Talk:LessWrong

From RationalWiki
Jump to: navigation, search
Icon internet.svg

This Internet related article has been awarded BRONZE status for quality. It's getting there, but could do with improvement.

This article is of MID importance to the wiki.

See RationalWiki:Article rating for more information.

Copperbrain.png


This page is automatically archived by Pibot
Archives for this talk page: <1>, <2>, <3>, <4>, <5>, <6>, <7>, <8>, <9>, <10>, <11>, <12>, <13>, <14>, <15>, <16>



Contents

Similarities between LessWrong and Objectivism

Some similarities that stuck my mind:

  • Both are outspoken atheists (I never said they were all bad).
  • Both preach reason. Or at least, their versions of it.
  • Both have problems with mainstream science. Objectivists have declared modern physics to be "corrupt", LessWrongians have pushed fringe science such as singularitarianism and cryonics.
  • Both have/had cultish leaders.
  • Both think that the world could soon be at an end and that they must do something. For Objectivists it's regulation of capitalism that could bring the end, for LWers it's the supposed imminent coming of the singularity.
  • Both have their writings which will Change Your Life. For LWers it's Yudkowsky's sequences, for Objectivists it's Atlas Shrugged.--Baloney Detection (talk) 08:06, 20 May 2012 (UTC)
Hmm. Another data point: I was reading about Ron Hubbard (the starter of scientology) on wikipedia and tbh the parallels are amazing: dianetics the science of mental health, raising the sanity with it to save the world from nuclear war, etc. The nuclear war was a real credible threat, but it was used the same. Dmytry (talk) 10:20, 20 May 2012 (UTC)
I don't know enough about Scientology to be honest, but it looks entirely plausible.

I'm thinking more LaRouche, at least in the followers - someone writes a shedload of stuff that's not obviously wrong and someone else goes "EVERY WORD IS GENIUS!" Though EY does better than LaRouche or Hubbard in writing stuff that isn't actually wrong to a non-phygist - David Gerard (talk) 13:03, 20 May 2012 (UTC)

You forget the bit while EY goes 'every brainfart is true' about himself first, with the AI crap and other stuff of this kind. I have no sympathy for that. The cult is no random coincidence, it starts with being your own cultist first. People want to feel good about themselves, some are 'strong' narcissists, they can form delusions of own importance, some are 'weak' narcissists (that could be the majority!), they want to join a cause where they'll feel more important. Dmytry (talk) 13:50, 20 May 2012 (UTC)
Actually, it seems to me that the personality cult of Yudkowsky isn't really focused on his supporters on LW, no matter how sycophantic they may be, and more between a bunch of his critics on this talk page. You seem to be hanging on every single possible comparison and criticism, completely regardless of whether its either warranted or relevant. "Oh look! He's got a Jewish background!" "Oh look! There's parallels with L.Ron Hubbard!" - thing is, we can draw parallels between L.Ron Hubbard and J.K. Rowling if we tried hard enough, it doesn't mean its relevant or even remotely applicable to anything at hand. This is almost at Cult of Schalfly levels, its borderline insane. Scarlet A.pngnarchist 16:59, 21 May 2012 (UTC)
Never seen mentioned the Jewish background, I don't know what background Hubbard has, and I don't care. The point is one has 'dianetics' other has some sorta 'bayesianism' in the making, neither really makes any sense (Bayes rule is good math and everything, but you can't just throw it at incomplete inference graph with cycles), both talk about saving the world, and the same sort of people joins similar causes. — Unsigned, by: Dmytry / talk / contribs 18:20, 21 May 2012 (UTC)
Hitler had a personality cult too. And he had his writings which will Change Your Life. WěǎšěǐǒǐďWeaselly.jpgMethinks it is a Weasel 18:22, 21 May 2012 (UTC)
This discussion strikes me as similar to "atheism is a religion!!!!!" ones: you really have to stretch to find similarities between them and it really doesn't say much even if they were similar. Cow...Hammertime! 18:31, 21 May 2012 (UTC)
Some people's atheism IS a religion, though. Dmytry (talk) 18:32, 21 May 2012 (UTC)
Care to elaborate on exactly what is the "stretch"? I didn't feel I stretched the similarities between LW and Objectivism, and of course there are differences (the former mostly refuse to talk politics, the latter are happy to do so). But you can't deny that there are clear similarities. And no, LW is nothing like Nazism (btw, who talked about EY's Jewish background?). You just did a Godwin.--Baloney Detection (talk) 19:17, 21 May 2012 (UTC)
Doh.gif ωεαşεζόίďWeaselly.jpgMethinks it is a Weasel 19:22, 21 May 2012 (UTC)
Care to explain why someone upon learning of Roko's basilisk should NOT point and laugh at LW?--Baloney Detection (talk) 20:03, 21 May 2012 (UTC)
Huh? ΨΣΔξΣΓΩΙÐWeaselly.jpgMethinks it is a Weasel 21:13, 21 May 2012 (UTC)
I would not point and laugh if not for the monetary aspect of it. If they were living entirely off the Thiel's tit, it'd be ok also. But they are taking advantage of mentally unstable. The person that lost sleep over the basilisk was a donor to SI btw. Dmytry (talk) 06:43, 22 May 2012 (UTC)
That's a fair point. Though Roko's basilisk indicates to me that Yudkowsky is an honest madman, he really believes his own stuff. But of course it doesn't make it any less bad.--Baloney Detection (talk) 22:51, 25 May 2012 (UTC)
One thing they sort of got right is the 'dragon in the garage' parable, when it comes to beliefs like that... the roko thing was also an insult because it leads very directly to 'my FAI idea sucks', and that hurts the ego, the reaction is adequately explained without assuming the degree of belief poorly compatible with monetary success. Dmytry (talk) 08:58, 29 May 2012 (UTC)
"Though EY does better than LaRouche or Hubbard in writing stuff that isn't actually wrong to a non-phygist" You can find glaring errors if you look more deeply. Such as Bayes' theorem being opposed to science, the pseudoscience (singularitarianism, cryonics) they peddle etc. And of course Roko's basilisk. "they want to join a cause where they'll feel more important" Clearly true of the SIAI, whose members apparently believe they are going to save the world.--Baloney Detection (talk) 18:02, 21 May 2012 (UTC)

Richard Loosemore

Who is he? ( http://lesswrong.com/user/Richard_Loosemore/ http://www.richardloosemore.com/bio ) What I can find is this: http://www.sl4.org/archive/0608/15895.html , banned for some disagreement about "conjunction fallacy". From even skimming the wikipedia article on conjunction fallacy http://en.wikipedia.org/wiki/Conjunction_fallacy you can see that it is a misnomer: people do answer it correctly when presented in 'how many people out of 100' form instead of 'how probable it is', likewise people don't expect 6 coin tosses to give FHHFHF to be more common than sequence of 3 coin tosses FHH, so it is clear this fallacy is not robustly a feature of conjunction itself. It can be more likely that "a bank teller active in the feminist movement" is Linda, than it is likely that "a bank teller" is Linda, and people often use 'is' the wrong way in language. People also use the questions themselves as evidence, when you are asked of probability of a very detailed scenario, assuming non manipulative circumstances the scenario did come from some detailed analysis. (Ironically the best counter example and demonstration that there's something wrong with the way people do conjunction would be the AI shit at LW, which is an enormous conjunction in which they are nonetheless pretty sure). Dmytry (talk) 12:28, 24 May 2012 (UTC)

Never heard of him, though he seem to be pretty in to the standard pablum of LW woo (singularitarianism). I guess his crime was not to conform enough to the LessWrong phyg/cult standards.--Baloney Detection (talk) 22:58, 25 May 2012 (UTC)

Not completely unaware

Shit Rationalists Say - David Gerard (talk) 23:20, 1 June 2012 (UTC)

I don't really get why anyone would want to be a rationalist. I'd much rather want to be an empiricist. These terms do have historical roots, but presumably a knowledge of history isn't a virtue for a Rationalist. I can understand the general criticism that RW is snarky, but that's not just to LW, that's to everyone. I'd very much like to see how a more proper skeptical resource, say the Skeptic's Dictionary, would take on LW. For a starter, SD isn't friendly to cryonics: http://www.skepdic.com/cryonics.html --Baloney Detection (talk) 17:12, 10 June 2012 (UTC)
(Claimed) neuroscientist on LessWrong goes what the hell are you people on about with this freezing brains rubbish. Extensive thread follows in which he goes "no really, you're fucked, this is actually ludicrous." It's anonymous and I don't know enough to write it up, but wow, if he's on the money he makes our cryonics article actually look startlingly optimistic - David Gerard (talk) 17:46, 10 June 2012 (UTC)
Cryonics is ludicrious, that's not just his view, it's the mainstream view in medicine. Shermer wrote a good piece about it ( http://www.michaelshermer.com/2001/09/nano-nonsense-and-cryonics/ ) and QuackWatch considers it to be, well, quackery: http://www.quackwatch.com/04ConsumerEducation/QA/cryonics.html --Baloney Detection (talk) 17:53, 10 June 2012 (UTC)

Reactions to LW elsewhere

Some random forum's reaction to LW: http://www.poe-news.com/forums/sp.php?pi=1002430709 http://www.poe-news.com/forums/sp.php?pi=1002405574 The next time they complain rationalwiki is bad, you can link this...

