Often, enemies really are innately evil.

Scott Alexander, you totally fell down as a rationalist when you saw one of the biggest effect sizes ever, in a study that controlled for so many things

Controlled for so many things = A lot of potential for P-hacking 

A single P-hacked correlational study is generally not strong evidence and doctors are trained to require more evidence for the medical actions they take.

This poll asked people if they did "malicious online activity directed at somebody they didn't know", and 28% said that they did.

If someone answers "yes" to that question that tells us little about whether the person benefited from engaging in the malicious activity. It also doesn't tell us whether the actions are prosocial. A person who gets people on facebook banned by reporting their posts does harm to other people but they might think that it's good to do so because it removes people from facebook who are harmful.

I don't think you have done a good job at defining evil. I would expect that a good portion of evil behavior in the sense that it does damage to another person while producing no personal gain is due to heuristics build for determining group status.

If Bob is together with Dave, there's a question of whether Bob or Dave is higher in social rank. Engaging in an action that does damage to Dave without Dave being able to do anything about it is a way for Bob to assert that he has higher social rank. 

Evolution trained some heuristics into us to engage in such status moves. 

I have found myself telling a mean joke about a friend which might have made the friend uncomfortable while providing no value to myself. I have the self awareness to flag the behavior as an error but most people don't in a similar situation.

A mean joke that's doesn't do more then a few dozens of seconds of discomfort isn't a big evil but it's nevertheless evil in the above sense. I would expect that most bullying arises out of instinct that exist for making moves to assert status in ancestral enviroment. 

Bullying in schools often happens in enviroments with strong status competition and is likely a result of status dynamics (with the caveat that it's following heuristics that produce status in the ancestral enviroment and not all of those necessarily actually produce status in the enviroment of the school). When a group bullies together toxoplasma is involved where it's a strong signal of group loyality to bully and defection to the group when speaking up in favor of the victim.

I don't believe a significant percentage of people is innately evil, and at the end of part I of this post, I don't think you've given me significant evidence to chance my mind. The study is not convincing, and not because of effect size -- people could have misunderstood the game or just pressed the red button for fun since we're talking about cents. I would have predicted few people to press the red button if the payouts were significant (thinking at least 100$ difference); I genuinely don't know what I would have predicted for the game as-is.

there are many, many, MANY more pieces of evidence from (almost) every internet troll, bully, and rapist, and many other criminals too.

I mean, rape has a pretty obvious advantage for the rapist. "Troll" is so overloaded that I think you'd have to define it before I can consider it seriously for anything. Bullying is the most convincing case, but my model of bullies, especially if they're young, isn't that they're innately evil. If I remember correctly, I have participated in bullying a couple of times before thinking about it and deciding that it's morally indefensible. I imagine most bullies are similar except that they skipped the part where they think about it, or that they have thought about it, maybe decided to stop, but then proceeded anyway because the instinct was too strong.

Anyway, this is quite speculative, but my point is that I don't think you're making a strong case for your leading claim. I realize that this may come across as nitpicking details in a mountain of obvious evidence, but that's often just how it feels like if someone doubts what you consider an obvious truth.

There's also an issue that we may have different ideas of what 'innately evil' means.

"Sadism exists and is popular" is something I think of as a major blind spot for mistake/error theorists.

We need to taboo [? · GW] "evil", and split this confusing mental bucket [LW · GW] into multiple less confusing ones.

everyone partly fulfils the definition of psychopath

What exactly does this mean? Psychopaths are incapable of certain emotions (such as compassion), which allows them to develop certain skills (such as lying and manipulating people) to a degree almost unimaginable for an average person (who would be aware, and hampered by, consequences of the lies and manipulation on other people; and would therefore every day miss thousand opportunities to practice). Are normies also capable of lying and manipulation? Sure they are. Does this make them partial psychopaths? Nope.

People commonly have a terminal value of dragging other people down.

I have a problem with the word "terminal", because it is easy to imagine instrumental reasons (and difficult to verify whether they actually exist outside of my imagination), such as intuitively perceiving the world as a zero-sum game, and trying to increase one's relative status by dragging other people down. In an ancient tribe, sometimes you only need to drag down about dozen people to get on the top.

On the other hand, what is instrumental from evolutionary perspective, may still be terminal for a human mind [LW · GW]. Humans, as a species, are evil to certain degree. (Also good. Also inconsistent.) Another question is whether some groups of humans are more evil than others. Seems plausible; if we have people who are taller than others, or smarter than others, why not also people who are more innately evil than others? That is, they have a stronger version of the universal instinct for evil. (This is different from psychopathy, because the natural assumption would be that this kind of innate evil exists on the bell curve. Psychopathy is bimodal. The person with strong version of universal evil would probably enjoy hurting people in some ways a lot, and yet feel genuine regret for accidentally hurting them in a different way. On the contrary, an ethical psychopath would not give a fuck about human suffering, but might refrain from hurting people for some non-emotional reason.)

