All of Cakoluchiam's Comments + Replies

A free full-cast Audiobook of Planecrash is currently in production at https://shows.acast.com/project-lawful-aka-planecrash, using AI-generated voices. It is quite excellent, albeit missing a few of the glowfic-specific elements such as character portraits and reaction tags (posts with no text, only character portraits). I highly recommend it for anyone.

There is a parallel analysis podcast and book club hosted by myself and members of The Bayesian Conspiracy's discord channel, formerly It Makes Sense If You Understand Decision Theory, now We Want Headband... (read more)

I dunno about yours, but my lampshades don't usually spin, particularly not with a "vroop".

0Kindly
Right, EY just threw that in to make it harder to guess.

On second thought, I have an alternative solution to what it is lampshading, that is the broken suspension of disbelief that after stating that the terms of Quirrell's contract prevented him and others from investigating Quirrell's identity, Albus would leave the room, allowing his conspirators to investigate Quirrell's identity.

This theory sounds incredibly plausible from the perspective of the author wanting to use the lampshade trope, but from the perspective of the reader, that action was completely in-character for Dumbledore and doesn't actually break suspension of disbelief.

You can't say it's obvious unless you can point to something it is specifically lampshading. The best answers I've seen so far in this thread are that it's lampshading itself, in which case there's no reason for it to have been in the story. Traditionally when you hang a lampshade on something, it's something that the author needs as a plot device but actually wouldn't make very much sense if the story were playing out realistically, that is, it threatens suspension of disbelief. I don't think any of us would disbelieve that Dumbledore would have a strange... (read more)

2ygert
I am sorry for being confusing about what the "obviously" was supposed to imply. I meant from the physical description, if you visualize the object in your head, the object is pretty obviously a lampshade. From there is is a pretty reasonable guess to say that this lampshade represents an attempt by EY to hang a lampshade on something. Of course, it is not so obvious what is being lampshaded.

The vrooping thing sounds like a centrifuge to me, though the pulsing light isn't something I'm familiar with in such apparatus.

If it is indeed a centrifuge, it would make sense that it was only mentioned -after- Dumbledore left the room. If they had somehow obtained a sample of Quirrell's blood, they might be separating it to do a DNA comparison against any candidates for his identity, which if I were HJPEV would have been one of my first (dozen) solutions to the problem of identification.

0ygert
The vrooping thing is very obviously (as has been previously pointed out) a lampshade.

It's a reference to an episode of The Simpsons, wherein Lisa's boyfriend states: "I'm a level 5 vegan. I won't eat anything that casts a shadow."

Edit2: Primary source found.

Fair point, though that also removes the point of evidence that casting requirements are removed with practice.

"It takes a cracked soul to cast." and "Murder tears the soul." just says that if you've gotten to the point where you could cast it once, that particular pre-requisite is already accomplished, so the work to crack your soul is already put in. It doesn't say anything about removing the requirement of wanting someone dead.

Though, so long as we're looking at evidence, if we take Quirrell at his word, then his ability to cast the spell despite not wanting his opponent dead is pretty strong evidence that the requirement is in fact removed. ... (read more)

1Desrtopa
Just because he didn't intend to kill him doesn't mean he didn't want him dead. As Moody said, you have to want it, not just for the greater good, but as an end unto itself. Quirrell might have wanted to kill Bahry as an end unto itself, whereas for matters of convenience it was better to leave him alive.
5Waffle_Iron
Harry didn't cast the patronus then, it was already active, he just moved it.

In canon, Moody used the unforgivables on a spider, and given the prevalence of ostensibly non-sentient things-to-fear in the magical world (e.g. boggarts), it's conceivable that they could have found a particular magical creature that even the most PETA-supporting student would have no trouble excising from the world. Also, as far as I can tell, there's nothing in canon to contradict that curses' targets are limited to Kingdom Animalia (see also: Harry's existential crisis about sentient plants), and I seriously doubt there are any 7th level vegans at Hogwarts.

