= 27d567bc2d48d9f40e9ccc20ea370b15)
Daniel_Starr comments on Harry Potter and the Methods of Rationality discussion thread, part 16, chapter 85 - Less Wrong Discussion
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
Comments (1106)
HPMOR is making me rethink human nature -- because of how people react to it. This is a story full of cunning disguises, and readers seem reluctant to see past those disguises. RL rkcerffrq chmmyrzrag ng ubj many readers took forever to decide Quirrell = Voldemort; I think I now know why.
I suggest that humans are instinctive "observation consequentialists." That is, we think someone is competent and good if the observed results of their actions are benign. We weigh what we observe much more strongly than what we merely deduce. If we personally see their actions work out well, we'll put aside a great deal of indirect evidence for their incompetence or vileness.
In HPMOR, Quirrell's directly observed actions are mostly associated with Harry getting to be more of what he thinks he wants. Even rescuing Bellatrix amounts to Harry getting to save a broken lovelorn creature in terms of what we directly observe. To believe Quirrell evil we have to bring in all kinds of expected consequences to weigh against those immediate positive observations.
Does the resistance to saying Quirrell=Voldemort maybe reflect a broader unwillingness to overlook what we directly witness in favor of abstract deduction? If it does, this implies some interesting predictions about human behavior:
if you can be kind and moderate in your personal behavior, you can get away with incredible amounts of institutionally-mediated violence and extremism, especially to anyone who feels like they "know" you. Hypothesis: the most dangerous people are those who can give us the illusion of "knowing" them while they command an institution whose internal operations we don't see.
More generally, an institution "wired" to do us harm can get away with it much longer than an individual doing it personally and directly. Faceless corporate evil, faceless societal evil, and faceless government evil are much more deadly than our emotional impulses realize. Hypothesis: we are biased to confuse the institutions with our attitude toward their leaders, or to refuse to act against the institutions because of the outward manners of their leaders.
if this 'observation consequentialism' bias is heuristic, then maybe it evolved as an anti-gossip function. In that case we should expect that people are too quick to believe outrageous things about people they can't observe. Hypothesis: the further away someone is from your understanding, the less likely you are to think of them as mostly a typical human being, and the quicker you are to believe them a saint, a monster, or something similarly exciting.
And, alas for EY, hypothesis: telling a story about cunning disguises, in which the protagonist of the story does not see through those disguises, is almost always going to lead to lots of readers also not seeing through those disguises.
Additionally, abusive relationships persist because the victim just can't help but forgive the abuser when the abuser is choosing to be nice. It can be hard to even believe your own memories of abuse when the abuser is smiling at you and giving you compliments.
I try to recall Quirrell murdering Rita Skeeter in cold blood every time I catch myself feeling like he's the good guy in the story.
I think the reason I was reluctant to accept that Quirrell is Voldemort is that Harry is a lot smarter than me and he trusted Quirrell.
That's actually a surprisingly good reason. In real life, the best rationalist you know is probably not a character in a story and feeling a sense of opposing pressure when you disagree with them is probably a pretty good idea.
This should cause you to update down your view of Aumann's Agreement theorem.
(I am reminded of many professional scientists tricked by charlatans when magicians were not fooled- because the scientists knew where to look for truth, and the magicians knew where to look for lies.)
I have updated by learning of it's existence.
Could you explain what you mean by this? I'm having trouble parsing "update down your view of".
Aumann's Agreement theorem is a neat true result about fictional entities. Its applicability to real entities is subjective, and based on how close you think the real entities are to the fictional entities. Increasing that distance makes AAT less relevant to how you live your life, and increasing that distance is what I mean by "update down your view of."
My feeling is that those entities are really distant, to the point where AAT should not seriously alter your beliefs. "I trusted X because Y trusted X" is a recipe for disaster if you trust Y because of different domain-specific competence, rather than their deep knowledge of X.
Right, ok. I'd already thought that AAT is essentially irrelevant to actual human behavior, so I was confused what brought it up.
ETA: No idea why you were downvoted so far.
On fictional evidence?
Harry is eleven.
I'm twenty-one, and I'm hell of a lot dumber than him in every aspect - despite having an IQ in the top one percent of humanity (135).
I generally expect that learning who to trust is something that comes from age and experience more than IQ.
Or it's just the halo effect, since Quirrell is awesome and of course awesome people are always good. You are making things up!
