See also: Everything I Needed To Know About Life, I Learned From Supervillains

Mr. Malfoy would hardly shrink from talk of ordinary murder, but even he was shocked - yes you were Mr. Malfoy, I was watching your face - when Mr. Potter described how to use his classmates' bodies as raw material. There are censors inside your mind which make you flinch away from thoughts like that. Mr. Potter thinks purely of killing the enemy, he will grasp at any means to do so, he does not flinch, his censors are off.

-- Quirinus Quirrell

A while back, I claimed the Less Wrong username Quirinus Quirrell, and started hosting a long-running, approximate simulation of him in my brain. I have mostly used the account trivially - to play around with crypto-novelties, say mildly offensive things I wouldn't otherwise, and poke fun at Clippy. Several times I have doubted the wisdom of hosting such a simulation. Quirrell's values are not my own, and the plans that he generates (which I have never followed) are mostly bad when viewed in terms of my values. However, I have chosen to keep this occasional alter-identity, because he sees things that would otherwise be invisible to me.

-- Quirinus_Quirrell

I was once asked whether I would rather be a superhero or a supervillain, and I probably shouldn't tell you how little time it took for me to answer "supervillain."

Being a superhero sounds awful, at least if you intend to keep being recognized as a superhero. Superheroes are bound by the chains of public opinion. A superhero can only do what people generally agree is good for superheroes to do. If you stray too far off the beaten path in search of how best to use your superpowers to actually save the world, you could easily end up doing things that look, at first glance, somewhat to incredibly evil. And if people are going to turn against you once you start actually optimizing, you might as well just be a supervillain to begin with. They look like they're having more fun anyway. 

You probably won't get the chance to decide between being a superhero or a supervillain, but you do get the chance to decide what kind of person you think of yourself as, and I think you should think of yourself more as a supervillain than as a superhero. Why? 

In the same way that being a superhero limits what you can do, thinking of yourself as a superhero limits what you can think. And if you want to save the world, you can't afford to limit what you can think. Humanity faces many difficult problems, and the space of possible solutions to any one of these problems is large. If you have censors in your mind that are preventing you from looking at parts of this space because some of your moral intuitions don't like them ("that's not the kind of thing a superhero would do!"), you're crippling your ability to search for solutions to problems. For example, your moral intuitions are likely to flinch away from solutions to problems that involve you causing bad things to happen but be okay with solutions to problems that involve you failing to prevent bad things from happening (think of the trolley problem, or Batman's policy of not killing his enemies). 

Edit (2/19): But thinking of yourself as a supervillain has the opposite effect. It's easier not to flinch at certain kinds of ideas, which now come more easily to mind and may not have otherwise occurred to you. For example, on Facebook, Eliezer recently mentioned a thread where people were posting examples of things that they valued at a billion dollars or more, such as their cats. With a supervillain module running in the background, I noticed and pointed out that this constituted a thread where people publicly described how they could be ransomed. I can't exactly test this, but I don't think this kind of idea would have occurred to me before I installed the supervillain module. (This is a tame example. I won't give less tame examples for obvious reasons.) 

There are many things you can't say, but you don't have to say everything you think. Until someone discovers a technique for reliably reading human minds, think whatever thoughts best help you accomplish your goals without worrying about any moral labels they may or may not, upon reflection, ultimately warrant. Moral labels are for a later step in the decision process than the part where you generate ideas. 

New Comment
111 comments, sorted by Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

There's another related aspect that's worth noting: supervillains are active, superheros classically reactive. The Joker hatches a plot and Batman stops him. Brainiac threatens to take over the Earth and Superman stops him. Doc Oc tries to blow up New York and Spiderman stops him. Etc. Etc. Ad infinitum et nauseam. If there's not any supervillain active on any given day in Gotham, Batman sits around preparing to fight them, letting most of the status quo stay unchanged.

To think about changing the status quo, think like a supevillain.

I so adore tropes, they give me something to subvert.

"According to all the stories, this ordinary world is what the extraordinary people try to protect. If you read a comic book about superheroes, it would be about superheroes defending all those everyday lives. The superheroes wouldn't be trying to cure AIDS or feed starving children in Africa or otherwise change the world. We have scientists for that sort of thing. No, a superhero is someone who defends that ordinary, everyday life from the forces that try to change it. Even if those stories come from our imagination, still, those are the people we praise above all others."