2DanArmak
How do you level up in being vegan? And what powers does it grant?

I interpreted the ease of casting the spell as a specific application of scope insensitivity rather than a change in the requirement to cast it. That is, while casting it the second time might be just as difficult (i.e. take as much mental/magical/spiritual energy) as the first, the third and fourth time would together be only as strenuous as the first, as would the collective fifth through eighth time, etc. It is already established in-universe that some form of personal mana depletion exists, and my idea of this difficulty reduction is an extension of that form of energy to the spiritual energy (established in canon w.r.t. horcruxes, dementors, etc.).

I'm a little surprised that HJPEV didn't immediately update his probabilities regarding Quirrell's motives in Azkaban with the new knowledge from Moody that "You've got to mean it. You've got to want someone dead, and not for the greater good, either.", which would seem to discredit the Defense Professor's excuse that "a curse which cannot be blocked and must be dodged is an indispensable tactic."

8Nominull
I took this passage as saying that you don't have to be especially pathological to cast the killing curse a second time - Moody explicitly says it "doesn't tell us much". So if we trust him, it doesn't tell us much.
5buybuydandavis
Yeah, that kind of leaped out. It also made me wonder how Quirrell and Dumbledore thought they were going to teach students the spell.
3Exetera
Remember that Harry had also learned that Quirrell had successfully used Avada Kedavra on two Death Eaters. Moody says that it isn't hard to cast AK for a second time, and Harry already knows that this time would have been at least Quirrell's third.
Desrtopa310

Not necessarily; someone who's as deeply misanthropic as Quirrell might wish most people dead (having killed before, he can, as per Moody's explanation, wish people dead rather more casually than non-murderers.) If you're already capable of bringing intent-to-kill to bear on pretty much anyone who crosses you, you can probably use it strategically the way Quirrell suggests.

On the other hand, even if Quirrell's explanation holds true, it does suggest Harry should revise upwards his estimates of just how cavalier Quirrell is with other people's lives.

Either way, should we or shouldn't we have trusted the rest of their answers to be statistically reliable?

1EricHerboso
I see no reason to throw out their responses. They appear to just not be familiar with the terminology. To someone that does not know that "fair coin" is defined as having .5 probability for each side, they might envision it as a real physical coin that doesn't have two heads.

What, the myriad prophets of revealed religions and cults aren't enough of a hint for you?

If this were anywhere but a site dedicated to rationality, I would expect trolls to self-report their karma scores much higher on a survey than they actually are, but that data is pretty staggering. I accept the rejection of the hypothesis, and withdraw my opinion insofar as it applies to this site.

Hell, if the mathematical universe hypothesis is correct, then somewhere out there in the universe there is, with no intelligent priors, a collection of particles in the form of a computer, simulating a universe containing intelligent entities.

0Decius
And the universe it is simulating itself contains an entity which created a computer which simulates a universe... Given that the mathematical universe hypothesis is correct, what are the odds that the universe we experience A- is the mathematical universe B- is a computer analogue simulation which was generated spontaneously C- is a computer analogue simulation which was created by an intelligence but is currently unattended D- is a computer analogue simulation which is attended by someone who wishes to suppress the thought that the universe is atten---

Would someone who created a computer that created the universe count as a god? I can easily write computer games with more complex behavior than I feel capable of fully comprehending, but I would not consider that computer program an intelligent entity. I can imagine that someone more educated and with a higher mental capacity than I could similarly write a computer program that is capable of creating and maintaining in simulation a universe with the global constants and initial conditions necessary to produce intelligent life without the program actually ... (read more)

