Talk:LessWrong/Archive17

This is an archive page, last updated 3 May 2016. Please do not make edits to this page.

Archives for this talk page:

<1>, <2>, <3>, <4>, <5>, <6>, <7>, <8>, <9>, <10>, <11>, <12>, <13>, <14>, <15>, <16>, <18>, <19>, <20>

, (new)(back)

Where is their research?

These guys claim to conduct research, even having the word "research" their new name (perhaps the word "singularity" gave bad associations?). But what research have they actually produced? What keeps their donors motivated to keep donating? How do they know if any research is being produced, or if the institute employees just eat cheese doodles for the money all day?--Baloney Detection (talk) 11:56, 20 April 2013 (UTC)

Well, in the fairness, they did fund and Yudkowsky co-authored something on assigning (non unique) probability-like values to propositions, I don't have the link handy, but there's a post on Löb's theorem. Of course they oversell it to the point that the instinctive reaction is "this got to be BS" but it looks like an OK paper, as much as I'd not expect to admit this. Still, by large, what happened is that they hired Luke Muehlhauser who doesn't know about anything technical but can adequately/objectively research what a research organization would look like, and then push towards outwards appearance of such, and consequently outwards image improved a fair bit. It's still batshit crazy though, judging by reactions to Basilisk. Dmytry (talk) 13:29, 20 April 2013 (UTC)

Not that RationalWiki will ever mention anything about it in this article or the one about Yudkowsky, since RationalWiki's mission is seemingly to just mock whatever it chooses not to like. Aris Katsaris (talk) 14:12, 22 April 2013 (UTC)

Using Thiel's money to list yourself as co-author is very weak evidence of competence. That piece about Fermi is strong evidence of extreme crackpottery - the crackpot needs the scientific establishment wrong, so the crackpot confabulates history. Dmytry (talk) 17:19, 22 April 2013 (UTC)

Thankfully I donate money to MIRI, not to Yudkowsky directly, so I don't actually need to give a bloody hoot about whether the research done at MIRI is done personally by Yudkowsky or by the other people MIRI employs. Aris Katsaris (talk) 17:26, 22 April 2013 (UTC)

There's much more effective ways to fund research than a "charity" where the top paid member is a mix of crackpot and scam artist. They literally live in an imaginary world where people of the Enrico Fermi calibre can't make simple conclusions from known facts due to extremeness of the outcome (or the like), so they need those conclusions made for them, or else we all die. The real world doesn't work like this. Dmytry (talk) 05:40, 23 April 2013 (UTC)

There is a pretty big LW post with all their stuff that has gone into actually published papers, I think.--^talk 14:16, 20 April 2013 (UTC)

Where? Also, they (MIRI) seems to have virtually no interaction with the scientists who research the subjects they write about. Raises skeptical hackles.--Baloney Detection (talk) 16:51, 22 April 2013 (UTC)

Deleted comment about LW cultish characteristics

Hello, Yudkowsky just deleted a commment of mine: http://lesswrong.com/lw/h8d/ritual_report_schelling_day/8t9r where I compared the LW community against a list of cult characteristics: http://www.prem-rawat-talk.org/forum/uploads/CultCharacteristics.htm

The comment was relatively well received (it wasn't even downvoted). I also have the html version of the deleted comment. V_V

I'm a little surprised you didn't yet get accused of being me, or vice versa. I think Yudkowsky's mental health is deteriorating. The Bayes vs Science was bad enough, but the recent piece about Enrico Fermi... sadly, Yudkowsky really lives in the imaginary world where it is normal that someone like Fermi can't make a "default conclusion" that fission ===> chain reaction, can't extrapolate known facts, just like now when people of similar calibre do not care about skynet and don't sign up for cryonics. Not in the world where Fermi was unsure that nuke is possible because little was known, but, once enough was known (once they detected secondary neutrons and measured things), he of course concluded it without shying from implications, and, being a fairly reasonable and rational man, even implemented quite clever safety measures for the speculative chance of some positive feedback. Dmytry (talk) 06:59, 22 April 2013 (UTC)

"it wasn't even downvoted". Apologies for not downvoting you. It seems to have given you a wrong impression about its perceived quality, which I personally found emetic in its slanderous quality (Yudkowsky "demanded" people's disposable incomes? Really??). Aris Katsaris (talk) 14:06, 22 April 2013 (UTC)