See also: HPMOR

There do exist at least a few superheroes who are trying to change the status quo proactively. I don't actually read that many comics, so I could be missing quite a few instances, but here are some examples:

  • Tony Stark a.k.a. Iron Man (movies): stops developing weapons and begins developing Arc Reactors instead, in order to eventually supply the world with limitless free power.
  • Tony Stark a.k.a. Iron Man (Civil War comic arc): dedicates a lot of effort to the Superhero Registration Act, in order to make superheroes accountable for their actions. Is ultimately proven wrong, I guess.
  • Professor X: wants to make the world a more tolerant place through a combination of charitable donations, public outreach, and, when absolutely unavoidable, mind control. Contrast his approach with that of Magneto, who wants to accomplish essentially the same thing through conquest.
  • Reed Richards: pretty similar to Iron Man, in that he is slowly trying to uplift the rest of humanity to his own super-technological levels. Creates the whole Negative Zone fiasco when he decides to think like a supervillain during the Civil War arc.
  • The Flash (cartoon, some comics): When he's not facing the existential thre
... (read more)

Daenerys Targaryen (books): Though initially motivated solely by revenge and personal survival, she stops long enough to overturn several existing social orders in order to improve the average quality of life.

Nearly all of which turn out terribly if I recall correctly. Her freeing of the slaves for example on a utilitarian scale is somewhere between a particularly bad natural disaster and The 30 years war, especially since it seems unlikely to last.

Besides being basically at the head of a marauding horde that is a menace to settled civilization there are further reasons to doubt she has made a positive impact so far. My girlfriend recently had an interesting fired monologue on the subject, she hates the character and sees her as behaving like the worst possible stereotype of Western (in particular American) doogooderism and interventionism. And yes I did say stereotype so I'm not saying what follows is an accurate description of real world affairs.

  1. Encounter stable working society
  2. Deem noticeable features of it immoral
  3. Not bothering to study the society use violence to enforce your morality
  4. Notice things becoming complicated and incredibly messed up, like obvious great suffer
... (read more)
I suspect the TV show may end up reducing, if not the scope, at least the emotional empact of the harmful fallout of her anti-slavery actions. Pop culture tends not to play well with values dissonance. It is known.
In terms of disliking her as a person, or as thinking she detracts from the story? I more or less agree with her description, but while I dislike her in the former manner, I definitely don't dislike her in the latter. I'd be pretty disappointed if she comes out on top in the war though.
Haha yeah, that analysis is pretty much correct. Of course, we could argue whether a "stable working society" based on (f.ex.) mass slavery is moral or not; i.e., whether military action against it (and thus the resulting deaths) would be justified.
Both Dany and Jon, to point to obvious examples, are almost classically heroic in their actions and cbffvoyl gurve fgnghf nf zrgnculfvpnyyl qrfgvarq urebrf. Samwise, Brienne, Stannis, Ned, &c. are pretty straighforwardly heroic. The universe is written with a bell curve rather than bimodal distribution of morality, and it assumes that things like nitty-gritty politics actually matter, so it's easy to pattern-match it "there are no heroes," but I don't think that's particularly true.
I find it difficult to pattern-match any of the characters to the classic hero template, especially when you compare them to traditional hero archetypes such as Green Lantern or someone like that. As I said, Dany is initially motivated by personal survival, with a dose of revenge fantasy on the side. Her actions are impressive, but hardly selfless. Her motivations do change as her character develops, at which point she begins to think in terms of social structures and armies -- again, as opposed to a more classical hero who would think in terms of beating up bad guys in person. You are right about Jon being more of a heroic archetype, but even he ends up making several distinctly un-heroic choices that cause a lot of damage to the... well... not the "good guys", exactly; I guess you'd call them the "comparatively less bad guys". Ned is probably the most heroic character in the entire story, which is why ur trgf xvyyrq bss engure dhvpxyl. Urebrf qba'g ynfg ybat va gur jbeyq bs NFbVnS. Oh, and I am reasonably sure that the [quasi-]supernatural properties of any of these characters will have little, if anything, to do with their ultimate fates (other than in terms of PR). At least, this has been the pattern so far.
I only have indirect knowledge of most of these characters through friends who actually follow their series, but I'll note that Reed Richards is iconically ineffective at the whole uplifting deal (tvtropes link).
Well, his character changes between different comics and cartoons. In some, he is indeed useless. In others (IIRC), he introduces things like cheap clean energy, fast transportation, and even competent economic planning at some point. This applies to most of the well-known characters, actually, as every team of writers put their own spin on them.
Doesn't he act in characteristically supervillainish ways during that arc? Cloning Thor and all that.
He does, it's true, I forgot it was him who did that. In Iron Man's case, thinking like a supervillain had worked out... poorly.
IIRC, there was an Elseworlds comic where he acted slightly differently and it worked out fine. It was framed with a Mysterious Stranger offering to show him what would have happened if he chose differently, as I recall. So it was just bad luck, I guess?
I don't know, I haven't read the Elseworlds version. People tend to make their own luck though, good or bad...
checks If he hadn't threatened Captain America just before Clor went crazy, Cap would have joined forces with him to fight the Thor clone, ended up joining the pro-registration side, and he's so damn trustworthy that everyone would have signed up. So yeah, it was basically karma. Oh, and it was actually a What If?, not an Elseworlds. Those are DC, aren't they?

