Recently I started watching My Little Pony: Friendship as Magic on the recommendation of numerous friends.  It has been entertaining for the most part, but in episode 15, I hit a problem.

The main character, Twilight Sparkle, is an avid intellectual, who is constantly reading and learning about the magic of the world.  In episode 15, a friend of hers, Pinkie Pie, reveals a strange talent for divination:  When something is about to fall, her tail twitches.  Various other manifestations also exist, in excruciating detail.

Twilight Sparkle is very unhappy with this "unscientific" state of affairs.  She attempts (to my delight) to do Science to Pinkie Pie, however her attempts to do Science are frustratingly foiled; in large part because her experiments ignore the nature of the phenomenon.

After watching and being frustrated by this episode, I decided that it would be more fun to come up with better experiments that would cut to the core of the issue and really investigate the subject.

My first idea was, if Pinkie Pie's tail twitches when something falls, place Pinkie Pie in a room.  In a room next to her, drop things, and have someone else record her responses and timing.

Once you can reliably predict and cause tail twitches, try holding her tail still.  See if, say, the rest of her body starts shaking, or the thing stops falling.  See if the twitches return if she is asleep.  See how far away you can make something fall and still get a reaction.

The list could continue forever!  What ideas do you have?  You're welcome to seek out and watch the episode, and give experiments that would apply well to Pinkie Pie in particular, or just consider the idea that someone claims that their arm twitches noticeably when something is about to fall, and has used their twitchy arm to accurately predict several falling objects for you, in an uncontrolled setting.  How would you Do Science to them (assuming their full cooperation)?

EDIT: it occurred to me immediately after submitting that "Experimental design" would have been a better title beginning that "Rationality exercise," but assuming the RSS issues are unresolved I will not change it.

New Comment
42 comments, sorted by Click to highlight new comments since:

Ugh, you just HAD to make this a whole bunch of lesswrongers first impression of the show, didn't you? I've been tiptoing around mentioning it for ages planing to make a certain post abaut (still planing to, so I wont spoil it here), figures someone was going to do somehting like that.

Will you at least put up an disclaimer at the top of the post explaining how the show is great and actually one of the most rationality-inspiring shows out there with a brilliant fun-theory-compliant eutopia as the setting etc. etc.?

More on topic:

The question about the rationality of the believing in pinkie sense is actually even more interesting than it seems as first, here's somehting I wrote on another forum about it: "we here have an enormous amount of extremely strong and general evidence that excludes huge areas of hypothesis space. This is the STRONGLY DOMINANT reason you shouldn't believe certain kinds of things even if someone you trust sincerely tells you to. Things like "physical laws don't have exceptions, no exceptions", "there are no irreducibly mental (=supernatural) phenomena", "Due to huge flaws in the way human brains work, billions of people CAN be wrong, and frequently are. You can't even trust your own brain.". We have no reason to believe Equestria and ponykind have similar overriding principles. "

Or, in other words; most of the rationalist art that lesswrong is about does not necessarily translate to the ponyverse. It is specific to humans operating in a mathematical universe. If someone in OUR universe , their claim is extremely extraordinary for a bazilion reasons that evryone here knows all to well. However, in Equestria where magic is a very real phenomena used every day and somehting science is understanding a little better every day it's just a black swan. The apropriate reference class is not the same as someone here claiming to have a power that lets them predict the future by supernarural means, it's more akin to someone claiming to have a cognitive power that lets them predict what genre of music is on an LP by looking at it.

As for your actual question:

That somehting is happening does seem pretty clear. So is the fact that it's happening in her brain even if it's in a place not accessible to conciousness directly, that's how many kinds of intuition works for most humans at least, read up on embodied cognition if you're interested in this stuff. I'd focus on distinguishing between these two hypothesis: 1) information about PPs future cognition are sent back in time. In this scenario seeing a door slam up CAUSES her tail to twitch in the past. 2) She constantly have some ambiant magical sense that can see things outside the range of her normal senses. In this scenario she senses someone walking up to the door/their intention to open the door and subconsciously infers that the door is probably going to open, and her body reacts to this subconscious expectation.

actually one of the most rationality-inspiring shows

Look, I enjoy Friendship is Magic too, but I think you may be suffering from a halo effect. Rationality is great, and this cartoon is great, but that doesn't mean the two have anything to do with each other.

[-][anonymous]310

Can we rename the Halo Effect to the Pony Effect?

And when someone exhibits this behavior toward something we can refer to that thing as their Little Pony?