Kinda fascinating. Though from which perspective do they come from? Do they just throw dirt at everything and everyone on the Internet?--Baloney Detection (talk) 19:57, 13 June 2012 (UTC)
No idea... seems like a fairly normal forum, making fun of conservapedia and liking rationalwiki : http://www.poe-news.com/forums/sp.php?pi=1002364960 . Also in LW thread someone explained interferometer mistake. If i search for random bashing I tend to find the stuff that well deserves bashing, like http://www.poe-news.com/forums/spshort.php?pi=1002307636&ti=1002307636 , not Justin Bieber/Lady Gaga/etc bashing. Dmytry (talk) 09:00, 14 June 2012 (UTC)

LessWrongian misunderstandings of science

Part of the goal of RW is to:

  • Analyzing and refuting pseudoscience and the anti-science movement.
  • Documenting the full range of crank ideas.

Given this, what misunderstandings of science does Yudkowsky (and thus most LessWrongians) have that you can think of? What runs through my mind is that Yudkowsky and most of his acolytes believe that science and Bayes' theorem are in conflict with each other.

What LessWrongians also for the most part don't get is that you still have to perform the experiment. You can't simply gloss it through with a naive application of Bayes' theorem.--Baloney Detection (talk) 17:51, 4 June 2012 (UTC)

Well, frequent misuse of the terms like Solomonoff induction and Kolmogorov complexity would be an example of pseudo-mathematics (as would most references to Bayes and Bayesian something). They just use math as technobabble. In the cognitive science they promote the view that human mental function can be replicated with very very little computational power compared to what is commonly believed to be the lower bound (i.e. basically they promote the views of most optimistic AI researchers of the 1950s). The pseudoscience is not about incorrect predictions or incorrect theories; the pseudoscience is about incorrect process. What they promote is an incredibly sloppy process of reasoning. They promote the very idea of believing in results of sloppy thought on terms you poorly understand. Back when I was still trying to contribute constructively, I would speak against the sloppiness: http://lesswrong.com/lw/ai9/how_do_you_notice_when_youre_rationalizing/5y3w . Note that the edit only added expanded explanation when it was at -1 or -2. This was at -7 or -8 peak, before someone wondered why. You could say that Eliezer is a moron and get fewer downvotes. This illustrates what they are against: the notion that you need rigorous/careful reasoning to be correct. They believe some fuzzy ill defined 'rationality' suffices (and not only suffices, but is superior). And also they don't like experiments. Dmytry (talk) 06:39, 5 June 2012 (UTC)
This post makes me question how well Yudkowsky really understands probability theory. My conclusion is, not very well. So what if "Positive and negative infinity are not integers"? Yeah, they aren't integers, but they are affinely extended integers. So what if odds ratio gives positive infinity for 1? We need not limit ourselves to the reals; we can use extended reals instead. Same for his silly log odds argument. "Using the log odds exposes the fact that reaching infinite certainty requires infinitely strong evidence, just as infinite absurdity requires infinitely strong counterevidence." So? What's wrong with the idea we have infinitely strong evidence for some propositions? How strong is the evidence for 1+1=2 - infinitely strong or only finitely strong? If only finitely, can you quantify it? I think not - which supports the position that we have infinite evidence for this proposition. Yudkwosky needs to say, what is the probability that Bayes Theorem is true? And should Bayes Theorem not be modified to account for the non-zero probability that it is wrong? Zack Martin HolyMaratreanSigil.png 12:47, 5 June 2012 (UTC)
In addition, I would say that they have a poor understanding of human capabilities and intelligence in general. Not woo-like 'everybody is a genius,' just the normal sort of capabilities we can see from observation of humans... any human. Although yes, people are more than capable of being stupid, the biggest idiot in the world is still very intelligent. Human perception and analysis is very very complex and very nuanced and we're capable of all sorts of amazing things that we take for granted. Visual recognition and pattern recognition are ridiculous in the natural world even for the proverbial left pointy end of our bellcuve, for example. We have our glitches and we have societal pressures that change our conclusions (don't get me STARTED on their dismissal of social sciences!) but on the whole we're pretty impressive as far as animal life on this planet goes. The belief that humans are naturally 'stupid' (but not them; they think in the magic golden formulae) I find is often used as a strange hubris, and causes the crux of their science fail. The belief that mastering human-level intelligence in an AI is attainable within the century, the idea that they don't have to test things, the dismissal of non-secret-clubhouse-people and their perhaps well-supported conclusions or objections... they all spring from the 'humans are dumb' falsehood that supports the massive egos of at least a few there. It just takes one bad idea to give rise to a host of bad ideas. Just one concept. Even in otherwise smart people. ±Knightoftldrsig.pngKnightOfTL;DRlavishly loquacious 12:55, 5 June 2012 (UTC)
"don't get me STARTED on their dismissal of social sciences!" But evo psych is a social science...right? Nebuchadnezzar (talk) 16:34, 5 June 2012 (UTC)
I find they are often all-too willing to use evolutionary psychology to confirm their position that humans are stupid, but comparatively less willing to address social issues that guide the thoughts of individuals and groups just as much as (if not more!) than genetics do. I usually see much less emphasis on education (how can people understand their philosophy if they aren't actually taught critical thinking skills?), on understanding cultural reasons why people believe what they do. It goes along with the 'politics is the mindkiller' idea; they quite often avoid actual issues that are relevant to their discussion in order to hold their own position. ±Knightoftldrsig.pngKnightOfTL;DRmore at 11 00:44, 6 June 2012 (UTC)
I was being more than half sarcastic -- forgot the dick quotes around "science." The problem is not that they think humans are stupid, but that they think "non-rationalists" are stupid. Either that or they think "rationalist" is not a subset of "human," which I find to be less plausible, though they could be confusing their imagined future selves with their current selves. Nebuchadnezzar (talk) 01:35, 6 June 2012 (UTC)