This suits extremely well with both local communities relationship to known criminals and to historical figures. Politics is a mind-killer and so on, but a lot of heroes of different nations have done some downright nasty stuff, but managed to keep their reputation due to perceptions about their personal manner. It has recently been used by leaders such as Chavez and Khomeini, but American presidents have also used this effect extensively (why kiss babies?) and historical figures from Cesar to Richard Lionheart and countless of medieval kings have also garnered good will by the actions they have undertaken in public while at the same time doing something in the opposite direction of way greater magnitude through their institutions of power.
Exactly! That's just like what all the most infamous dictators did, and what Machiavelli recommends in The Prince.
I'm skeptical that people who've taken a long time to accept that Quirrel is Voldemort constitute a significant proportion of HPMoR readers. Sure, I've noticed a considerable number of them too, but HPMoR has a lot of readers. There's a risk of availability bias here; a reader who expresses skepticism that Quirrel is Voldemort automatically attracts attention from anyone who thinks it's obvious, whereas other people who think it's obvious don't.
Personally, I've had no trouble at all accepting that Quirrell is evil ever since his first class, where he praised Harry's killing instinct. Villains pointing out and encouraging protagonists' darker impulses is a time honored trope, and praising an eleven year old in front of a whole class of other children for his drive to kill seems pretty indicative of evil to me.
Part of the problem is what 'he is Voldemort' really means: he isn't like canon Voldemort or even with how MOR Voldemort is reported to be.
As for his obvious evil: it's too obvious, he seems to be the sort who enjoys playing the cynical villain but is actually, if not nice, at least nice to his friends. And Harry seems to be a friend. If he was trying to manipulate Harry he wouldn't have called it intent to kill, he'd have called it being decisive or intelligent or somesuch.
Oddly enough, open villainy can be a great cloak for subtle villainy.
To be honest, I'm not even sure if Voldemort is Voldemort, in the sense of being the man behind the proverbial curtain here. Everything about him from the name up screams "assumed persona": he's far more theatrical a figure than a blood-purist demagogue would need to be, and some aspects of what he does even look counterproductive in that context. And while the canon Tom Riddle did all the same stuff and all for no particularly good reason, in the context of MoR I think we can assume that there's an agenda behind it.
I don't know for sure what that agenda is yet, but a good first step seems to be this question: why would you want to pose as a supervillain? As it happens, Eliezer has touched on that before.
More proof:
Added to this...
...would seem to suggest that Quirrelmort was pretending.
As you pointed out, Eliezer has suggested that humanity might benefit from a Dark Lord to unite against.
And Quirrell has used Voldemort as a reason for magical britain to unite.
To clarify, this is only weak evidence in favor of Nornagest's theory, but it seems like we shouldn't be postulating evil mutants without considering other possibilities.
Quirrell and Harry are both horcruxes of Voldemort, and there is a decent chance that Quirrell has guessed that this is the case by now, if he didn't always know. Quirrell thus has a very good reason to be nice to Harry...they are partially the same person.
But just how much similarity does hpmor Voldemort bear to cannon Voldemort?
Intelligence boost aside, both Harry and Quirrell have the exact same motives as canon Voldemort (power and immortality). The only difference between them is that Harry has an ethical component to his utility function - that's pretty much the only difference between Harry and Quirrell. Tom Riddle for his part is not against ethics - he just doesn't care about them. There are different varieties of evil: let's not confuse amorality with sadism.
So there is absolutely no reason why Quirrell should view Harry as an enemy, except where Harry interferes with his plans because of his morality. If Harry succeeds at all his goals, so does Quirrell (to some extent. There is still the "dominance" component of power, which is a zero sum game. It's hard to tell how much Quirrell cares about that.)
Harry's view of Quirrell is slightly more problematic. Because of Quirrell's lack of ethics constraints, Quirrell has many more options open to achieve his power/immortality goal than Harry does. So while Harry doesn't need to kill Quirrell, he does need to prevent him from achieving is goals in unethical ways.
In fact, my current prediction is that Harry will "win" by achieving Quirrell's goals ethically, thereby making it unnecessary for Quirrell to behave immorally.
Some thoughts...
When reading through the first time, it did seem really obvious that Quirrell was an improved, much more rational version of Voldemort; so blatantly obvious that it made me think if it was a clear red herring. (In the same way that Snape is the canon red herring.) I wondered if Eliezer had reversed things, so that Snape is the real villain and Quirrell the real good guy...
However on re-reading, my prime suspect is now Professor Sprout (Chapter 13):
Of course, everyone knows that, just like they know Dumbledore's not really insane, it's just a cover!
I don't think anyone failed to see the signs that Quirrel is Voldemort in HPMOR. There are just those of us who believed it to be a Red Herring, because "that's how stories are supposed to work." If a potential solution to a mystery seems very obviously true in the first quarter of the story, then in most stories it's probably not the true solution. . Of course, at this point there's just no denying it.