1DaFranker
This seems to unreasonably deviate from the way almost every type of simulation I've ever heard of works. You can pause/resume, you can increase/decrease the "timesteps" to make the world go "faster" (with larger quanta levels though), or you could just arbitrarily increase the raw processing speed of the machine running the simulation to make the ratio of simulated vs external time proportionally higher. Of course, if what you're proposing instead is that our "minds" are actually outside the simulation and sending input into it, rather than being fully contained within the simulation, then yes, the real-time constraint does apply. ETA: In the latter case, I would argue that the term "Virtual Reality" is more appropriate and the use of "simulation" here is misleading and prone to conflating or confusing the two scenarios.
3CCC
Given that all the 'players' are running in the universe in question, being able to make a large number of fine-grained adjustments to local variables in an instant (in-universe time) is simple; simply pause the simulation. ...unless some of you out there are actually players from outside the universe, in which case the rest of us would appreciate a hint.
1Cakoluchiam
Hell, if the mathematical universe hypothesis is correct, then somewhere out there in the universe there is, with no intelligent priors, a collection of particles in the form of a computer, simulating a universe containing intelligent entities.

TROLL TOLL POLICY: Disapprove: 194, 16.4% Approve: 178, 15%

So more people are against than for. Not exactly a mandate for its use.

Hypothesis: those directly affected by the troll policy (trolls) are more likely to have strong disapproval than those unaffected by the troll policy are to have strong approval.

In my opinion, a strong moderation policy should require a plurality vote in the negative (over approval and abstention) to fail a motion to increase security, rather than a direct comparison to the approval. (withdrawn as it applies to LW, whose trolls are apparently less trolly than other sites I'm used to)

gwern240

Hypothesis: those directly affected by the troll policy (trolls) are more likely to have strong disapproval than those unaffected by the troll policy are to have strong approval.

Hypothesis rejected when we operationalize 'trolls' as 'low karma':

R> lwtroll <- lw[!is.na(lw$KarmaScore),]
R> lwtroll <- lwtroll[lwtroll$TrollToll=="Agree with toll" | lwtroll$TrollToll=="Disagree with toll",]
R> # disagree=3, agree=2; so:
R> # if positive correlation, higher karma associates with disagreement
R> # if negative correlati
... (read more)

I don't agree that the first doesn't count. The Relationship Style question was about preferred style, not current active situation. It could be that 2/3 of the polyamorous people just can't get a date (lord knows I've been there). (ETA:) Or, in the case of not looking, don't want a date right now (somewhere I've also been).

2DaFranker
I'm in the "no preference" camp, not the poly specifically, but I'm certainly there. LessWrong does seem to indirectly filter for people who are there, simply because people who aren't are less likely to take an interest in things that would lead them to LW, IME.

It was stated that they should give the obvious answer and that surveys that didn't follow the rules would be thrown out... but maybe 50% isn't as obvious as 99.99% of the population thinks it is.

Is there any reason the prompt for the question shouldn't have explicitly stated "(The obvious answer is the correctly formatted value equivalent to p=0.5 or 50%)"?

0Eugine_Nier
My working theory is that they were trolling.

Any results for the calibration IQ?

3gwern
The original question: Well, the predictions spread the usual range and look OK to me: R> lwci <- as.numeric(as.character(lw$CalibrationIQ)) R> lwci <- lwci[!is.na(lwci)] R> # convert tiny decimals to percentages & put a ceiling of 100 (thanks to Mr. 1700...) R> lwci <- sapply(lwci, function(x) if (x<=1.00) { x*100 } else { if(x>100) { 100 } else { x }}) R> summary(lwci) Min. 1st Qu. Median Mean 3rd Qu. Max. 0.0 20.0 50.0 44.8 70.0 100.0

This question was biased against people who don't believe in history.

For my answer, which was wildly wrong, I guesstimated by interpolating backward using the rate of technological and cultural advance in various cultures throughout my lifetime, the dependency of such advances on Bayesian and related logics, with an adjustment for known wars and persecution of scientists and an assumption that Bayes lived in the western world. I should have realized that my confidence on estimates of each of these (except the last) was not very good and that I really shoul... (read more)

0Manfred
Have you read Malcolm Gladwell - Blink? It's a fun book that doesn't take too long, which hella makes up for the occasional failure of rigor. Anyhow, the conclusion is that even on hard problems, expert-trusted models will still have very few parameters. And those parameters don't have to be the same things you'd use if you were a perfect reasoner - what's important is that you can use it as an indicator.