If there's not any supervillain active on any given day in Gotham, Batman sits around preparing to fight them, letting most of the status quo stay unchanged.

Well, his secret identity spends his time making large amounts of money and giving it to charity.

I think you mean "reactive". Unless you're thinking of some superheroes who do an aikido-esque "I'm just going to get out of the way and ... oops, your plot seems to have blown up in your face" sort of thing.
Yes, fixed thank you. (Although the newspaper comic version of Spiderman might actually fall into the passive category pretty often- that seems to be the exception.)
To be fair, I think this is more a fact about the medium (written stories) than about the characters. It's much easier to write something in which your protagonist reacts rather than being the first mover. This would not necessarily be the case when extrapolating the character outside the context (e.g. when faced with a dictatorship, a superhero may act to overthrow it).

The point is that a superhero can't take preemptive action. The author can invent a situation where a raid is possible, but for the most part, superman must destroy the nuke after it has been launched - preemptively destroying the launch pad instead would look like an act of aggression from the hero. And going and killing the general before he orders the strike is absolutely out of the question. This is fine for a superhero, but most of us can't stop nukes in-flight.

A dictatorship is different because aggression from the villain is everywhere anyway - and it's guaranteed that we will be shown at least one poor farm girl assaulted by soldiers before our hero takes action against the mastermind. Only when the villain is breaking the rules egregiously and constantly is the hero allowed to bend them a bit.

If you have a situation with both an antihero and a hero in it, the hero can be easily predicted - as opposed to the antihero,who is actually allowed to plan. Superheroes end up quite simple, since the rules they obey are so strict, they can only take one course of action (their choices tend to be about whether they follow the rules or not, and not between to courses of action that are both allowed). And that course of action often isn't the most effective.

It really depends on what we mean by "superhero". If we stick to the archetypal Western examples, you're probably right. But things becomes less clear if we consider something like Watchmen, V for Vendetta (V is pretty super), or the many gray area types Marvel and DC have (I'm leaving that vague because I'm not too familiar with the canon), not to mention various manga and anime heroes (Lelouch?) But maybe we wouldn't call them superheroes precisely because they don't fit the "only act in response to clear, certain evil". Mostly, I think this points to the fact that, to no one's surprise, {Supervillain, Superhero} is not a comprehensive summary of thinking styles, nor are they sharply defined categories.
Many of those examples are probably classed as antiheroes. Punisher would be a great example of a hero who changes the status quo (by shooting it.)
Aim to be an antivillian? Someone who wants to conquer the world and rule it with an iron fist, but is unwilling to use evil means to do so...

Security thinking. If you want to keep your systems secure, you have to model intruders quite well, such that if you look at a device or environment you immediately notice how a malefactor could do things you don't want. You can do this and be the good guy.

Poacher turned gamekeeper works too. I worked at an antivirus company a while ago. Many of the researchers had been virus writers and l33t h@xx0rs in what they now regarded as their idiot youth, but had spent many more years using hacker thinking to defend against the people still doing that, and greatly enjoyed it.

You don't want to think like just any old supervillain. Most of them have systematic flaws in their behavior too. Besides the obvious, in a lot of stories the motive that best explains the villains' behavior is not really to succeed but the one the author has for them, which is to put up a good fight and be awesome, but lose. If you try to be like them, you might end up just trying to try. Your mind will censor out the non-grandiose but effective plans. You do not want to be about demonstrating a good effort. You want to be a winner.