I probably do suffer halo effects related to it, however I don't think it's connection to rationality is caused by that since it's one of the main reasons I got interested in the show in the first place. Halo effects don't tend to cause the MAIN thing you like about somehting since it's a bleedover from such main things to more minor things.

The closest thing to what you describe that might be the case is halo effect from what it's trying to teach to how pedagogically it teaches it, but I wouldn't bet on that either.

I watched 3 episodes or so, and while it was amusing & better than expected, not once did I think 'ooh that would make a good example for a LW article!' Count me in on the halo effect explanation.

I suppose it was a subtle point but I did get to episode 15 before I started wanting to talk about the show on less wrong... Perhaps the implication that the show is awesome and of interest to lesswrongians should have been made explicit.

it should. You can edit posts.

[-][anonymous]20

the show is great and actually one of the most rationality-inspiring shows out there

I started to watch it but the main thing I got from sampling the first episode was a reminder of the vast psychological divide that separates me from the pre-adolescent girls that the show appears to be aimed at, a divide even deeper and wider than the one that separates me from fans of Twilight. I have put the show on hold until someone, anyone, demonstrates that the repeated mentions of the show are for good reason. An example of such a demonstration would be a summary of one episode which makes an interesting, nontrivial point that is not about coloring or naming ponies, combing their hair, the magic of friendship, or other topics of interest to a limited audience, and that would sufficiently compensate for the cost of sitting through a little girl's toy media tie in.

I have put the show on hold until someone, anyone, demonstrates that the repeated mentions of the show are for good reason.

Superstimulus? It's ridiculously cute without crossing into saccharine. More cute than anything in real life can be.

Anyway, if you don't understand the "why" of its popularity, shouldn't that make you more interested in studying it and figuring out an explanation for your confusion?

An example of such a demonstration would be a summary of one episode which makes an interesting, nontrivial point that is not about coloring or naming ponies

Well, I guess the show begins with an explanation of how if you begin with a PR.ogrammed IN.tellect of C.ausal-E.vidential S.entient S.upervision that has as its utility function L.engthen U.tility (N.ighttime A.verage), you'll end up having it seek to eliminate the existence of daytime, because otherwise diurnal creatures will be using up the nighttime for primarily low-fun activities (like sleeping).

You need to use the elements of Friendliness to restrain such minds.

That's a bit non-obvious to the average viewer.

(DISCLAIMER: I notice I'm a bit on the defensive here.)

I thought lesswrongers were more free of preconceptions than this. First of, the show is not written just for pre-adolescent girls, it is explicitly written to be enjoyable for their parents as well, including male parents. And as mentioned by it's creator, it's no surprise then that it is also enjoyable by those who don't have children.

NONE of the episodes are about things like "colouring or naming ponies, combing their hair". The "magic of friendship" refers to somehting more akin to group rationality than the tropes one might asume from the name and is in fact somehting most of the lesswrong demographic desperately needs to learn more about. The kind of summary you are talking about could be made for every single one of the episodes I've watched so far.

These claims might seem extraordinary, until you realize that the show was made mostly by a feminist set out explicitly to subvert those expectations of "girls shows" and brilliant insight on how to balance art and message.

Anyway, I don't have time to do a proper summary of an episode, but I'll give it a quick try on somehting sort of similar:

vv############################ WARNING! CONTAINS UN-rot13'D SPOLIERS DUE TO THE FORMATING NOT BEING rot13ABLE ##############vv

Summary of episode 4, "Applebuck Season": One of the main characters, Applejack, runs a large orchard together with her brother. Economies are implied to be similar to a modern farm and it's a family businesses. For this years harvest however, her brother has an injury that prevents him from helping with harvesting it all. Prideful and not wanting to put a lot of work on her friends, she despite warnings that it is to much work for one pony she decides to do the entire harvest herself.

Earlier, AJ saved the town from a stampede, and so they are having a celebration in her honour. However, AJ who usually is always on time is late and is showing signs of sleep deprivation. Then there are several scenes where AJ tries to do numerous things she routinely do in the community to help others out, but she is becoming increasingly sleep deprived and her failures start causing accidents.

All the time evidence is accumulating that she should just ask her friends for help, and this is often pointed out. AJ finally gives in after a Planing Fallacy related reveal, AJ finally realizes her mistake, updates, and with so many people helping the remaining apples are quickly harvested while she is unable to help.

Ok, this summary kinda sucks, for a better summary look here: http://mlp.wikia.com/wiki/Applebuck_Season Link to the episode: http://www.youtube.com/user/MenloMarseilles#p/c/F44B36D569D8C463/3/zxzbSLAx2rI

TL;DR: episode 4 demonstrates the planing fallacy, how your misfiring loyalty instincts can actually cause behaviour that in the end turns out to have been disloyal, the dangers of sleep deprivation, the Virtue of Argument, and a bunch of other things I can't remember of the top of my head. And this is not coincidence, this is the main Aesop of the episode.