I want to say that my own origin lies in having been raised Unitarian Universalist with the most amazing minister who never invoked "God" as anything more than the common good or interpersonal kindness. I want to believe that UU Sunday school attendance, or, more interesting to me even at that young age, ditching class and sticking through the "adult" section of the worship, where she would give the most awe-inspiringly inspirational sermons, would be enough to awaken any child as a rationalist. Alas, I am fairly certain I was prepared ... (read more)

Two relatively simple rebuttals to your premise:

(1) One can easily create a number with a non-terminating decimal expansion which makes use of a finite quantity of the digit "5". Therefore it is conceivable that one could also exist in an infinite universe which makes use of only a finite quantity of atomic structures identical to "you".

Similarly, and working in the opposite direction (complex-to-simple as opposed to the former simple-to-complex extrapolation), it is strongly believed that we exist in a universe with fixed universal con... (read more)

A potential error for the second conclusion is that we have incorrectly predicted the nature of consciousness, and the true solution is that one is somehow able to perceive without a physical avatar functioning in the way we expect of a human capable of perception. Thus, "you" are able to experience the branches of the MWI where everyone else perceives you to be dead.

Interesting you don't consider what I thought would be the obvious interpretation of counting Y(+Z) alone, even after you considered adding adopted and foster children, which would, over the population, double-count any choice with the inclusion of X.

If I, given a universal interface to a class of sentient beings, but without access to that being's language or internal mind-state, could create an environment for each possible truth value of the statement, where any experiment conducted by a being of that class upon the environment would reflect the environment's programmed truth value of the statement, and that being could form a confidence of belief regarding the statement which would be roughly uniform among beings of that class and generally leaning in the direction of the programmed truth value, th... (read more)

Nu, lrf, V sbetbg nobhg jrrxraqf. Ubjrire, zbfg bs gur ybj-vapbzr crbcyr V xabj ner npghnyyl jbexvat zhygvcyr cneg-gvzr wbof, be n cneg-gvzr wbo naq fpubby (juvpu, va vairfgzrag inyhr, qrcraqvat ba gur ntr bs gur vaqvivqhny, vf ubhe-sbe-ubhe pbzcnenoyr gb n ybj-jntr wbo), bsgra nzbhagvat gb zber guna sbegl ubhef n jrrx. V guvax (ntnva, nffhzvat gur urnqnpurf qba'g nssrpg fyrrc) n svsgl creprag punapr vf fgvyy n ernfbanoyr nffhzcgvba, rfcrpvnyyl fvapr fgerff, fhpu nf jbex, vf n cevznel pngnylfg sbe urnqnpurf.

Thanks for the idea. I know he referenced OvercomingBias for a while before LessWrong, and I read a handful of articles there, but I think my sparse interest in following anything, aside from what was crossposted or linked at Yvain's blog, makes any claim of seniority a potential false representation of my expertise.

This is sort of the inverse problem of the way I counted the start of my last relationship more than a year earlier than my then-partner would. One good thing I can say for weddings (aside from marriage) is that at least they solidify a date to count as anniversary. So, I suppose now that I've finally registered an account, I can count today for future anniversaries between myself and LessWrong.

V rfgvzngrq ol nffhzvat gur gerr jnf nobhg gjragl srrg va qvnzrgre (fbzrguvat V erzrzore sebz zl Ylprhz qnlf jvgu pvepn avargl creprag pbasvqrapr), naq unq nobhg na rvtugl-avar qrterr vapyvar (bar qrterr gncre), juvpu V pbhyq rfgvzngr ol bofreivat n inevrgl bs rireterraf va zl ivpvavgl. Sebz gurer, vg jnf fvzcyr gevtbabzrgel.