Edit: the next paragraph has spoilers in it for Watchmen. So does the single line after it.

If there's one supervillain people should maybe try to emulate, it's this guy). He defeats an invincible character, stronger than Superman in a world where no one else has significant super powers, with nothing but mind games, (and in the movie at least), nukes several cities and blames it on said godlike character (the comic book ending has him blaming it on aliens, I like the movie ending better), tricks the US and USSR into thinking they have a common enemy, and prevents a larger nuclear war. He exemplifies the utilitarian answer to the trolley problem you ment... (read more)

The rparen in your Ozymandias link needs escaping: [this guy](\)) --> [this guy]( (I estimate p=0.5 of my not having screwed up the markup myself, but I'll fix it if I have.) EDITED to add: harrumph, it appears that LW undocumentedly implements an extension to Markdown where a bare URL is turned into a link, and I haven't found any way to stop the linkifying without also changing what is displayed. So you might notice that the colon above is italicized, which is a really crappy workaround. Maybe there's a better one.
If you wrap text in a pair of back ticks (`) then it gets displayed as "code" so left unmodified by the markdown parser. (E.g. [this guy](\)))
fixed, thanks.
I'd consider rot13-ing the spoilers, just in case.
I put in a text warning before the spoilers, this should work I think. Rot13 is annoying in my opinion.
I used to be annoyed by it, but then I added a rot13 selection bookmarklet to my toolbar.
There's no "spoiler" tag available in LessWrong? Too bad. Anyway, even for the comic book villain, it probably won't work out; at the end of both comic book and movie his big secrets are clearly about to be leaked to the public, which will be grossly hazardous to both his goals and his health.
Moore leaves it ambiguous; Rorshach's journal is being pulled out of the slushpile but that's at something Moore has modeled on the Birchers. I suspect that that name doesn't mean much to most LWers, so I'll elaborate. To put it in perspective: suppose you read in the Bircher periodical The New American (which in the same issue alluded to how the US government is controlled by Russians and how fluoridation is a government plot, probably due to the Jews*) an expose based on an anonymously submitted handwritten journal supposedly written by the serial killer Gary Ridgway about how 9/11 was a plot by philanthropist Bill Gates to provoke military intervention against failed states. I think the odds are very small anyone would take it seriously. * This is pretty much what the Birchers have officially held at various times; it was the Birchers who Dr Strangelove was mocking.
Sure, in the same way that if I wrote a post called "think like a scientist" about how you should test your hypotheses it would be reasonable to respond "you don't want to think like just any old scientist..."

This post is in Main and not Discussion so I can get some clarification on what belongs in Main and what belongs in Discussion, and also because of what looks like a demand for more non-meta posts in Main. I'd be happy to move it to Discussion if people think this is too short or doesn't have enough citations or something to be in Main.

I would expect some more "meat" from a Main post; for example, some hard empirical data that points to the effectiveness of thinking without internal censors -- assuming that anyone had ever studied this issue, of course.

Fair. Ordinarily I would've given examples, but given the nature of the subject...

Edit: I think this point is more important than I originally thought and have moved the post to Discussion accordingly.

Agreeing that it should have been in Discussion.

I approve of this post being in Main and I approve of your accommodation to the appearance of that demand and I approve of your testing what might be but I hope are not the boundaries of what people want to see in Main.

If your point is that the best way to achieve the terminal values of a superhero is to have instrumental values of a supervillain, then you better offer a way to cope with the unreliable wetware your utility computation runs on.

If this comment was written by a bot that produces phrases maximizing the ratio of the number of usages of pleasant-dopamine-buzz-in-group LessWrong language to non-in-group language, it would produce something like this. I say this even though I really appreciate the comment and think it has genuine insight.
When in Rome... But yes, you are right, of course :(
I can't reasonably claim to know the best way to do anything nontrivial like that. I'm only claiming that your brain has certain kinds of reactions to certain kinds of ideas and that it would be a good idea to tweak these reactions so that they're more like the reactions a supervillain would have than the reactions a superhero would have.

I am not saying it's not fun. Note, however, that it is reasonably easy to let go of your morals once in a controlled way, just to see what other options are available, and then return to your moral ways if you find nothing acceptable. It's much harder to do it repeatedly. It's harder yet to go back to the status quo if you actually took one of the darker paths for the greater good, "just this once".

