one-box. There is no set of norms that cannot be undone by a demon hellbent on punishing you for following them.
English
@NathanpmYoung It doesn't, actually. If the predictor models your decision procedure, the correct answer is to one-box in the problem as described.
English
Something that cuts at the disagreement:
Arjun Panickssery@panickssery
@robertskmiles @peach2k2 I've committed to one-boxing if the situation appeared, but if I hadn't, then it would be rational to two-box the correct answer is to two box in Newcomb's Problem as described because there's no retroactive ability to commit to acting, so two-boxing dominates one-boxing
English
@fmgs31 @lmldias @robertskmiles @peach2k2 My point is that accuracy is not the only relevant factor here. The basis of the prediction matters as well.
English
@Heighn @lmldias @robertskmiles @peach2k2 Well, all this buzz created by the Veritasium video, which states the limit case: a predictor that's almsost 100% accurate, doesn't matter how. But even if it weren't, it only has to be 51% accurate to still be preferred to take one box given the expected value.
English
@fmgs31 @lmldias @robertskmiles @peach2k2 That actually matters a lot! If the predictor actually knows how you make your decision and bases her prediction on that (i.e. it models your decision procedure), then you should one-box. If she uses some correlating factor, it's just Smoking Lesion and you should two-box.
English
@lmldias @robertskmiles @peach2k2 It doesn't matter how it does it, the premise is that it's accurate.
English
@robertskmiles @lmldias @peach2k2 Yes, but this does depend on how the predictor makes her prediction. If she bases it on the color of your shoes (in a hypothetical world where wearing blue shoes strongly correlates with one-boxing) you should two-box, regardless of your shoe color.
English
@xriskology Please demonstrate why utilitarianism leads to this conclusion.
Also, EA is not necessarily utilitarian in nature.
English
100%. Example number one million of why utilitarianism is a terrible moral theory.
David Z. Morris@davidzmorris
A spectacular example of the barren ethics of Effective Altruism in the wild. Anthropic taking a minimal, reasonable ethical stance is *unethical,* ackshully!
English
@realtwitcourt "Klopt dat ik goed verdien". Precies, en dat is een van de redenen dat ik niet doneer: het deel van mijn geld dat ik doneer maakt elders een (veel) groter verschil.
Nederlands
@PstafarianPrice @americnhumanist @gm_skeptic I'm trying to understand your point here. I mean, surely victims of (attempted) assassination can still be bad guys?
English
I can't express how disgusted I am with @americnhumanist, and whoever @gm_skeptic is, for this disgusting advertisement
"They're after me!" it claims, framing Trump and Kirk, two LITERAL VICTIMS OF ASSASSINATION as... the bad guys? Jesus Christ.
English
@skdh Whether it makes intuitive sense also strongly depends on the interpretation. MWI makes much more sense to me (and is not at all counterintuitive to me) than the Copenhagen Interpretation.
English
Also, most of the quotes from physicists who supposedly said that no one understands quantum mechanics don't exist.
The one person who clearly said this was Richard Feynman, but it was also clearly a joke.
It is frequently attributed to John Wheeler, but I found no evidence that Wheeler said anything of the sort.
There is a second-hand quote (related through Heisenberg) that Niels Bohr said if you aren't shocked by quantum mechanics you don't understand it.
The only other verifiable quote I have been able to find came from Sean Carroll.
Personally I think it comes down to what you mean by "understanding". We understand the maths just fine, it's not all that difficult. Does it make intuitive sense? Depends on how much time you have spent with it.
x.com/Leophilius/sta…
English
@davemotorcycle1 (And, of course, not only that: Masley claimed she should be purged from polite society.)
English
@davemotorcycle1 The bigger point is not whether "don't associate with fascists" is a good norm; it's that Masley says Grimes went "full fascists" with his evidence being that she went to Yarvin's wedding and a party he attended.
English
quite disheartened by how much negative feedback Andy is getting here
Andy Masley@AndyMasley
Had only recently learned that Grimes went full fascist and is palling around with Yarvin now, what a fucking lunatic. Surprised she hasn’t been completely purged from polite society yet, she should be.
English
@davemotorcycle1 That is, had @AndyMasley just said that Grimes should not have gone to these events, this would have been a very different debate.
English
@ilex_ulmus @NathanpmYoung Out of curiosity, what is your view regarding being friends with people who eat meat?
English
@NathanpmYoung Ah you were so close. Yes you should cut off people involved in AI— I call it #UnfriendAI
English
Seems like taking this seriously, AI worriers should cut off all friendships with those involved in AI. That seems (in their view) to cause more harm than fascism.
Seems like a bad equilibrium to push for.
Andy Masley@AndyMasley
Had only recently learned that Grimes went full fascist and is palling around with Yarvin now, what a fucking lunatic. Surprised she hasn’t been completely purged from polite society yet, she should be.
English
@xChessPuzzles 1. Qe5
If pawn move or Ra7 move 2. Qxb8#
If Rbxh8 2. Qxh8#
If Rb move on line 8. 2. RxR(c..g)8#
English
@Benthamsbulldog I think your EA writings are often inspiring. Your arguments for God are interesting but, I believe, ultimately incorrect. Your writings on Decision Theory seem quite confused, and you're (unknowingly) strawmanning FDT and being quite (over)confident about it.
English
My sense is that among LessWrongers, I'm a somewhat polarizing figure with some having a very low opinion of my writing and others having a pretty high opinion of it. Would be curious to hear from the people who think it's really bad why.
English