I spent a lot of time analyzing that question and came up with the following solution, which, granted, assumes at least three things, and "only a fool would attempt a plot that was as complicated as possible", but...

Vs jr nffhzr gung svsgl creprag bs gur urnqnpurf qverpgyl nssrpg jbexvat ubhef, gura gur pbfg bs nal bs gur guerr qehtf vf fvtavsvpnagyl ybjre guna gur bccbeghavgl pbfg bs ybfvat gubfr jbex-ubhef ng zvavzhz jntr. Qeht N, juvyr abg gur zbfg pbfg-rssrpgvir jura pbzcnerq qverpgyl gb Qeht O (be rira P), unf gur terngrfg rssrpg naq fgvyy p... (read more)

0A1987dM
Gur dhrfgvba fnlf “ure” (abg gung V gbbx gung vagb nppbhag jura nafjrevat).
3Desrtopa
Svsgl creprag punapr bs urnqnpurf qverpgyl nssrpgvat jbexvat ubhef frrzf vzcebonoyr tbvat ba gur fbeg bs nffhzcgvbaf V gnxr sebz gur cerzvfrf bs gur dhrfgvba. Vs jr nffhzr n sbegl ubhe jbex jrrx, gura nffhzvat gur urnqnpurf qba'g vagreehcg fyrrc, gung tvirf pybfre gb n bar va guerr punapr. Gernqvat vagb ernyzf bs "cebonoyl bireguvaxvat guvatf," V fhccbfrq gung vs gur crefba jrer na hafxvyyrq, ybj jntr, shyy gvzr jbexre, naq gur urnqnpurf jrer frevbhfyl phggvat vagb gur gvzr gurl pbhyq fcraq jbexvat, naq gur zrqvpngvba abg pbirerq ol vafhenapr, gurve rzcyblre jbhyq cebonoyl whfg sver gurz. Bhe pheerag wbo znexrg urnivyl snibef rzcyblref, naq rzcyblref bs hafxvyyrq jbexref pna nssbeq gb or rkgerzryl cvpxl. Fb tvira gur nffhzcgvba gung gurl fgvyy unq n wbo ng nyy, V gubhtug vg jnf yvxryl gung vg jnf n cneg gvzr bar jurer gurve urnqnpur gvzr qvqa'g bireync zhpu jvgu jbexvat gvzr, naq/be gurl znantrq gb cbjre guebhtu gur urnqnpurf naq jbex naljnl gb nibvq trggvat sverq, fb vg jnfa'g ceriragvat gurz sebz jbexvat.

That Myers-Briggs test was a lot less thorough than what I remember from a lot of the ones I took online back in TheSpark era. Though, part of me is kind of glad that each of the extra credit questions could be completed in under an hour.

I probably never would have heard of the idea if someone hadn't pointed out its conspicuous omission on the census. I read completely through the original test census and it didn't even register as something so noteworthy on first pass... just another thing that I would probably understand better if I actually read more LessWrong, but since I hadn't, I'd leave my answer blank. Now I know a lot more about it and could probably (p=70%) actually put an answer down with some confidence.

Since it appears the final version of the census has been backedited onto t... (read more)

As I mention elsewhere, ... gur grfg vf vaperqvoyl rnfl gb tnzr. Vg qbrfa'g punatr gur cbfvgvba bs cbgragvny nafjref orgjrra gnxvatf, fb trggvat n cresrpg fpber ba lbhe frpbaq gel vf cerggl rnfl.

3Nornagest
That's what I thought. So if jeremysalwen is right and they're calculating scores as if their sample of results is representative, then their scoring algorithm is going to be utterly useless (perhaps outside a narrow regime around 100): the right-hand side of the curve is going to be populated entirely by people gaming the test in various ways (retakes are almost certainly common, and more sophisticated methods are likely to represent at least a couple percent of results), and I'd imagine a good chunk of the left-hand side would fall to guessers.