Kind of both. Obviously not all of them think this, but I always walk away with a taste of 'We have overcome our humanity, we are better than normal humans' even out of the gate, no AI required. Like their rationalism is a hack or a fix for all of the things that come with not only being homo sapiens, but also being a participant in whatever society they individually happen to belong to. Also, sorry... I am not so good at sarcasm on the internet. I had a feeling but then the urge to type happened. :( ±Knightoftldrsig.pngKnightOfTL;DRsufficiently advanced argument still distinguishable from magic 03:50, 6 June 2012 (UTC)
My impression is that they confuse play-pretending to be Spock with gaining powers of correct reasoning. The scientists don't aspire to play-pretend to be Spock and are therefore more wrong. Dmytry (talk) 13:09, 7 June 2012 (UTC)
But look at the results when you shut up and multiply! (The process was: 1. Read LessWrong article. 2. Decide that superior rationality means you can make a bundle just showing up in Australia with a working visa. 3. Be surprised when reality doesn't cooperate. Hopefully they're having decent tourist experiences.) - David Gerard (talk) 22:11, 7 June 2012 (UTC)
Ghahaha... perfectly illustrates my point. The fool haven't done even 1 percent of the work required for declaring a country optimal for employment, he only did very incomplete work with serious rational Vulcan face on. All of the rationality there works like this: Do some insufficient or entirely flawed reasoning with serious Vulcan face on, and expect to be magically better than all the 'non rational' folks. Dmytry (talk) 06:34, 8 June 2012 (UTC)
No, Maratrean, that's you being a bloviating fuckwit again with startling delusions of relevance. What he means is that there is no such thing as philosophical certainty, thus 0 or 1 are not probabilities; this does not rule out practical certainty, where the difference is pretty much epsilon. Your problem with log odds of probability is your problem, they work perfectly well and make the numbers in Bayes' Theorem easier. tl;dr Shut the fuck up. You're a net negative. Just shut the fuck up. - David Gerard (talk) 19:11, 5 June 2012 (UTC)
Ya the article looks ok to me, elementary math. If EY has troubles with statistics, that's at more abstract level of suggesting that people should be doing those 'Bayesian updates' on incomplete graph with cycles where the new nodes are created based on probabilities of existing nodes, and where two nodes that affect probability of 1 node may, in turn, be both affected by third node, which would get double counted unless everything is tracked. Which is BTW where science is 'at odds' with his understanding of Bayes because scientists know math better and know not to risk double counting evidence and circular-updating. Also btw really awesome for accepting only arguments you like: if you like it proclaim you 'update' on it, if you don't like either call it non-bayesian or argue evidence may be double counted. That really helps with beliefs like the AI stuff. Dmytry (talk) 21:35, 5 June 2012 (UTC)
@DavidGerard — the definition of "probability" is not directly connected to philosophical certainty; an event with probability 1.0 is not necessarily philosophically certain, and can even fail to happen — almost surely happening and surely happening are not the same thing. And I have no problem with log odds; if there is anyone here who has a problem with them, it is Yudkowsky not me. Zack Martin HolyMaratreanSigil.png 01:43, 6 June 2012 (UTC)
@Dymytry: Where does he argue that science is "at odds" with Bayesian statistics? I don't see how he can even make that argument, esp. since he is in AI/cog sci where Bayesianism is all the rage these days. Nebuchadnezzar (talk) 01:47, 6 June 2012 (UTC)
I'm not Dymytry, but I'll answer anyways. Here he does argue exactly that: http://lesswrong.com/lw/qa/the_dilemma_science_or_bayes/ --Baloney Detection (talk) 08:09, 6 June 2012 (UTC)
He argues against Science (with a capital S), which is a word he's using for the social process of science (which of course has many fucked bits), not against the notion of science itself. (Beware negative halo effect.) - David Gerard (talk) 13:47, 6 June 2012 (UTC)
My impression is that he just doesn't like the fact that science doesn't assign some probabilities to unfalsifiable bullshit which he is full of when he's trying to do something new (such as argue about AI). His idea is that under Bayesianism, people should assign some probabilities to what ever he makes up (or what ever idea he takes from science fiction and presents as serious concern of his). That's obviously faulty because you go from effectively zero probability (which is the probability you have effectively assigned to every hypothesis that you haven't considered), to some nonzero number, which is then either too low and you can be gamed by giving bullshit arguments for true facts so you set them low probabilities, or too high and then you can be gamed with straight bullshit such as AI risk argument of his. Also note the reference to many worlds. He really doesn't understand jack shit about either Kolmogorov complexity or QM. The notion of lowest Kolmogorov complexity of objective model within which you can find yourself, does not favour many worlds. It favours the iterator over all possible physical theories. Or even worse, simple iterator over all numbers (within which you'll find yourself as a sort of Boltzmann brain). You have to discount for size, i.e. you have to discount against huge theories that predict everything at once. The MWI might or might not be most useful interpretation depending to what sort of physics you are doing. Also, his idea of how physicists understand CI is a strawman, he doesn't understand the whole point of why it is called 'interpretation', and as pointed out by others he probably confused phase of 180 with complex number i when doing the interferometer example. Dmytry (talk) 15:02, 6 June 2012 (UTC)
@David Gerard: Did you actually read the article? He did put science and Bayes against each other, and asked which one the reader "pledges allegiance" to. He was strongly criticized for his lack of understanding of science by philosopher of science Massimo Pigliucci ( http://rationallyspeaking.blogspot.se/2010/09/eliezer-yudkowsky-on-bayes-and-science.html ). To my knowledge Yudkowsky has never responded to that, and probably never will (he is usually too full of himself to respond to critics). As we all know, Yudkowsky has several other problems with science when it comes to his pet issues, for example the fact that mainstream science is ambigous to the singularity hypothesis and dismissive of cryonics.--Baloney Detection (talk) 16:30, 7 June 2012 (UTC)