"I would be asking for more people to make as much money as possible if they’re the sorts of people who can make a lot of money and can donate a substantial amount fraction, never mind all the minimal living expenses, to the Singularity Institute." (from Kruel's list of quotes). And there was a considerably worse one as well. I think I vaguely recall you defend one of such instances, so screw off, you know Yudkowsky demands money from the followers. Dmytry (talk) 17:13, 22 April 2013 (UTC)

In short you want to use the word "demand" without any actual interest to its actual meaning, in short yet another bloody lie by the usual bloody liars, much like the lies you've spoken against me in the past. Do you know about the difference between "I demand" and "I would ask" you idiot? Aris Katsaris (talk)

Ahh, right, if he didn't literally say "I demand" then it's a bloody lie to describe as demand, no matter how much he'd go on about how we all die if you don't donate the money. I see. Dmytry (talk) 17:29, 22 April 2013 (UTC)

Yes, once again you see, if the truth was really bad you wouldn't be constantly needing to misrepresent it with such fierce dedication. "Demand" at the very least must mean a penalty if you don't oblige the person so 'demanding'. I don't know of any lesswronger being penalized at *all* for not donating. Aris Katsaris (talk) 17:36, 22 April 2013 (UTC)

http://www.merriam-webster.com/dictionary/demand . You know, it is a common crazy phenomenon to see everyone else as a liar due to their use of words in their linguistically appropriate meanings. Dmytry (talk) 17:42, 22 April 2013 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Seems like a persuasive request for money, which while perhaps not a "demand," in the technical sense of the word, given the bullshit being spouted by the uneducated, non-contributing member of society about how he's the last hope for us not to get fucking SKYNETTED OMGERHERD, is somewhat reprehensible, unless, of course, he lives like a hermit. Let's go to the tape.... In 2011 the institute paid 32 year old high-school graduate $99k, so let's go with a big fat NOOOOOOPE. Hipocrite (talk) 18:12, 22 April 2013 (UTC)

Yudkowsky and LW in general don't have the same "any comment can stay up no matter how shit" ethos as RW does (different stroke and all). So, not only is this factoid entirely unsurprising but also depressingly unworthy of any note whatsoever. You basically trolled a site that applies the DFTT principle with tremendous success, and then came here to bitch about it not going down well? ^sshole 19:23, 22 April 2013 (UTC)

They're getting heavy with deletions, though. E.g. this guy http://lesswrong.com/user/PrawnOfFate/overview/ , looks like a rather polite-disagreeing fellow despite his non-LW-ian beliefs on topic of moral realism/absolutism, but still labelled a troll, his comments deleted, etc etc. Dmytry (talk) 08:07, 23 April 2013 (UTC)

Also, he just got mass downvoted (with 50+ "karma" loss), likely using a bot. Dmytry (talk) 09:40, 23 April 2013 (UTC)

Well, this website was founded as a place to obsessively monitor another website and bitch and brag about vandalism there. :D

But yeah, this and the recent psychic incident (complete with a talk-moving that was either extremely incompetent, or CPesque "burning the evidence") does indicate that RW may be developing an issue with one trick ponies...--ZooGuard (talk) 19:30, 22 April 2013 (UTC)

A case of inventing new terms for already well-known concepts

Yvain wrote a piece for LW called The noncentral fallacy - the worst argument in the world?, as usually highly upvoted as his writings tend to be. Yet what he came with was not something new, just dressing up poisoning the well (a well-known, since-long identified logical fallacy) with new jargon. Of course, for people who are not familiar with logical fallacies Yvain and the rest of LW look like geniuses, as if Yvain came up with a revolutionary idea nobody ever concieved before. Which he didn't, even if his post looks like he is trying to pass it off as he did. That SpaceBattles thread about LW is an amusing resource!--Baloney Detection (talk) 19:42, 28 April 2013 (UTC)

No, the two are clearly different. To refer to examples relevant to this talk page, "Poisoning the well" is e.g. speaking of Roko's basilisk whenever someone discusses LessWrong. "non-central fallacy" on the other hand would be e.g. describing Eliezer Yudkowsky as a "school dropout". Aris Katsaris (talk) 12:51, 30 April 2013 (UTC)

Your examples do nothing at all to clear things up. Both seem like poisoning the well to me. It might be that the so-called noncentral fallacy is a subtype of poisoning the well. But frankly I find naming fallacies to be a waste of time to begin with. TallMan (talk) 13:13, 30 April 2013 (UTC)]