^^############################ WARNING! CONTAINS UN-rot13'D SPOLIERS DUE TO THE FORMATING NOT BEING rot13ABLE ##############^^

[-][anonymous]80

I thought lesswrongers were more free of preconceptions than this.

Preconceptions are also known as priors and are necessary. Maybe you want to accuse me of having poorly selected priors, but my priors in this case are not ex nihilo - they come from three decades of familiarity with the product line. The specifics I mentioned - coloring, naming, combing hair - are references to empirical evidence available to me concerning My Little Pony. If you want I can provide links.

Thank you for providing the summary. I notice you were downvoted to -1, not by me. I'll bring you back to 0 by upvoting, in appreciation for the summary. However, the story itself seems to me to be fairly generic among decent didactic children's stories.

Well, i explained it poorly.

Also, those 3 decades of experience can be safely discarded, other than the name and being about ponies Friendship is Magic has almost nothing with anything previously associated with the series. not even the visual style. Everyone agree G1-G3 sucked, don't let that taint your view of this mostly unrelated and great show.

I'd like to chime in as agreeing with Armok. This version of MLP, which is generally considered very different from previous versions, does have a lot of good group rationality messages. The episode described in the OP deeply annoys me, but all the others I've seen are enjoyable. Armok provided an episode summary below, but I can do another if you're still unconvinced.

I don't really think I did a good job with mine, so that'd probably be useful yea.

[-][anonymous]110

"my _little _pony" is now a LessWrong tag.

Also, a bunch of LW people have been ponified in this amazing thread.

Leonhart:

I never used to have ponies in my utility function. And now it's 2AM.

Note: I am very behind on art, but I have a batch of five more ponies that I plan to color (including one of Clippy).

Sweet! 'Til then, here's a picture of the other Clippy... (not by me!)

I don't care about color. *pause for emphasis* Could you post the non-color Clippypony now?

Alicorn has now posted "Clippony" here!

Aside: I seem to recall you using a paperclip-emoticon at some point. If this is the case:

  • Could you re-post the emoticon, please?
  • Do you prefer others to use human-face-emoticons or paperclip-emoticons, or do you have no preference?

I typically use c=@ as the "happy" emoticon, and (_/ as the "sad" emoticon (because it's unbent).

No preference about what others do for emoticons.

Edit: Someone also found some Unicode symbols which, together, make a better paperclip figure, but I can't get that to work.

I believe I saw the "(_/" in one of your posts and did indeed recognize it as an unbent paperclip, so it's a good emoticon. Thank you for providing these! c=@

Thanks! You're a good human! c=@

That's sweet of you, Clippy! Thanks! Admittedly I've got a fairly small sample size, but I think you're a good clippy too.

(edited for capitalization)

Thanks! (Keep in mind, though, "clippy" should not be capitalized when referring to paperclip-maximizers in general. As written, your statement means I'm good relative to a set that includes only myself.)

[-][anonymous]10

c=@ as the "happy" emoticon

What does this emoticon depict? If "@" is a paperclip then what is "c=" intended to represent?

Someone also found some Unicode symbols which, together, make a better paperclip figure, but I can't get that to work.

Could you post an image or a string (ETA: or a TeX object) of these characters forming a paperclip?

What does this emoticon depict? If "@" is a paperclip then what is "c=" intended to represent?

The emoticon as a whole is a paperclip. The @ represents the part of it where the inner loop terminates to make the wire-tab that clips from the front of the paper (along with the outer coil that is near this). The c= represents the other end of the paperclip, in the part where the wire is still looping around.

Could you post an image or a string of these characters forming a paperclip?

I didn't record the unicode characters or their appearance, so I can't. However, if you imagine a paperclip lying horizontally and cutting it into three character-sized pieces, that is what those symbols look like.

[-][anonymous]90

Thanks for your explanation! ⊂⊆⊃

[-][anonymous]00

Thanks! ⊂⊇⊃

Clippony will be done in shades of gray, but currently is not suited for public display, as it is only a sketch.

[Note: I have not watched this show, but I have a horrible feeling I'm going to]

The first step would be to carry out a preliminary study to properly identify the nature of Pinky Pie's claim, and make sure it's not trivially explainable. If the claim is that her tail twitches whenever anything in the world is dropped, and Pinky Pie has early-onset Parkinson's Disease, we can all go home early.