Your third sentence (at least up to the semicolon) should be rot13ed, although the proposition it expresses is pretty well known.
How about the first five words?
OK, I guess.
You deserve far more karma than what you received, my friend.
By the way, could you link me to the argument expressed here?
Right here.
Thanks.
I don't understand why this was downvoted... the original source has been deleted but Glumph posted a link to an accurate copy of it.
Could someone who has been reading HPMOR more assiduously than me say whether and where it has been explicitly revealed, in the story itself, that Quirrel is Voldemort?
It has not.
Ah. In that case, I choose to discount gur qr-choyvfurq nhgube pbzzrag ba gur znggre and predict that Quirrell, as we have seen him so far, is neither Voldemort, nor Voldemort's puppet.
ETA: Edited only to rot13 something and correct Quirrell's name.
I further predict, more speculatively, that Harry will wrongly come to the opposite conclusion, betray Quirrel, and only too late realise his mistake in turning against his strongest ally. Furthermore, Harry will make this mistake through applying what he has learned from Quirrel about good and evil to Quirrel himself.
All predictions based solely on my reading of the published story.
And furthermore, as a result of this, Harry's eventual victory will come at far greater cost than it otherwise would.
Beware the conjunction fallacy. Your scenario is complicated enough that its probability must be small, and also detailed enough that your brain is likely to try and overestimate that probability.
Of course. As I said in another comment, I rate the combined probability substantially below 0.5.
I think Quirrel is dying. He has lapses where he goes into "zombie-mode" and what is that, really? It could be some kind of disease or magical illness- perhaps at the end of the year Harry permanently loses his mentor because the illness has finally killed Q or put him in a coma.
What odds would you give for that?
I'm not interested in a monetary bet, but when I reach into the unknown depths within and pull out a number, it's 80%. For my more speculative predictions, I'd put the chance that I am right in every detail substantially below 50%. I shall be most gratified if it turns out that I nailed it.
Neither am I; I should have said "probability".
...Wow. Really? Bearing in mind that Eliezer is on record as saying he does not deceive his readers with red herrings?
Yes, really. Certainly, Quirrell has some significant relationship to Voldemort, and the questions of who Quirrell really is and what that relationship is have been raised in the fic. But I don't think Eliezer has been deceiving the readers.
Some bits of the foregoing discussion really ought to be rot13ed.
Spoilers ahoy:
Ryvrmre unf pbzzragrq gung Ibyqrzbeg == Dhveeryy zber guna bapr. Va gur puncgre jurer Dhveeryy fnlf ur pnfg n fcryy ba gur Cvbarre cyndhr, gur nhgube'f abgrf pbagnvarq gur fragrapr "Ibyqrzbeg ubepehkrq gur Cvbarre cyndhr!"
The only way you can be remotely correct is if Eliezer has outright lied to us. You are almost directly calling Eliezer a liar. I believe that the probability of Eliezer being a liar is way, way, way lower than the probability that your theory, which is quite crackpottish and heretofore unsubstantiated even without considering Eliezer's comments on the matter, is correct.
Beware the Löbian death spiral.
Lbh ner ersreevat gb qr-choyvfurq zngrevny, juvpu nf sne nf V'z pbaprearq qbrf abg rkvfg. Nalguvat gung Ryvrmre unf jvguqenja sebz pnaba, ur vf serr gb punatr ng jvyy. Zl vzcerffvba (V qba'g ernq zbfg bs gur UCZBE guernqf) vf gung gung pbzzrag bs Ryvrmre'f, juvpu ab ybatre rkvfgf, vf gur fbyr fbhepr sbe D=I. V nyfb abgr gung gur Cvbarre vapvqrag unccrarq jryy orsber gur riragf bs UCZBE; jurgure D jnf= I gura V qba'g unir na bcvavba ba.
I gave in-world reasons, based purely on published HPMOR canon, for thinking Q != V. A meta-consideration tending in the same direction is this. In Rowling canon, Q=V, but this is a dramatic reveal at some point in the first volume. Readers coming to HPMOR having read Rowling, seeing that this is an alternate version of the Potterverse, will immediately be wondering how all the characters of the two parallel universes correspond to each other. Hence the question, as soon as Quirrell appears: is Q V? Now, how can the answer to this question be revealed as a surprise, if the answer is that Q=V? The only way of making a mystery of it is to plant suggestive hints that Q=V and then, when the time comes, reveal Q != V.
Anyway, I'm right or I'm wrong, and the story itself will give the real answer soon enough.