I really think you need to qualify this statement a whole lot; more than you have done in the OP. Otherwise, it sounds like you're saying, "before you do anything else, get yourself a huge leather chair and a fluffy white cat to pet while sitting in it". This would, admittedly, be a lot of fun, but it has little to do with rationality. In addition, assuming I understand your real point correctly (and I'm not at all certain that I do), I do admit that it feels really persuasive. Thus, I am instantly suspicious of it, and I would really like to see some evidence. Does the act of deliberately thinking like a villain improve a person's performance across any kind of a measurable metrics, or in fact affect it in any significant way ? I would love to see some sources.
I don't have evidence in the sense that I can't point you to controlled studies. I'm working off of general principles, e.g. the general principles behind the tragedy of group selectionism, as well as extrapolating from Quirinus_Quirrell's experience and mine. I guess I'm also making two points that I didn't do a good enough job of distinguishing. The first is that if your narrative identity contains more superhero aspects than supervillain aspects, you may be more likely to flinch away from what I think is the correct answer on certain kinds of problems, e.g. trolley-like problems. One way to test this would be to prime groups of people into thinking about superheroes resp. supervillains and then give them trolley problems. I don't know if anyone's done this. The second point is that thinking like a supervillain can also make it easier to generate certain kinds of unsavory but potentially useful ideas. One way to test this would be to prime LWers into thinking about superheroes resp. supervillains and then make them play the AI in AI Box experiments. I am pretty sure no one's done this. I don't have the resources to run the first test. The second test is not a very good test (it would be hard to control for the other factors relevant to success in the AI Box experiment). If the way I wrote the OP or the fact that I put it in Main suggested that I had strong evidence that this is a really good idea, that was a mistake on my part.
I don't believe that the Tragedy of Group Selectionism is entirely analogous, since that scenario deals with epistemic as opposed to instrumental rationality. And, while your and Quirinus_Quirrell's experiences are definitely interesting, that doesn't mean that they are representative (of course, it doesn't mean that they aren't representative, either). I think the experiments you propose are promising (the first one more so than the second, as you said). If you could dig up some studies on the topic, or run some yourself, the results would definitely make a good Main post (IMHO). No no, I did not get that impression at all -- that's why I originally pointed out that, due to the lack of evidence, this article probably belongs in Discussion more than in Main.

Thinking like a super-villain is the wrong advice to give to this demographic since it will prime them for counter-productive patterns of behaviour that work in fiction.

Don't most supervillains fail in fictional works?
They eventually do. But they often start with an origin story that makes little sense and during the story they often end up becoming stronger and stronger right up until the end.

I think this is just a limitation of comic book superheroes. They desire public recognition. In other traditions with analogous figures, particularly religion, being reviled is just another burden to be taken on by the hero. (Although this sometimes happens in comic books too. See the recent Batman movies.) I especially like the Tibetan Buddhist concept of "crazy wisdom." Tibetan folk heroes spend a lot of time shocking people out of their complacency and generally acting like supervillains. But it's all in the name of universal compassion. (Google "Drukpa Kunley" for a particularly entertaining example.)

I was once asked whether I would rather be a superhero or a supervillain, and I probably shouldn't tell you how little time it took for me to answer "supervillain."

One of the defining psychological characteristics of supervillains is the desire to defeat death. Count me in.

So are there good comics that have the protagonists thinking like supervillains and pull off something interesting? There's a bunch of supervillains-as-antiheroes stuff of course, but that doesn't seem to stretch that hard against story conventions. The interesting stuff has the protagonists actually successfully carry out good intentions using supervillain style problem solving. Unfortunately, it's hard to serialize a comic where problems actually get properly solved very far, and pathologically serializable comics rule the market.

Alan Moore's Miracleman, started in 1982, is probably the archetypal example of completely ignoring the concern of being able to write and sell the book five years from now when writing the plot. The superman will fix your planet and your children and you and he knows he knows better than you. And it ends up mostly working out okay. Neil Gaiman actually started an interesting looking arc from where Moore left off, with earth having become weird, but the series got cancelled before it got completed.