Because editing long sections is hard on my phone

I have to agree with Mara on this one, that post reeks of does not understand the concept. But the non-zero concept of Bayes thm being incorrect confuses me, unless he is engaging in pointless pyrrhonism again. Pi 3:14 (talk) 02:00, 6 June 2012 (UTC)

No, my point is — if Yudkowsky wants to argue that no proposition has probability 1, then it follows P(Bayes theorem is true)<1.0. So, should not Bayes theorem be made more accurate by taking into account the non-zero probability of its own incorrectness? (I'm not claiming it could be wrong; but if Yudkowsky is to be consistent then he must claim that.) Zack Martin HolyMaratreanSigil.png 02:14, 6 June 2012 (UTC)
Hey Mara, SHUT THE FUCK UP! LongStandingUser (talk) 02:54, 6 June 2012 (UTC)
But then you have to take into account the incorrectness of what you used to assess the correctness of Bayes' theorem, and then the incorrectness of that assessment, and the incorrectness of that... or you can just say that these mathematical statements are proven assuming your formatulation of number theory is correct and that your underlying logic that builds up mathematics is correct. Otherwise, what you're merely saying is that "because you can't be 100% sure, therefore you are actually 0% sure and ANYTHING can happen, right?" The post you have a problem with is a fairly hand-wavy tutorial on philosophical certainty based on analogy to odds ratios. Scarlet A.pngbomination 08:13, 6 June 2012 (UTC)
An easy way out of infinite regresses is to assign certain assumptions a probability of 1. Anyway, part of the problem with Yudkowsky's post is he seems not to understand the relationship between probability theory and philosophical certainty. Probabilities of 1 and 0 are not necessarily philosophical certainties; an event with probability 0 can happen and an event with probability 1 can fail to happen. Zack Martin HolyMaratreanSigil.png 08:19, 6 June 2012 (UTC)
No, if you set a probability to 1 or 0 you get a division by zero, but you're too obsessed with the structure of your own anus to notice you're still talking. The problem is that you are still talking. You have nothing to say on the topic and you are still saying it. Shut the fuck up - David Gerard (talk) 13:04, 6 June 2012 (UTC)
I'm no math expert (in fact I am a math idiot) but I was under the impression that it's not that 'things with probabilities and 0 and 1 happen' it's that in some cases the probability can approach 0 or 1 infinitely but never reach 0 or 1. Meaning that something can be so close to 0 it's pretty much 0. Unless the rules of the cosmos suddenly change. Which may be even less probable. So you might as well relax; for all that it matters to us there may as well be 0 and 1 probability because the alternative is so staggeringly unlikely that it makes 'improbable things happen' look like a coin flip. I could be very wrong, I don't remember where I heard this; if I am, please correct me. This is kind of a plea for clarification because learning stuff is cool and I am a math idiot. ±Knightoftldrsig.pngKnightOfTL;DRlavishly loquacious 13:30, 6 June 2012 (UTC)
Yes, that's the difference between philosophically possible (the egg might unscramble itself) and practically possible (the egg is not going to unscramble itself). If you read anything whatsoever by Maratrean, you will end up less informed than you were when you started, per this - David Gerard (talk) 13:47, 6 June 2012 (UTC)
I was under that impression, yes. Iiinnntteerreessttinnng. I find this stuff way more fascinating than I should, possibly because I'm not naturally great at math and hearing it explained is cool. Except what you said about maratrean. I was pretty aware of the uncoolness of that. :S ±Knightoftldrsig.pngKnightOfTL;DRwalls of text while-u-wait 13:51, 6 June 2012 (UTC)
Talking of mathematical stupidity but how is the probability of throwing a number between 1 and 6 using a standard die not exactly 1. similarly I'd like to hear how the probability of throwing a seven is not zero. Bad Faith (talk) 13:46, 6 June 2012 (UTC)
No one going to mention The Last Hero in answer to this one? Scarlet A.pngd hominem 16:23, 6 June 2012 (UTC)
The atoms in the die have a probability greater than zero (but on the order of ten to the minux so many I don't even have enough numbers for it) of spontaneously rearranging themselves to form a seven on one face. This is of course not going to happen, and for conventional uses you can ignore it. The trouble is that humans are very bad at ignoring it (as Yudkowsky noted in the excellent "But There's Still A Chance, Right?") and conventionally behave as though any probability greater than zero must be non-negligible, which is of course wrong. Which is why people reply "No. No chance." But this is technically incorrect, and it's occasionally important to note it's technically incorrect - David Gerard (talk) 13:54, 6 June 2012 (UTC)
It's particularly important when you are making a living by pascal-mugging people using incredibly tiny risks (or the risks you made up which are difficult to estimate) to big words like Mankind. Speaking of which, another instance of making incredibly low probabilities: long sequences of sloppy inferences. Dmytry (talk) 14:51, 6 June 2012 (UTC)
@DavidGerard, You don't divide by zero. If you are using Bayes theorem, P(B) can't be zero, and so if it is you can't use Bayes theorem. But there is a lot more to probability theory than just Bayes theorem. It's funny how you combine cluelessness with a continual insistence that I "shut up". You are anger bear. Zack Martin HolyMaratreanSigil.png 20:30, 6 June 2012 (UTC)
Maratrean, I can't find DG being either clueless or especially even saying once you should shut up. So shut the fuck up. (Just kidding, I don't want to shut anyone up on the RWiki, but your accusation is way out of line.) ħumanUser talk:Human 03:12, 7 June 2012 (UTC)
He is clueless; for example, he accuses me of having a problem with log odds when I don't, and there is no evidence that I do. And he seems to think that "division by zero" somehow makes zero or one probabilities invalid, when it doesn't. It actually doesn't matter what I say, he'll just tell me to shutup anyway. Zack Martin HolyMaratreanSigil.png 08:45, 8 June 2012 (UTC)
There is in fact no evidence whatsoever that you have ever said anything useful, relevant or informative on any topic at any time, and considerable evidence that you consistently confuse your ignorance with knowledge and then insist on enlightening the world with it. You're a waste of electrons, time and oxygen. Just shut the fuck up, you bloviating prick - David Gerard (talk) 11:58, 8 June 2012 (UTC)
Wow, you are an amazingly hostile person. What is wrong with you? If you aren't interested in what I have to say, why not simply ignore me? Maybe someone else will find my contributions more valuable than you do. Zack Martin HolyMaratreanSigil.png 13:05, 8 June 2012 (UTC)
That would only be possible if they were a bloviating fuckwit in waiting. As such, Non-Bloviating-Fuckwit Decision Theory proves mathematically that they should be ignored at the least, and preferably actively hampered for the good of humanity - David Gerard (talk) 14:32, 8 June 2012 (UTC)
I don't expect anyone to agree with everything I say, but the fact that sometimes people (other than you) agree with me indicates that they find some value in at least some of my contributions. Like Pi beginning this section with "I have to agree with Mara on this one"... is Pi a bloviating fuckwit in waiting ? Zack Martin HolyMaratreanSigil.png 08:13, 9 June 2012 (UTC)
If you said the sky is blue, this would not be evidence that your existence has a fucking point - David Gerard (talk) 10:32, 9 June 2012 (UTC)
I think everyone's existence has a point — even yours. You seem to be a very negative and hateful person. Zack Martin HolyMaratreanSigil.png 11:16, 9 June 2012 (UTC)
If you have a purpose, it's as what not to be. I love everyone, except you - David Gerard (talk) 12:17, 9 June 2012 (UTC)
Wow, you are something special. "I love everyone, except you" Zack Martin HolyMaratreanSigil.png 12:43, 9 June 2012 (UTC)
@David Why do you feel this way? I feel this way ('confuse ignorance with knowledge') about much what Eliezer is doing with quantum mechanics, Kolmogorov etc. especially the AI, knowing nothing and going on - quite explicitly so - how he's the most rational person. Dmytry (talk) 11:57, 12 June 2012 (UTC)
Large parts of existence as a human are about judgement under uncertainty - but there are times when people should damn well know that their ignorance isn't data (and yes, LW does it a lot). It's why nerds are nuts - David Gerard (talk) 23:14, 12 June 2012 (UTC)
@David Gerard : Engineers are the sane ones compared to non-engineers like Eliezer. I question your consistency - how exactly is Maratrean worse than a guy who actively promotes such 'judgement under not knowing jack shit'? The guy who takes money for work on 'friendly artificial intelligence'? The guy whose writings are the very pinnacle of being a 'bloviating fuckwit'? I've read some of Maratrean's essays, nothing under any reading gets even remotely close to EY. Dmytry (talk) 05:17, 13 June 2012 (UTC)
EY hasn't spent a couple of years posting tl;dr blether to RW - David Gerard (talk) 09:37, 13 June 2012 (UTC)

@ David Gerard And Maratrean hasn't spent more than a decade promoting pseudoscience on the internet at large. The real pseudoscience, the one that looks scientific but isn't, not the lame variations that don't even look scientific. EY doesn't just promote woo/bs, he creates woo (out of science fiction), promotes it, and promotes art of woo creation (aka 'rational' speculation without attention to any subtleties of the subject). He also created an enormous battery of 'cognitive biases' allowing to label any sort of reasoning which takes into account subtleties you don't know of, as cognitive bias (technically, being correct is having a biased distribution, too). Even if you actually make a demonstrable fallacy, there's the way out, via this very argument that the opposite side is just too good at knowing fallacies, for their own good (forgot what that article was called). Dmytry (talk) 11:49, 13 June 2012 (UTC)

Isn't that what they refer to as "undiscriminating skepticism"? Basically, if you are an atheist but not a Yuddite, you are an atheist because you hang out with moderately educated people who are mostly atheists. And similarly, people who are skeptical of cryonics are so because of tribal signalling due to Penn and Teller. Here: http://lesswrong.com/lw/1ww/undiscriminating_skepticism/ Do these guys seriously not see the big beam in their eyes? What is this ( http://lesswrong.com/lw/cfe/i_stand_by_the_sequences/ ) if not tribal signalling? Or even more than that, a list of articles of faith?--Baloney Detection (talk) 20:36, 13 June 2012 (UTC)
Yup. They basically slap the label "improving art of human rationality" onto "bloviating about irrationality of those who disagree with things for which we don't even have a well defined argument". There's nothing about how to actually finding the truth. The truth should magically appear if there's no biases. Is this actually a very bad and offensive comment: http://lesswrong.com/lw/ai9/how_do_you_notice_when_youre_rationalizing/5y3w ? What I am to think of getting -5 (-9 +4 or so) on explanation of rigorous reasoning and the accumulation of reasoning errors in fuzzy reasoning? (that was before I ever flamed their AI bullshit) Dmytry (talk) 09:50, 14 June 2012 (UTC)

Some thoughts on the 'bayesian' rationality.