Unless there are significant numbers of people, myself for example, who take the test multiple times with varied random algorithms just to see how it affects the outcome. I'd only put a (p=0.55) at the test underestimating your score, conditional that it doesn't correct for self-selection bias.

Though, given that the lowest score appears to be "less than 79", rather than an exact number, they may simply drop any scores under 79 from their pool, or at the very least weight them differently. Has anybody identified a similar maximum score which would support this hypothesis of discarding outliers?

1jeremysalwen
Analysis of the survey results seems to indicate that I was correct: http://lesswrong.com/lw/fp5/2012_survey_results/

Also, just spent an hour I should have spent sleeping upvoting all the comments that explicitly mentioned taking the test, and a few that were just insightful.

(yelling) Curse you squid-god of time, for reawakening the sleeping demon that is my love for census, long forgotten in the archives of naturalization! (/yelling)

I too had to estimate my time in the community (even though this is my first day posting anything)—I started lurking shortly after Yvain mentioned it in his blog, but he is a prolific enough writer that digging through those archives would be a mind-numbing task. Perhaps there's someone else who's done the dirty work for me? (hoping)

1Vaniver
Last time, I figured it out by writing a script that went through my comments until it hit the earliest one, but that only worked because I made an account as soon as I got here. (I probably could have gone hunting for the results of that script, but I didn't care as much this time around.) I imagine Yvain has mentioned LW on his blog enough times that it'd be difficult to do an automated solution (but you might be able to just make a list of links to LW from posts, then choose from those).

At first I wondered whether the test adjusts either direction for lack-of-patience; and then I realized I could run an experiment.

There don't appear to be any points granted for finishing early; I just took the test three times, guessing randomly as fast as possible, and scored 93 first then <79 (what appears to be the lowest score possible) twice and 93 the third time, and then took the test a fourth time, guessing randomly at a rate of 1 question/minute (finishing with 1 minute to spare), and got 83. This appears to reject the hypothesis that finishin... (read more)

I strongly suspect that a lot of the members of LessWrong have had a non-internet IQ test and will have entered their scores on the census. Those who also took the extra credit internet test and entered their scores to that as well could serve as a sample group for us to make just such an analysis.

Granted, we are likely a biased sample of the population (I suspect a median of somewhere around 125 for both tests), but data is data.

0A1987dM
I assumed that iqtest.dk would count as “respectable” rather than as “amateur” so I included its score in both questions. If you do such a calibration, you should ignore my entry.
4jeremysalwen
From what I could read on the iqtest page, it seemed that they didn't do any correction for self-selection bias, but rather calculated scores as if they had a representative sample. Based on this I would guess that the internet IQ test will underestimate your score (p=0.7)
2Nornagest
A lot of people in this thread have reported institutionally tested IQ scores much higher than those given by the extra-credit Internet test. (My own score on the former is two and a half standard deviations higher than on the latter, though I took the former many years ago.) I suspect that iqtest.dk's either normed very low (unusually for an Internet IQ test) or is suffering from other problems. The first possibility that comes to mind is that some people are taking it several times (one person elsewhere in the thread reported scores in the 100 range, then 120, then 140 over three trials), and that its scoring system is taking that into account improperly.

Survey taken: check! Account finally registered: check, please!

I was off by 50%ish on the two estimation questions, but I forgive myself Bayes' age since I really know nothing about history in "space-of-time" context. The redwood tree on the other hand was a geometry problem for me, more than anything else, and I misjudged its incline by half a degree.

0A1987dM
What?
5Cakoluchiam
Also, just spent an hour I should have spent sleeping upvoting all the comments that explicitly mentioned taking the test, and a few that were just insightful. (yelling) Curse you squid-god of time, for reawakening the sleeping demon that is my love for census, long forgotten in the archives of naturalization! (/yelling)