Mark Gruenwald's Squadron Supreme miniseries for Marvel in 1985 had a superhero group start fixing things and ended up making everything dystopic. J. M. Stracz... (read more)

How about Superman: Red Son? Superman grows up in 1950s USSR and quickly rises to leadership of the country, then takes over most of the world (save for the US) while liberally using lobotomies to maintain power. Nevertheless, he turns the world into a relative utopia, and continues to perform typical superheroic acts literally everywhere. His foil is Lex Luthor, a super-genius and manipulative bastard whose only goal is to bring Superman to ruin. (And then there's that question... those who have read it will know what this refers to.)
It's important to note that the incomplete later segment of Miracleman was apparently going to make things not work out okay.
In the X-Men Comic Cable and Deadpool (2004-2008) Cable makes a a proactive effort to improve things, attempts to create a little utopia. While it all goes pear shaped, I got the impression that was mostly for editorial reasons (actually I only got interested in the series after the cross-over in question de-railed things). Haven't re-read it recently, but it got me asking questions at the time about whether a superhero with knowledge of the future could actually make things better.

If this gets upvoted highly, I will update in favor of LessWrong continuing to become more in-group-y, more cutesy, and less attached-to-actual-change-y. It's becoming so much delicious candy!

This is the comment a super villain would make if he wanted less competition.

I don't think it's productive to think this way. Yvain wrote a great post which I currently can't find where he points out, among other things, that it's generally a bad idea for your primary reaction to an event to be a reaction to how you think it fits into an overarching narrative (e.g. "this just goes to show you can't trust those dirty Greens"). The LW community doesn't strike me as homogeneous enough that it's productive to model it using in-groupiness, cutesiness, and attached-to-actual-changiness parameters that can be inferred from current posts and that determine the value of future posts. Evaluate posts separately, and if you want to model something, model individual users. And for what it's worth, this post isn't typical of the kind of post I want to write. Would your reaction have been substantially different if this had been posted in Discussion?
I want to upvote your comment, but I can't bring myself to do so without hearing more about your reasoning. Without it, your comment reads like a mere personal attack.
I upvoted, but only because I thought the same thing and assumed adamisom's reasoning was the same as mine.

This is a persuasive argument that thinking like a superhero is bad. I notice that you haven't said anything about how supervillains think, much less argued that we should emulate them. Consider changing the title, maybe? The current one is eye-catching but inaccurate.

Good point. I could change the title or I could add another paragraph. I'm not sure what standard procedure is for editing Main posts, though. Is this frowned upon without a clear delineation of where the edit is? (I usually do this for comments, but in this case it might interrupt the flow of the post.) Edit: Paragraph added.
I don't know about standard procedure, but editing in another paragraph sounds like a good idea. If you're worried about interrupting the flow, you could put the note at the end of the post rather than at the point where you insert the edit.

There's a pen & paper roleplaying game being made about this.

Just the opposite, in fact: Better Angels is about being deliberately ineffective.

For example, on Facebook, Eliezer recently mentioned a thread where people were posting examples of things that they valued at a billion dollars or more, such as their cats. With a supervillain module running in the background, I noticed and pointed out that this constituted a thread where people publicly described how they could be ransomed. I can't exactly test this, but I don't think this kind of idea would have occurred to me before I installed the supervillain module.

I don't generally think of any of my thought processes in terms of supervillainy, ... (read more)

Large part of this comes from the fact that superheros rely on their special power whilst super-villains have to make do with more conventional means (intelligence), so this is largely fictional evidence.

In the real world I would not bet on Hitler/Stalin/Mao against Churchill, given the same resources. But to your point part of Churchill's power was understanding people (at least certain kind of people) - he called Hitler's end-game very early and against the prevailing political opinion at the time.

Amusing given your username. I might add that being reactive means that you won't go haring off in suboptimal or completely crazy directions because of the tyrrany of identity.

A supervillian who is trying to help is an anti-hero or anti-villain. That's not what I normally think of when I think of a "supervillain".

And if people are going to turn against you once you start actually optimizing, you might as well just be a supervillain to begin with.

Why? You might be able to build a better power base by acting as a hero.

I think you'd build the best power base by acting as a capitalist. Superpowers work for one thing, but you can buy anything with money. That includes other superpowers, so long as you're not the only on... (read more)

Lex Luthor occasionally averts this. Sure, he's generally the one piloting the giant robot, but at least he has snipers take out his enemies.
I'm pretty sure they all avert it occasionally. They just don't avert it nearly often enough.
Can't argue with you there ;-)

This is a really good post and (not but) it was smart to put it in Discussion. That said I think "thinking like a supervillain" also limits what you can and can't think and thinking like neither is likely to be more optimal.