I've put some of my thoughts there: Essay:LessWrong_Bayes_and_Rationality (i hope it is not against the rules). Basically, when you have incomplete inference graph with cycles, doing some theoretically optimal shit that works only on trees is a pretty damn stupid idea, along the lines of believing that you can walk off 8th floor and just keep on going because you just learnt of the awesome concept of inertia and ignored such low and specific details of Earth environment as gravity. — Unsigned, by: Dmytry / talk / contribs

Bayesianism as opposed to Frequentism

What's the hell is up with that local trope? I don't get it. You get Bayes rule from frequentist reasoning, that's how they derive Bayes rule themselves. I've heard it said that frequentism is somehow "bad epistemology" , what ever that means. The whole thing looks like a case of horrible fuzzy confusion of philosophy with mathematics.— Unsigned, by: Dmytry / talk / contribs

Both frequentism and Bayesianism are useful approaches in different circumstances — choose the right tool for the job. The idea that there is a single all-round best approach to reasoning is silly in my opinion. Zack Martin HolyMaratreanSigil.png 11:21, 9 June 2012 (UTC)
Shut up, you bloviating fuckwit. Your statement says nothing, provides no predictive power and is a perfect example of extrapolation from personal ignorance - David Gerard (talk) 12:17, 9 June 2012 (UTC)
If I'm playing a dice game, and want to calculate the odds of some particular throw, frequentist methods are useful, Bayes rule isn't really useful. As I said, use the right tool for the job. Zack Martin HolyMaratreanSigil.png 12:52, 9 June 2012 (UTC)
Yudkowsky got that trope straight out of Jaynes. Modern Bayesian textbooks (e.g. the one I'm ploughing through in fits and starts) point out that Bayes is a theorem, hence the mathematical facts; that the entire frequentist cookbook can be derived by Bayesian means; and that the big problem with pure Bayes is knowing what the hell your prior actually is in the first place. Bayes as ideal, frequentist cookbook as practical methods - David Gerard (talk) 12:17, 9 June 2012 (UTC)
Well it sounds to me like 'deriving' fluid dynamics from Archimedes principle. If you have classical laws of physics, you can derive probabilities in any dice experiment, along with deriving the Bayes rule, by taking the probability as limit of the frequency (and counting over the observations that ended up compatible with knowledge). The priors are your freedom-to-be-wrong when all you care for is not getting dutch-booked (being internally consistent), but do not care to be correct about any system. You don't get to choose arbitrary values when you want to be correct about physical system, but you do end up with a plenty of arbitrary values that you can set to be anything you want when you only care for some relations between your predictions. Dmytry (talk) 15:10, 9 June 2012 (UTC)
I know Bayes rule is often called a "theorem", but I'm not sure that is really appropriate. Attempts to demonstrate it run into difficulty, and many people seem not to bother. Jaynes attempt to justify Bayes rule follows Cox's approach, which has problems. Zack Martin HolyMaratreanSigil.png 12:52, 9 June 2012 (UTC)

Bob Carroll of Skeptic's Dictionary on LW (extremely briefly)

Here is an e-mail I sent him, and his reply. The hotlinks in the original e-mails are not included in this reposting (except the written-out link to LW). My original e-mail below:

Dear Dr. Carroll

I wonder if you would consider LessWrong ( http://lesswrong.com ) to be a trustworthy, reliable site, if you are familiar with it. It is based on the writings of Eliezer Yudkowsky whom you have a reference to. They write a lot about rationality, the human mind, and cognitive biases. But on the other hand, they also promote cryonics, the singularity and other topics of dubious scientific validity (Yudkowsky claims that parents who don't sign up their children for cryonics are "lousy parents").

Is it a site you would consider linking to, or a site you would dismiss as woo?

Sincerely yours, [Name]

And his reply below here:

woo

thanks.....

Bob Carroll The Skeptic's Dictionary Subscribe to our newsletter Try the SD app for iPhone Skeptic's Dictionary for Kids new! Unnatural Acts: Critical Thinking, Skepticism, and Science Exposed!

I hope my e-mail was fairminded enough. I tried to include both positive and negative aspects of LW in it, while still keeping it as brief as possible. --Baloney Detection (talk) 20:19, 10 June 2012 (UTC)

Speaking of which - you can maybe ask Holden Karnofsky if he knew of Roko thing when he was evaluating SI. As far as charity evaluation is concerned Roko thing is unforgivable sin - apparently you're supposed not to think through the faults in the moral philosophy of the FAI. However you look at it, it's total frigging idiocy. Actually I doubt he heard of Roko affair before he written review; if he did that would of saved him quite a bit of time. Dmytry (talk) 21:47, 10 June 2012 (UTC)
I've never heard of him, but Googling and checking his site, they appear to be in bed with the Yuddites.--Baloney Detection (talk) 21:28, 12 June 2012 (UTC)
Really? He criticized them pretty harshly: http://lesswrong.com/lw/cbs/thoughts_on_the_singularity_institute_si/ Dmytry (talk) 21:30, 12 June 2012 (UTC)
I see he did. My mistake, though at his own site he seemed more positive toward them. Perhaps I'll send him an e-mail.--Baloney Detection (talk) 15:10, 17 June 2012 (UTC)

Archive?

I think we should archive much of this talk page, given its length--il'Dictator Mikal 20:20, 10 June 2012 (UTC)

It shall be done. Peter Blessed are the cheesemakers 23:18, 10 June 2012 (UTC)
It has been done. Star of David.png Radioactive Misanthrope 00:53, 11 June 2012 (UTC)

What is bayesianism, hahahaha.

I didn't see this before:

http://lesswrong.com/lw/1to/what_is_bayesianism/

Would it be fair to say that they just reused name of a dead mathematician for some ideological thing similar in nature to 'dianetics' ? (edit: i.e. a bunch of truisms presented as awesome revelation that justifies the crazy beliefs) That's pretty clever. Everyone with a clue gets distracted to arguing fine points of statistics. And you can't describe it with a short phrase 'X sucks' without a zillion qualifiers. Dmytry (talk) 09:18, 11 June 2012 (UTC)

Much like how people who love to go on about Occam's razor often show very little or no interest in what William of Occam himself thought about the subject. Zack Martin HolyMaratreanSigil.png 20:25, 11 June 2012 (UTC)
More like how people whom go on about quantum anything typically know nothing about the actual math there and can't do an interferometer without confusing i with phase of 180. Dmytry (talk) 09:45, 12 June 2012 (UTC)
Or how people who have never opened a physics books confidently declare that MWI is "obviously correct".--Baloney Detection (talk) 20:33, 12 June 2012 (UTC)
The core tenet of "LW rationalism" (I think) : without knowing subtleties rationalists know better. I like to picture mission statements like "devoted to refining the art of human rationality" as a spring pulling together two fenceposts in the ground, one going hundred meters deep labelled "what we are doing" and other shallowly buried in the sand, labelled "what we consider rationality to be". Add tough string and the definition of rationality is sent flying. Dmytry (talk) 21:25, 12 June 2012 (UTC)
Some people seem to think that because LWians call themselves "rational", they are so. They are rational in the same way as Objectivists are, is word but not in deed.--Baloney Detection (talk) 18:29, 17 June 2012 (UTC)

A critique of the friendly AI "problem"

I wrote a short essay about the supposedly urgent problem that the SIAI has set itself to solve: Essay:Why there is no problem of friendly AI Constructive criticism very much appreciated!--Baloney Detection (talk) 22:37, 15 June 2012 (UTC)

Criticism of LW on Luke Muehlhauser's blog

In Luke Muehlhauser's (the executive director of the SIAI) now defunct atheist blog (where he started promoting LW), some interesting criticism of it came from commenters in this post[1]. Scroll down to the comment by antiplastic. I'll quote a part here:

"Yudkowsky’s manifest incompetence in technical philosophy, his utter lack of engagement with the professional computer science community, his petulant dismissal of even mild criticism from Luke, his endless invocation of the same handful of catchphrases repeated reflexively by his followers (“taboo your words!” “bayes!” “Bayes!!!” “BAYES!!!!”), the creepy interest these “bootcamps” seem to take in the moral and spiritual railroading of the participants — none of these things seems to register.
...
Speaking as the most militant atheist you will ever meet, it would have been better if Luke had remained some sort of moderated Christian with a host of silly supernatural beliefs who had nonetheless managed to outgrow the cult-mentality of fundamentalism, than to lose the former while hanging on to the latter like we’ve seen over the last few months. It’s especially painful because when he first appeared on the scene he was one of the more articulate and passionate and promising advocates for “my” side, and now he’s gone all Anakin on us."