Many informal fallacies look like or flow into each other -- I could also say they're both the "guilt by association" fallacy, the "appeal to emotion" fallacy, even the "fallacy of composition". (perhaps in the end all informal fallacies are just about assigning wrong conditional probabilities) Still the "non-central fallacy" as described by Yvain has a very specific definition of using the membership to a category which X truly belongs to but X is not a typical example thereof. e.g. "Aaron Swartz was a criminal" or "MLK was a criminal". That's obviously a much more specific fallacy than all the many ways that one can "poison the well" in a debate. Aris Katsaris (talk) 14:03, 30 April 2013 (UTC)

More specific, perhaps. But identifying a subtype of an old fallacy isn't particularly interesting, or useful - and it's not particularly original, either.TallMan (talk) 15:00, 30 April 2013 (UTC)

Is it a fallacy to describe you as mammal, vertebrate, or animal? Yudkowsky is an HS drop out, which is valuable information if you are to read him because otherwise you might just trust him when he goes on about what e.g. Enrico Fermi knew at the time of a quote - you can expect an HS drop out crackpot to be wrong about such kind of things, whereas you can expect crackpot with a PhD to be correct about such kind of things, and when reading the former types of crackpot, scepticism must be applied even to the easily verifiable assertions. edit: same goes for crap like that MetaMed's promotion I linked below - how do you know that you can't trust anything there? When someone compares CT scans to Hiroshima survivor, that relies on trust, because the radiation is highly confusing and it is hard for most readers to check. Dmytry (talk) 05:26, 1 May 2013 (UTC)

For extra points, count how many times Dmytry attempts to poison the well above. :-) As a sidenote, I believe that Yudkowsky dropped out of school at age 12, so he most probably wasn't at highschool yet, and thus "highschool dropout" is a factual inaccuracy. Aris Katsaris (talk) 10:07, 1 May 2013 (UTC)

What ever. If Andrea Rossi wants to get a job at a secure nuclear facility, and fails background check, is that poisoning the well too maybe? Dmytry (talk) 10:41, 1 May 2013 (UTC)

I would have said "using Bayesian evidence" - David Gerard (talk) 18:48, 1 May 2013 (UTC)

MetaMed, "evidence based" healthcare.

Endorsed by Yudkowsky, run by former director of singularity institute Michael Vassar and employing Yvain, de-constructed here. It's lovely these guys didn't even skip the trope of comparing CAT scans to Hiroshima survivors and being a few orders of magnitude off about it. Dmytry (talk) 05:21, 29 April 2013 (UTC)

Heh, they say you get jobs through connections these days. Funded by Thiel. Hmm, is Thiel (a seriously creepy nutjob) wasn't around, then I guess MIRI, and by extension, LW, would die out rather quickly.--Baloney Detection (talk) 19:53, 3 May 2013 (UTC)

The future of LessWrong?

As it currently stands, LessWrong is a fringe site whose impact on the world (let alone academia) is close to zero. Despite David Gerard boasting about how RW is irrelevant compared to LW and only read by its editors, [1] actually has far more visitors than [2].

So what do you think is the future of LW? Will they (not all of them, obviously) eventually grow out of their beliefs? Perhaps Muehlhauser will find another (credible or ridiculous) cause to devote his life to? Or will LW grow in influence and signficance? Right now LW is really under the radar, they spout their nonsense without hardly anyone noticing? If they grow, then perhaps Skeptic magazine and the like will start to deal with their claims like they do with homeopathy?

Or perhaps it will continue like now, a fringe, irrelevant site and they don't grow out of it? No signficant changes.--Baloney Detection (talk) 14:48, 14 June 2013 (UTC)

It's not just alexa, but also google trends: [3]. However lesswrong has much more "worthwhile" content than RW, and probably will have a steadier future than rationalwiki. RW got a kickstart from conservapedia: [4] (google trends again). Just compare http://lesswrong.com/comments/ to RW Special:RecentChanges, or http://lesswrong.com/recentposts/ to Special:NewPages. --81.175.227.88 (talk) 10:58, 16 June 2013 (UTC)

Do you believe Yudkowsky's claim that Bayes' theorem is superior to the scientific method? Or that cryonics works? Or that the singularity is near? LW is pretentious garbage.--Baloney Detection (talk) 19:08, 16 June 2013 (UTC)