But that's not being a supervillain, is it? That's being an ANTI-HERO.

I get what the article is saying (don't let your thinking become lazy by not going where conventional morality doesn't allow, even if that's what puts other people at ease and makes you accepted), but you're not advocating becoming a supervillain - abandoning all morality for selfish gain.

What you are saying that being an anti hero is better than a regular hero, ie. it is better to face possible orchestration for the sake of doing the truly right things and getting results.

What your say... (read more)

Maybe, but "supervillain" sounds better in the title and is more fun to say.
It might be more fun to say, but "I think like a supervillian" seems like a particularly dangerous cached thought to have, in that it will inspire overconfidence if you /ever actually encounter a supervillian/. A very smart agent whose goals are in direct conflict with yours AND who has no scruples AND hasn't forsworn the dark arts is /not someone you think like/.

The most limiting thing that you have not pointed out is that as a Superhero, you want to save the world. Saving the world [from supervillains] is by definition reactive. A Supervillain's goals have much more room for variation, and one could argue that Supervillains actually are optimizing the world, it just happens to be sub-optimal for everyone else.

It depends on what you mean by "reactive", I suppose. For example, if you as a superhero dedicate years of your life to reducing hunger in the world, then technically you are reacting to the hunger that exists, but still, this is much more similar to "optimizing the world" than to "stopping Lex Luthor".
You are correct. I was interpreting "saving the world" in this article to mean "saving the world [from supervillains]". (fixed in comment now)

But that's not being a supervillain, is it? That's being an ANTI-HERO.

I get what the article is saying (don't let your thinking become lazy by not going where conventional morality doesn't allow, even if that's what puts other people at ease and makes you accepted), but you're not advocating becoming a supervillain - abandoning all morality for selfish gain.

What you are saying that being an anti hero is better than a regular hero, ie. it is better to face possible orchestration for the sake of doing the truly right things and getting results.

What your say... (read more)

If you have censors in your mind that are preventing you from looking at parts of this space because some of your moral intuitions don't like them ("that's not the kind of thing a superhero would do!"), you're crippling your ability to search for solutions to problems.

Reminded me of this.

Some of the more sophisticated supervillains do have noble goals, but suffer from nothing more than an imbalance of coefficients in their utility function. For example, a non-supervillain who wants to prevent nuclear war might undertake public awareness campaigns, invent anti-ICBM defence technologies, run for office in order to implement disarmament policies, etc. A supervillain would enslave humanity instead, then rule with a titanium fist, ruthlessly eradicating any humans whose actions could conceivably increase the probability of nuclear war.

EDIT: the... (read more)