Pretty damning!--Baloney Detection (talk) 23:42, 16 June 2012 (UTC)


Yeah, except the discourse level on that blog was for shit, and most of the whining philosophers (including the one you quote) were butthurt Christians given to obfuscatory apologetics as if they were saying something that would actually refute atheism. I appreciate you're searching for any stick you can, but this isn't a good one - David Gerard (talk) 13:21, 17 June 2012 (UTC)
I think it is true though that Luke just gone from religion into a cult. The cults prey upon religious de-converts. He's been given truth in form of EY the guru who supposedly understands some badass mathematics of truth (Solomonoff induction), except not only the guru haven't got a clue, there's formal proof that you can't just 'do' Solomonoff induction, and in any case LW is no mathoverflow, nobody would post a hard math problem here and expect an answer. Dmytry (talk) 14:26, 17 June 2012 (UTC)
That sounds like an interesting experiment, actually - David Gerard (talk) 16:59, 17 June 2012 (UTC)
Hmm. Dmytry (talk) 19:11, 17 June 2012 (UTC)
"most of the whining philosophers (including the one you quote) were butthurt Christians" Did you even read the quoted part? He wrote "Speaking as the most militant atheist you will ever meet", so no, not a Christian. A few other commenters were though. But keep sticking up for pseudoscience promoters (no, cryonics is not feasible, whatever Yudkowsky might tell you: http://www.quackwatch.com/04ConsumerEducation/QA/cryonics.html ) who have nightmares about robots.
@Dymytry: Yes I think that's true. Before Muehlhauser was into LW, he was pretty deeply into Alonzo Fyfe's desire utilitarianism (I couldn't find any difference between Fyfe's brand of utilitarianism and preference utilitarianism which is what most modern utilitarians advocate, hence I fail to see what's so amazing about Fyfe's stuff) and often wrote about it on his blog. But I think he is fundamentally smart, and as the quoted blog commenter said, his engagement with LW will probably blow over in five years or so. He is all too prone into believing he is saving the world, leading to some odd behavior.--Baloney Detection (talk) 15:06, 17 June 2012 (UTC)
Meanwhile he's making a crackpot organization be less honest and collect more money via not understanding the technical details even at the level of being a honest crackpot. Actually, now I see what is honest about Eliezer. He named his thing 'singularity institute' because all he ever worked on was hyper-singularity scenario. Add someone like Luke - entirely technically clueless, to the point of incapability of being a honest crank - and it's transitioning into a bona-fide fraud. edit: actually they have an article on that, 'belief in belief'; the EY was or is a believer in some crackpot hypotheses. Whereas Luke doesn't believe in the crackpottery, he just believes it's true. At which point he gets himself moral carte-blanche to just go ahead and optimize strings (such as name and descriptions) for the purpose of maximizing the donations, at which he has no restraint, lacking the technical understanding to be a honest crank.Dmytry (talk) 05:24, 20 June 2012 (UTC)

Yudkowsky as a crank

I checked Yudkowsky against some crank trait lists, and he certainly has more than his fair share of them.

Common characteristics of cranks (from Wikipedia[2])

1. Cranks overestimate their own knowledge and ability, and underestimate that of acknowledged experts.
2. Cranks insist that their alleged discoveries are urgently important.
3. Cranks rarely, if ever, acknowledge any error, no matter how trivial.
4. Cranks love to talk about their own beliefs, often in inappropriate social situations, but they tend to be bad listeners, being uninterested in anyone else's experience or opinions.

How Yudkowsky does, in my estimation:

1. Yes. 2. Yes. 3. Yes. 4. I don't know.

Dunno about 3, I think not quite plus people are dicks online in general. Dmytry (talk) 21:32, 17 June 2012 (UTC)

Wikipedia continues:

Many cranks:
1. seriously misunderstand the mainstream opinion to which they believe that they are objecting,
2. stress that they have been working out their ideas for many decades, and claim that this fact alone entails that their belief cannot be dismissed as resting upon some simple error,
3. compare themselves with Galileo or Copernicus (or in a religious context, Noah), implying that the mere unpopularity of some belief is in itself evidence of plausibility,
4. claim that their ideas are being suppressed, typically by secret intelligence organizations, mainstream science, powerful business interests, or other groups which, they allege, are terrified by the possibility of their revolutionary insights becoming widely known,
5. appear to regard themselves as persons of unique historical importance.

Again my estimation of Yudkowsky's performance:

1. Yes. 2. No. 3. No. 4. No. 5. Sort of. He doesn't necessarily view himself as being of unique historical importance, but he clearly views is friendly AI stuff in that light.

the fanboys do 2 for him, http://lesswrong.com/lw/cef/how_many_people_here_agree_with_holden_actually/6l5t . See +22 Dmytry (talk) 21:32, 17 June 2012 (UTC)

Martin Gardner in his Fads and Fallacies in the Name of Science[3] also had a list of crank characteristics:

The second characteristic of the crank (which also contributes to his or her isolation) is the tendency to paranoia. There are five ways in which this tendency is likely to be manifested.
1. The pseudo-scientist considers himself a genius.
2. He regards other researchers as stupid, dishonest or both.
3. He believes there is a campaign against his ideas, a campaign comparable to the persecution of Galileo or Pasteur. He may attribute his 'persecution' to a conspiracy by a scientific 'masonry' who are unwilling to admit anyone to their inner sanctum without appropriate initiation.
4. Instead of side-stepping the mainstream, the pseudo-scientist attacks it head-on: The most revered scientist is Einstein so Gardner writes that Einstein is the most likely establishment figure to be attacked.
5. He has a tendency to use complex jargon, often making up words and phrases. Gardner compares this to the way that schizophrenics talk in what psychiatrists call 'neologisms', "words which have meaning to the patient, but sound like Jabberwocky to everyone else."

My estimation of Yudkowsky's performance below:

1. Unsure. Does he? His fans certainly do. 2. Yes (in the field of AI). 3. No. 4. Yes (except he is not attacking any particular scientist). 5. Yes.

Another characteristic of the crank according to Gardner is that they work almost isolated from the scientific community (see the Wikipedia link for more on this). This is also true of Yudkowsky.

Do you disagree with my judgements on the various scores here? Are they correct? I can only base my judgements on what I have seen.--Baloney Detection (talk) 19:53, 17 June 2012 (UTC)

Well I don't consider him to be a honest crank TBH. Too much success getting money. More like a cult starter, the cranky beliefs at the level of dragon in garage. Dmytry (talk) 21:32, 17 June 2012 (UTC)
I'm conflicted about if Yudkowsky is honest or not. Perhaps he once believed it, but now only keeps it going to make a living. On the regular job market, he doesn't have much to offer really. Hardly anyone has any use for hand-waving about friendly AI. I know though that certain LW beliefs are very likely to be false, almost certainly false (like cryonics and singularity). I added a crank section on the Eliezer Yudkowsky entry, please help improve it.--Baloney Detection (talk) 21:55, 18 June 2012 (UTC)
I think he honestly believes he might be right (as in, not certainly wrong), and honestly believes he's doing nothing morally wrong, but such sort of belief usually is produced by simply not caring enough to think things through. It may also be the case that he and rest of SI don't see anything but material utilitarianism (where the utility is necessarily material in nature), see it as the only philosophical doctrine, or the best one, see anything not adhering to such doctrine not generally intelligent, and mix up their notion of intelligence with everyone else's. I.e. thorough confusion and absence of any motivation to think anything through. I imagine a clinical psychopath of the kind that would be swindling people out of money just like this, would necessarily be genuinely concerned about AI as well (edit: i.e. assuming he believes AI will be built), so the AI fear got to be genuine in some way. Dmytry (talk) 06:33, 19 June 2012 (UTC)

Question to the defenders of LW on RW

I know there are quite a few here who I think have a net positive view of LW (David Gerard and Armondikov spring to mind), and tend to water the criticism of LW down. While I'm all in favor of having multiple viewpoints around (though it does make the entry look contradictory), I'd like to know how you view the LW promotion of cryonics, singularitarianism, Bayes being opposed to science and such wooish beliefs. And of course the infamous Roko's basilisk. Do you agree with LW on these issue? Do you not? Why or why not?--Baloney Detection (talk) 18:45, 19 June 2012 (UTC)