What reason do you have to claim that Bayes' theorem is NOT strictly superior to the scientific method? The way I see it, the scientific method is dependent on an informal quasi-understanding of what Bayes' theorem actually formalized (how evidence and hypotheses connect to each other), so yeah, Bayes' theorem could be called strictly superior to the scientific method, since it's formalized and more powerful: it can achieve all that the scientific method (namely the formulation and rejection or acceptance of alternate hypothesis based on testing) does and more. But it's far harder to apply of course, same way that calculus is harder than counting on your fingers. A human mind can't keep and constantly update a million different hypotheses, nor keep all the relevant evidence in mind, so when we simplify simplify simplify down to the underdeveloped human mind, all that remains is the scientific method. But a computer wouldn't ever need to fully "reject" or "accept" a hypothesis, it would just need to assign and constantly adjust the respective certainties, according to, guess what, Bayes' Theorem. Aris Katsaris (talk) 20:16, 16 June 2013 (UTC)

You see, here is the problem of taking Yudkowsky as your gospel. Yudkowsky has committed a category error. Take a read at how impressed (which is not very) a philosopher of science is with Yudkowsky's idea: http://rationallyspeaking.blogspot.com/2010/09/eliezer-yudkowsky-on-bayes-and-science.html --Baloney Detection (talk) 20:41, 16 June 2013 (UTC)

If that's the severest criticism of the argument that you can find, then it's praising with faint damnation. The comment section beneath that article has a good enough discussion, but the article itself is mostly complaints about the writing style of Yudkowsky rather than about the idea itself -- Massimo Pigliucci focuses on Yudkowsky's application of Bayes on judging over quantum mechanics interpretation and YET PIGLIUCCI ADMITS HE DOESN'T UNDERSTAND QUANTUM MECHANICS. So how is he qualified to judge Yudkowsky's understanding thereof? But let's forget that for a sec, and use a simpler example everyone can get: The most recent well-known example of the triumph of Bayes's Theorem is Nate Silver's election predictions. Which scientist made the same predictions using the scientific method of formulating and testing hypotheses instead? Btw, I saw that the same blog you linked to call Bayes' Theorem the ["Great Equation of Power and the Timeless Secret of the Universe" http://rationallyspeaking.blogspot.gr/2012/11/odds-again-bayes-made-usable.html], so if you want to escape people who treat Bayes as the coolest thing ever, it may not be your best choice. Aris Katsaris (talk) 21:42, 16 June 2013 (UTC)

The idea is still a category mistake. Bayes' theorem is (quite obviously) a formal statistical theorem. The scientific method is an abstract philosophical concept. And if you read, say, your Kuhn or Lakatos or Feyerabend, there is no such thing as a unified scientific method. It is simply a reification. Furthermore, Bayes theorem has been incorporated into scientific methodology. It makes no sense to say that one is better than the other. It's like saying elephant is superior to purple. Nebuchadnezzar (talk) 22:02, 16 June 2013 (UTC)

If someone argues that a comparison between Bayesianism and science doesn't make sense, then the proper followup question is to see in what way it was meant by the person making such a comparison (rather than assume the person is stupid). I can *say* "elephant is superior to purple", if I'm discussing whether I should buy pyjamas-with-elephants or purple-pyjamas for a baby nephew that likes the color purple but completely adores elephants. And we can e.g. similarly discuss whether Bayes theorem or Science is superior to discovering epistemic truth if we are given the task to program a computer with an understanding of the former or with an understanding of the latter. In that case, for example, the very fact that Bayes is much more formal than Science makes it by itself superior on that one specific criterion. Another point is that Bayesianism has a way to distinguish between theories that make identical predictions in a way that Science does not (between theories equally supported by the evidence, the theory with the highest prior will end up the theory with the highest posterior). Aris Katsaris (talk) 22:42, 16 June 2013 (UTC)

"And we can e.g. similarly discuss whether Bayes theorem or Science is superior to discovering epistemic truth if we are given the task to program a computer with an understanding of the former or with an understanding of the latter." I don't see how this answers any of my objections. Comparing a well-defined statistical theorem to a philosophical abstraction is still nonsensical. What does it even mean to program a computer with an understanding of science? "Science" cannot be summed up in a single formula like Bayes' theorem.

"Another point is that Bayesianism has a way to distinguish between theories that make identical predictions in a way that Science does not" Which is why scientific methodology has been incorporating Bayesianism for some time now. This statement only makes sense if you arbitrarily define science to exclude the use of Bayes' theorem. Nebuchadnezzar (talk) 03:27, 17 June 2013 (UTC)

You miss the point. RS is hardly anti-Bayes, it (or at least Pigliucci) is simply disagreeing with Yudkowsky's idea. You'll have to get used to that whenever you go outside of LW. As for QM, Pigliucci was referring to Feynman's quote, not expertise in the field (which Yudkowsky doesn't have either, even though his fanclub thinks so).--Baloney Detection (talk) 22:19, 16 June 2013 (UTC)