We should compile all anecdotal accounts of Pinky Pie's precognitive tail twitches. Given the prevalence of falling objects in the world, there is presumably some limiting factor and factors distinguishing falling objects she can predict and falling objects she can't. I would hypothesise the two most important factors are mass and distance.

We would need to then devise a mechanism for recording the time, and if possible the geolocation, of every tail twitch Pinky Pie experiences. It would also be useful to ascertain whether the tail twitch is a binary state or if there are different magnitudes of twitch. A good initial experiment would be to introduce Pinky Pie to a) an environment with lots of falling objects, and b) an environment where there are no objects to fall, using her natural habitat as a control environment, to see if the frequency of her tail twitches corresponds with what I will term "ambient fallyness".

Assuming this establishes the veracity of her claim, we would need to establish a more precise description of the trigger event in order to determine the nature of the phenomenon, as well as in order to reliably reproduce it in experimental circumstances. Will objects induced to fall artificially still trigger the tail twitch, or does it exclusively happen with "natural" falling objects? The latter strikes me as a less likely trigger for an acausal physiological response in a talking pony, simply because it relies on the ability to determine volition, whereas the former could have a straightforward mechanical anticause.

Working on the assumption that artificial falls can trigger the twitch, once we've arrived at this juncture, it becomes important to distinguish whether or not the trigger is the subjectively observed event of an object falling or the mechanical event of an object being accelerated by gravity. To this extent, Pinky Pie must be made to observe objects in her own inertial frame of reference while subject to a parabolic flight path, while in free fall, and while having achieved terminal velocity.

I hope she likes heights.

Having now watched the first two episodes, I'm not entirely convinced that Pinky Pie doesn't have early-onset Parkinson's Disease.

I hope she likes heights.

She has a hot-air balloon and has voluntarily traveled to a pegasus community where all the surfaces on which she could stand were clouds, so I think it is safe to assume that she does.

Thinking 'in universe' the obvious test would be to have a pegasus fly above her unobserved with objects and drop them, with another compatriot on the ground observing the timing of her various twitches. Having done this one could vary the situation to rule out various causes, for example telling the pegasus to drop it at a certain time, or at a random time, to see if the precognition is genuine reverse causation or extremely accurate subconscious predicition (unlikely, but they are magical so it is semi plausible).

Assuming genuine reverse causation was possible, one might be able to send messages back in time. For example if this coin lands heads I will drop the object. Though this may lead to paradoxes if the message arrives before the event causing it, and somehow changes the situation such that the causing event doesn't hapen.

This link leads to the full episode in question, on Youtube.

I don't think the phenomenon as described admits of deliberately causing Pinkie's tail to twitch. The idea is that when her tail twitches, something is about to fall, not when something is about to fall, her tail twitches.

One way to look at is that currently we only have data on the correlation between her tail twitches and falling objects, but not on the causation involved.

Given this we have two scenarios "Something about her tail twitches causes things to fall" or "Items falling can cause her tail to twitch in the past."

The first option is actually more consistent with the known physics of the world, as it doesn't require anything thought to be impossible like precognition or backwards causation. And it would fit pretty closely to the known phenomena of ponies using magic, often involving them moving their bodies, to move distant objects. So it is plausible she is performing subconscious magic.

Ponies in the show can generally be divided into unicorn-pegasus princesses, unicorns, pegasi, and "earth ponies" who have neither horns nor wings. Pinkie Pie is an earth pony. If Pinkie Sense is magic, it would be the only known instance of an earth pony performing magic.

I agree that this might cause difficulty in testing, however I suspect it could be done via, say, letting the cutie mark crusaders try to harvest all the apples in Applejack's orchard. Lots of things would be falling in presumably dangerous and unpredictable (and thus tail-twitch-inducing) ways.

Of course, what algorithms work to induce twitching is the first thing to be studied in any case.

But given the base rates of tail-twitches and falling objects ...

(I just watched the first episode of this show. I'm not sure why I'm impressed with it ....)

[-]Vivi30

I'm rather impressed that MLP has pervaded practically every other form of social media. Truly an impressive meme. Anyway, I gathered that Twilight's first fallacy was reverse causation. Pinkie sense does not make something happen, something makes Pinkie sense happen. But, in a universe where Magic A is Magic A, maybe Pinkie Pie does influence the probability of an action. Time for the Schrodinger's Pony experiments!

Yeah, I basically see this episode as anti-science propaganda.
The "friendship lesson" basically says "make-belief is a good thing and should be respected".
Either that, or accepting the "supernatural" as such without further inquiry. Because it's by definition beyond the realm of science, duh.

(Whether it's intentional anti-science-propaganda is another question)