The good bits are good (the writing on cognitive biases is good stuff, as noted) and the bad bits are bad. You appear to be having difficulty comprehending that something can have good bits and stupid bits at the same time, insisting that anything must be entirely one or the other. This is incorrect - David Gerard (talk) 19:07, 19 June 2012 (UTC)
It's quite possible that I'm falling victim to the human tendency of black-and-white thinking. But even so, do you endorse cryonics and singularitarianism? Do you believe Bayes' theorem is opposed to science?--Baloney Detection (talk) 19:10, 19 June 2012 (UTC)\
Do I endorse cryonics? Well, I wrote large chunks of cryonics and its talk page, so why don't you turn your baloney detector on the history of each - David Gerard (talk) 06:46, 20 June 2012 (UTC)
One can probably find good and true bits in how scientologists justify their nonsense. Pretty much any screwed up ideology has a bunch of truisms and generally good things as the core tenets, too. First there was the AI belief and AI charity and cryonics. Then there came the writings on biases, created during the process of rationalization of why ignore everyone else. You got any clue about the topic? Availability heuristics. You are a programmer? Of course you're wrong about Solomonoff induction b/c you're used to making software fast (and someone who probably just read some popularization book would know better and have all sorts of earth shattering God-disproving insights right away which somehow none of the mathematicians who came up with that stuff could see (note: i am an atheist, but i am sick of pseudomathematical disproofs of God just as much as I am sick of pseudomathematical proofs of God)). And so on and so forth. Also, what's so good about the biases work? I see made up pop psychology that qualifies for pseudoscience status. Biases are complicated. I see also the belief that somehow the lack of biases would lead to truth, i.e. that the study of biases is improving 'art of human rationality', etc, a baseless assumption that doesn't even make much sense (even complete absence of particular form of stupidity is not intelligence, for there is pretty much infinite number of ways to be stupid). Dmytry (talk) 22:36, 19 June 2012 (UTC)
No, LessWrong is nothing like Scientology, and that's a completely silly comparison. And trivialises Scientology the way casually comparing people to Hitler trivialises the Nazis and what they did.
I'll note here that I happen to know really lots and lots about Scientology, so I can speak knowledgeably about the comparison.
L. Ron Hubbard was an amazing bullshitter. He was pretty smart, but his real genius was knowing how to manipulate people. He was a true sociopath: empathetic but not sympathetic; both willing and able to do whatever it took to get other people to do what he wanted. He had no ethics and no sense of true or false, and frequently believed his own lies.
Scientology was founded after the failure of Dianetics: Hubbard wanted to make lots of money without work, so he came up with a popular psychology. When that collapsed (because you can't bullshit forever), he formed Scientology to hide behind cover of religion. This was enough and it did really alarmingly well, and was only finally taken out by the Internet. It still exists, but it's a cripple like Christian Science (which was the same kind of vicious in its day, though not the same degree).
Hubbard was also a terrible writer. In the pulp days, he churned out barely readable tosh that could be edited into printable form, and did so reliably in first drafts, so editors loved him. (The one good novel he wrote, Fear, was only good because John W. Campbell absolutely rode his arse to make it good.) He continued the same writing habits in Scientology. He made it up as he went along, and the contradictions, stupidities and impossibilities became a feature, as he got people to blame themselves for the failure of his stupidity. He didn't just let them do this, he made it a rule that if the text seems stupid, it's because you misunderstood a word. The mechanism was the basic religious one of "everything good is from the religion, everything bad is from me, I must try harder."
Scientology was conceived as deeply evil shit and carried forward as deeply evil shit. It was a pretty efficient machine for separating marks from their money, for the glory of Hubbard.
LessWrong isn't like any of that, in any way at all. The SIAI was an obscurity only transhumanists cared about. Yudkowsky started the blog Overcoming Bias with Robin Hanson (a non-Austrian economist at GMU who has two degrees in physics; he invented the idea of the prediction market) to blog about cognitive biases, and about what he thought about stuff, because he was blocked on writing and thought blogging every day or two would help him get this stuff out. What is now "The Sequences" was a bunch of blog posts in sequence, and didn't pretend to be anything else.
The SIAI is not a cult headquarters. It's a semi-incompetent charity, as is really incredibly normal amongst charities. It's only just getting its shit together with Luke doing the daily running. Asking for money does not make an organisation inherently evil, even if they're asking for money for things you think are incredibly stupid; this does not constitute financial abuse.
There are many lesswrongers who seem to treat it as if it were an object to worship. This is because humans are stupid, and because they're smart and socially inept kids who've never had an interest in their lives before. They usually get over it - I read all of LessWrong from the beginning to 2011 and saw names come and go on the typical online community cycle of 12-18 months, often with a dramatic door-slam. LessWrong is basically a waste of time, and is best consumed as Internet television.
Scientology is basically the Godwin example of cults; comparing any damn thing to Scientology makes actually abusive cults that are not as bad as Scientology seem benign. LessWrong is not Scientology. Not even slightly. I know what cults are, and they really aren't it. I have looked at it very closely with this precise question in mind, because a friend was getting into LW jargon-spouting and several other friends and I were actually a bit worried. (I'm not worried now, and the friend is fine and still on LW.) If Yudkowsky really wanted a cult, he's quite smart enough and knows enough about human minds and the highly exploitable biases they are prone to that he could start one. Starting a cult is not hard at all - the main prerequisite is no ethics and not caring about truth or falsity. (There are, of course, people like me on the internet who simply delight in giving cults a hard time, so the environment is not so cult-friendly unless they cut people off from the start. Examples about, but not very much on the internet.)
Their stupidities are stupid, their fans are fanboys, their good bits are fine, that's quite sufficient. The stupidities fully warrant horselaughs, as well as the 10,000 syllogisms. You're mistaking stupidity for malice, and this is leading you to make silly comparisons - David Gerard (talk) 21:26, 21 June 2012 (UTC)
Goodpost.gif Nebuchadnezzar (talk) 22:30, 21 June 2012 (UTC)
Yeah, pretty much what I'd say. There's good, there's not-so-good. I'd be lying if I said that wasn't applicable to RationalWiki. Or back when I was on Star Trek forums. Or the theatre group I work with. Or my own research group. Or my own writing. Hanging on every word Yudkowsky says and bigging up LW as some hub of extreme pseudoscience is pretty much turning this page into Cult of Schlafly Mk2. Not entirely a bad thing, of course, as RW should always aim to diversify in who it pisses off, but as DG says, putting it on par with Scientology is laughable. Scarlet A.pngbomination 23:16, 21 June 2012 (UTC)
"Hubbard wanted to make lots of money without work" Well and EY wanted to make some money without work, and isn't so talented. The LWism got the bootcamps and thought reform and every other feature of a nasty cult (complete with doomsday prophecy), which all those things Armondikov speaks of do not have, but cults do. The scientology been extremely demonized but it's just the same mundane thing at larger scale and after longer time, and done by a more talented more ambitious asshole. edit: and also, lesswrong.com is just online support group where they draw dedicated folks from, into various rationality bootcamps n stuff Dmytry (talk) 07:19, 22 June 2012 (UTC)
You can indeed find superficial correspondences. But that still doesn't make the comparison anything other than specious. You're still holding to a completely wrong track here.
As I note: if EY aspires to be a cult leader, where's the cult? He's smart, he's a good writer (certainly a hell of a lot better than Hubbard) and he knows all he would need to about human cognitive biases. LW is a terrible idea for a cult, if he were an actual sociopath (another word not to be thrown around lightly - it actually means something, it isn't a synonym for "weird" or "stupid" or "talks stupid philosophy"), he'd do a tremendously better job of it.
I don't intend to trivialise LW/SIAI's problems - there's plenty of stupidity there that needs documenting (which is why this article is here). But that doesn't require assuming the level of malice you're assuming.
Please imagine for a moment: what if LW and SIAI's stupid bits were not the products of malice, but of perfectly normal stupidity? What would the world look like? How would it look different to how it looks to you now? - David Gerard (talk) 10:22, 22 June 2012 (UTC)
Another thing to consider: if Yudkowsky were like Hubbard, he wouldn't have deleted Roko's post, but rather enthusiastically embraced the idea. Scaring your members into ponying up cash for the cause is a key feature of cults, yet he rejected this method, exposing himself to the censorship charge instead. LW also featured the results of GiveWell's audit of SIAI despite their negative impression. I highly doubt that Scientology would ever conduct a transparent review of how its leadership spends the organization's money. Cults do all they can to keep outside criticism from reaching their members, LW is somewhat open to it at least as far as technicalities and practices rather than core beliefs are concerned. Even the most incompetent cult leader would by now have figured out that such liberties massively undercut their solicitations, so it's unfair to insinuate that they're just after the money. Röstigraben (talk) 12:37, 22 June 2012 (UTC)
The Roko's idea was pointing out a fault with the concept of friendliness and with 'decision theory' that he invented. Hence EY getting very angry. He also did in fact act out the entire taking it very seriously thing, after it gave nightmares to a donor, which I don't think helped to do anything about donor's fears. I do not buy into dichotomy between 'malice' and 'stupidity', especially as applied to utilitarian who rationalizes. What is exactly the difference between incompetence and believing in your own lies anyway? Ohh I see: if it is a past villain whom we vilified to the level of caricature James Bond's antagonist, then it was beliefs in the own lies and all around evil, while otherwise, it's honest self delusion. The way I see it, it is same mundane shit, no caricature arhvillains whom wake up in the morning and think, 'how can I be evil today', everyone's thinking they're doing good, using some entirely broken logic to justify their actions (which is awful easy if you believe that prevention of dust speck in sufficiently many eyes can justify even 50 years of worst torture, and when you seriously believe your actions affects the immense number of people who will ever live). Nobody evil ever puts their effort into making sure they're wrong about what they're doing. They do not care, that's the thing. Only caricature villains properly evaluate their actions, deem them evil, and then do them. Regarding the GiveWell's 'audit', I imagine the choice was very simple: publish it on LW or have it published on GiveWell, with the former being a lower loss. edit: and I don't believe the argument that "he could of done better if he wanted to make a cult" works. Look at this: http://www.timesunion.com/local/article/Secrets-of-NXIVM-2880885.php . Consider that this whole improving art of human rationality thing is very new compared to NXIVM. Consider that Keith Raniere is by all metrics most likely to be much, much smarter than EY, with much better talent at cult making. Consider that Keith had prior experience doing nasty shit. Consider that EY's working online and has to reach people via internet, which is a novel task for a cult. Dmytry (talk) 13:54, 22 June 2012 (UTC)
That's actually another point - how, exactly, would you run a cult or sect when all you have is an online platform plus occasional small-scale meetings? Cults are all about isolating their members from the rest of the world, so nobody can interfere with the brainwashing. Personal face-to-face interactions are very important when you're trying to exert psychological pressure on people, not to mention really controlling their lives. On a platform such as LW, you can always leave without a trace, you can switch usernames, nobody has any clue about whom else you're in contact with, and most importantly, communication happens in one huge forum for everyone to see. That's totally anathema to all the methods of social control that cult leaders like to employ. Seriously, just because there are some slight tendencies towards a cult of personality surrounding Yudkowsky, that does not mean that the community actually resembles a cultist organization. Röstigraben (talk) 14:35, 22 June 2012 (UTC)