Discussing this example in public constitutes failing at supervillainy forever.
Unfortunately, this is a recurring theme on LessWrong.
My point is that committing yourself to thinking "like a supervillain" is almost as inefficient as committing yourself to thinking "like a superhero". It's better to think like "a well-informed agent who makes optimal decisions".
Who do you think of when you think of well-informed agents who make optimal decisions?
Is this why people cheer for the thieves in heist movies?
As jooyous points out in a sibling comment, thieves in (some) heist movies come to mind. They don't steal things because that's what thieves are supposed to do, or because they set out to do the sneakiest thing possible, or whatever. Instead, they steal because they have done the research, and they believe that theft is the best way to acquire the wealth that they need in order to accomplish some other task (even after accounting for risk). They could of course be wrong about that, and, in movies, they often are; but they still make the best decision possible given the information available to them. Other people either already possess the wealth, or can acquire it through other means; if these heist movie protagonists were all CEOs of Stark Industries, they wouldn't steal. They are thinking like optimizing agents, as opposed to thinking like thieves. Another example is scientists and engineers like Sebastian Thrun (just to name a recent example). His goal is to reduce automotive fatalities. If he was thinking like a villain, he could've set out to ban cars, or shoot drunk drivers in the head, or whatever; instead, he created the self-driving car.
How well do you know Sebastian Thrun? Can you pretend to be him easily enough for your brain to start generating different kinds of ideas than it normally generates?
I don't know him personally at all, but the bit about his reasons for creating a self-driving car comes from one of the interviews he gave -- I'm not putting words in his mouth, or anything. He did say something to the extent of, "I wanted to save lives, so I embarked on this project". Granted, he did not say, "I thought about shooting drunk drivers in the head, but, on reflection, this other way is more efficient". My point, though, is that (IMO) inventing a self-driving car is a very superhero-ish thing to do; in fact, Tony Stark does things like that every day (which is easy for him, what with being fictional and all). In this case, it happens to be an efficient solution to the problem, despite the fact that it aligns quite well with social mores.
Sorry, that comment didn't seem to have the desired effect. There is a paragraph that does not exist in the OP that I am pretending exists, and I should probably fix that. (Edit: Fixed.) My point is that the technique I'm suggesting is for you to pretend to be supervillains you like in order to come up with interesting ideas. You probably know the supervillains you like better than you know Sebastian Thrun, so I expect using supervillains to be more effective at generating interesting ideas than using Sebastian Thrun.
Yes, that sounds like a good idea :-) I think I'll need to read that missing paragraph before I can properly respond. I like quite a few supervillains, but I don't necessarily want to think like them. For example, I like Lex Luthor as he is presented in DCUO, but his thinking is demonstrably flawed and inefficient, due to some glaring mental biases and the aforementioned secrecy-induced positive feedback loops. I like him, but I want to think better than he does. But it is quite likely that you were thinking of something else when you typed your comment...
I like MoR!Quirrell. (His superpower is his intelligence.)
I don't know enough about him yet to like or dislike him. His actions seem like they're smart and efficient, but it's hard to tell whether they truly are, since we don't know much about his true goals. Due to all the mystery about him, he feels less like a character and more like a plot device -- though I fully expect that to change in the chapters to come. That said, intelligence is not MoR!Quirrel's only superpower. He is also an incredibly powerful wizard (at least, on his good days) in terms of sheer energy output (metaphorically speaking).
Nobody. I'm not going to intentionally limit myself to imitating people that actually exist. And that I've heard of, no less. I'm definitely not going to imitate fictional characters who bias their decision to whatever makes the story most interesting.
It's not as if you can choose not to imitate anybody. By default, you're imitating yourself. I've found imitating people other than myself to be a useful technique for generating ideas I wouldn't normally have thought of (which maybe I should have explicitly mentioned in the OP). What are your experiences with this technique?
By definition, either you trivially are imitating yourself or it is impossible to imitate yourself (if being yourself is mutually exclusive with imitating yourself). If you imitate others, then by imitating others you are doing exactly what you do.
Encouraging others to be villains is a classic supervillain move.
From this I take it that "let's shoot all these scientists in the head" is, by your reckoning, the most optimal solution under this scenario ? I personally don't believe so, but I could, of course, be wrong. Also, I realize full well that if you are truly committed to thinking like a supervillain, you will attempt to deceive me in your reply (should you choose to reply at all). But this exposes another problem with supervillainous thinking: it sets up a positive feedback loop in your own mind, which is rather difficult to break out of. This is part of the reason why Lex Luthor keeps trying to kill Superman instead of doing something (indeed, anything) more productive.
What he said sounds a lot more like "whether or not you plan to implement that idea, you shouldn't discuss it in public" than "that is a good idea that should be implemented, but we shouldn't discuss it in public."
That's a good point; I agree that I interpreted Qiaochu_Yuan's words somewhat uncharitably. That wasn't my intention. On the other hand, as per my comment above, one problem with thinking like a proper supervillain is that you can't share your plans with anyone -- as you have pointed out. This makes it a lot more likely (though, of course, by no means certain) that you'd end up falling prey to your own mental biases, thus embarking on some highly inefficient and/or disastrous path by mistake.
Supervillainy seems perfectly compatible with having some plans that you can share on the internet where everyone can see them, and some plans that you can share with a few other particular people but not everyone, as well as plans that you can share with no one. Your point about the last category of plans is a very good one.
-1Eliezer Yudkowsky

I rewrote my comment to more closely comply with the policy; if you still feel it is in violation, go ahead and delete it (not that I need to tell you that, of course).

This said, I disagree that my original comment listed any "identifiable targets". The comment policy brings up "people with purple eye color" as an example of potentially acceptable fictional targets; IMO, the targets I listed are no more specific than that.

1Eliezer Yudkowsky
Acceptable edit. Undeleted.
Have you by any chance read Watchmen?
Yes, but the character you're thinking of acts more like a classical supervillain. He is a good example of things working out well with that sort of mentality... although we don't really know how things turn out in the end, so maybe he's not a perfect example.