Let me find again that post from some girl who was dropping out of school and moving to work on existential risk together with Eliezer, and her father was concerned. LW is a recruitment ground. There is an enormous advantage to online operation: you can reach a very huge number of people to pick from; if you are going for sheer devotion its hard to beat. Also if you look at NXIVM it is clear it takes a lot of intelligence/talent/luck to be this successful; I don't see the argument "he could of done better if he was making a cult" working. Dmytry (talk) 18:06, 22 June 2012 (UTC)

here's example : http://lesswrong.com/lw/1sa/things_you_cant_countersignal/ but I think I had another person or another post in mind. Also on all the 'he could of done better', it's way beyond majority of people to even get 1 person to join like this, I am not buying that he could of done better, especially as he prefers to see what he's doing in positive light via rationalization (which is how bad things typically get done). Dmytry (talk) 18:14, 22 June 2012 (UTC)
This talk page is starting to remind me of that chapter in Them where a bunch of activists organize a protest against that grave and dire threat to our society, David Icke! Nebuchadnezzar (talk) 20:13, 22 June 2012 (UTC)
That's it? After years of insidious brainwashing and alarmism about the dangers of unfriendly AIs, Eliezer's got...one teenager dropping out of high school so she can visit SIAI? Worst. Cult. Ever. For the vast majority of his would-be disciples, their "sheer devotion" apparently exhausts itself in logging on to a website to discuss his pet ideas and contribute to their shared cognitive framework, all of it entirely voluntary. No traces of cult-like structures, power relationships or control. Admitting that outside critics and dissenters may have a point and exposing members to said criticism, instead of demonizing them and denouncing their statements as lies. Giving up your potentially most effective tools for fleecing believers and keeping them in line (as Roko's idea would essentially have worked out to the same threat of eternal torture that proved to be quite effective with Christians). Writing lengthy texts about how people can avoid misperceptions instead of manipulating them to your advantage. Sorry, but nobody but an idiot would have made that many obvious mistakes. The far more reasonable explanation is that he's not trying to build a cult. Yes, he wants to inspire a movement, he clearly enjoys the attention and flattery which he's getting from his community, and he vastly overestimates the kind of contribution SIAI could possibly make to the field, but that's about it. Simple human failings that nobody's immune to, not even if you've spent years analyzing cognitive biases, but nowhere near the kind of intentional deception and lust for power that cult leaders exhibit. Röstigraben (talk) 08:56, 23 June 2012 (UTC)
It's just 1 teenager that I could find a post of on prompt (edit: and as a counter example to your idea that it is all happening on the internet. There's this real world part, which is poorly visible, and into which people join). There's also the folks wasting considerable money on SI promoting pseudoscience. Also, are you sure that creators of other cults actually "wanted to make cult", movie villain style? I think it just sort of grows onto people who at start genuinely think they can improve the world (or if we vilified them, 'believe in their own lies'), while getting a little well justified slice for themselves. Dmytry (talk) 13:04, 23 June 2012 (UTC)

Trope

The point of this one wasn't clear to me:

A rational person must defend a non-mainstream belief.[ref]http://lesswrong.com/lw/1ww/undiscriminating_skepticism/[/ref] Apart from the fact that what is mainstream varies enormously around the world (the mainstream beliefs of Southern Baptists are not mainstream beliefs of the scientific community and vice versa), being contrarian just for the sake of it is not rational.

That ref is already linked in the article ("Any group that resorts to saying its critics are doing skepticism wrong[22] is already fucked and is never going anywhere.") and isn't an example of what you seem to be talking about (it's an example of a different stupidity). Do you have examples that would make your point clearer? - David Gerard (talk) 19:10, 19 June 2012 (UTC)

I don't quite understand what the part you quoted is intended to mean. I provided the link as an example. Yud says that nobody dares to consider cryonics after watching Penn and Teller, and that Shermer "blew it" by mocking one of his holy cows.--Baloney Detection (talk) 19:31, 19 June 2012 (UTC)
I think cryonics may work, but that'd be by pure serendipity or by future science being way good at reconstructing data in the shredded beyond repair brains. The research appears to be crackpottery for the most part, there's no clear feedback or standards. What would be needed is e.g. scanning microscopy study of a brain sample frozen under some conditions, demonstrating that gate concentrations on membranes can be read off it. Proof of concept micro uploading essentially. edit: or preserving neural network so that it retains original electrical function along with training. I do believe there's something we can do right now that would preserve brains well enough for future scanning if any appears, the problem is knowing *what* to do right now, and the cryonics people aren't really working on that, the cryobiology people are; as is always the case with such futurism. Dmytry (talk) 05:38, 20 June 2012 (UTC)
If it's a "trope", it should have a few examples. Please give some so your text can be written to make sense - David Gerard (talk) 06:39, 20 June 2012 (UTC)
I think the undiscriminating scepticism article is good enough example (edit: i.e. it refers to Penn and Teller). One has to somehow demonstrate that the scepticism is discriminating by agreeing with Eliezer on the pet issues or one blows it. Believing that AI could in principle be made doesn't suffice (implied) as demonstration of discriminating scepticism. Dmytry (talk) 07:43, 20 June 2012 (UTC)
A rational person must defend non-mainstream beliefs? Are they defining rationalists or hipsters? ±Knightoftldrsig.pngKnightOfTL;DRmore at 11 23:54, 21 June 2012 (UTC)
@Dymytry: Actually crynics has been tested on animals and failed.--Baloney Detection (talk) 19:49, 27 June 2012 (UTC)

Science vs Bayesianism

Yudkowsky has declared that Bayes' theorem is superior to science. Luke Muehlhauser (unsuprisingly) agrees. Massimo Pigliucci found the whole notion flawed.

So what do you say? Is Bayes' theorem superior to science as a way of knowing? I'm not very well-versed on probability theory, but I do get suspicious when someone proclaims he/she has a better way of knowing than the scientific method, especially if that individual is an avid promoter of pseudoscience.--Baloney Detection (talk) 19:44, 27 June 2012 (UTC)

Personal tools
Namespaces

Variants
Actions
Navigation
Community
Toolbox
support