People often turn off empathy for an outgroup. I don't even know how to correctly classify this, because the instinct is obviously innate, but the definition of outgroup is learned.

Now, having said this all, I believe this article points towards an important thing. Yes, demonizing one's outgroup is classic propaganda / cognitive bias. But we should not revert this stupidity [LW · GW] blindly. Sometimes people are different. And if we assume that people differ on the scale of innate evil, you may happen to be more good than the average... in which case you are likely to meet people or groups of people who are more innately evil than you. (Okay, but really what is the chance that Less Wrong just happens to be the place where good people gather? Dunno, it's a website about saving the world, maybe this resonates with some. Another possible hypothesis is that Less Wrong is for nerds, nerds are more likely to be on the autistic spectrum, and maybe autists are less evil... because evil, in some sense, is a social skill? No idea, just made this up.)

It seems to me this article would gain a lot of quality by refocusing around bullying. Bullying's definition is more clear-cut and uninamously accepted than evil and good definitions are. I find the title especially misleading, since nothing in the article applies more to anyone's enemies rather than their allies, or themselves.

When talking about school bullying, my pet theory is that it has the same psychological causes as prison bullying. There's a form of violence that arises from being constrained for too long, like when battery chickens end up pecking each other to the death.

People commonly have a terminal value of dragging other people down.

I'm doubtful this is true, because it's easy to see that dragging people down is not a terminal value, since there must be something gained by dragging others down, and getting that thing is the thing terminally valued. For example, maybe dragging people down makes someone feel good in some way, then getting that good feeling is the thing valued, not the dragging people down. Dragging people down is a strategy, not something terminal.

This suggests the core thesis is mistaken: it's not that people are innately evil, it's that they've learned a bad strategy to get what they want and are trapped in a local maximum where that strategy keeps working and other strategies are locally worse even if they would be globally better if they took the time to reorient themselves to those alternative strategies.

I'm not sure this has much of an impact on some parts of this post, like the bits about bullying, but it does call into question many of the inferences you try to draw about people in general.

To me, 'evil' means 'should be destroyed if possible'. Therefore I don't like to hand out the label recklessly, as it leads generally to impotent rage, which is harmful to me.

People would often reduce their own prize if it means that their opponent's is reduced more.

This tells me we care more about relative status than absolute. See: anyone saying anything remotely critical of capitalism in the 21st century in the United States.

This poll asked people if they did "malicious online activity directed at somebody they didn't know"

You mean the default way to gain status on Twitter?

But yes, pure cruelty does exist. What of the fact that chimpanzees are cruel but have no concept of evil? This tells me maybe cruelty serves a self-interested purpose in dominance-based status hierarchies. If the human bullies don't know that, it wouldn't be the first evolved behavior that humans do without fully knowing why.

Whenever I try to analyze evil, I find banality all the way down.

Evil is when you "know that some things are damaging to someone else, gain no tangible value from doing them (or even expect that their life would be worse off!), know it is not a virtuous act, and do the harmful acts anyway without expecting future good to come from it."

The first failure point is "gain no tangible value." Imagine any prototypically evil character, maybe a person who is bullying once, maybe a chronic bully, maybe Dr. Evil, maybe Satan. Each of these gains some subjective value from their actions, if not "tangible" value. Either "tangible" is critical here, in which case you have way too narrow a definition of value, or it's not, in which case it is clear that these people are selfish and pretty legible.

What makes them evil is that their value system is so out of whack that they are evil (please just live with the circularity for now, I'm not trying to propose that as a formal definition). So the person who is bullying once and then learning it doesn't fulfill them that much...they may have done a bad, or even evil, thing, but they aren't evil! Same of the chronic bully - if they had a bad home life and are coping poorly, their innate value system may still be programmable to avoid evil acts. Dr. Evil is much closer to chronic evil right up until Goldmember (I can't believe I'm really going with this), when we find out he is a victim of circumstance, which anyone seemingly can be with enough compassion. Satan, well yeah, he's evil.

No one likes or endorses bullying, but you need a definition of evil that has validity, and yours is debatable. But even if accepted, it hardly encompasses a lot of people. You could actually stand to loosen the definition of evil, but you quickly run into selfishness. Construct definition is step one here. And it'll probably carry value judgments (see: virtue).

I am extremely dubious of the claim presented in part I, and in particular am dubious about the validity of the study cited.

Wait what?

What planet are these psychologists from where if you walk away from a bully, they suddenly become stuck in place and give up?

I think the "walk away" thing works better, broadening the situation to not just school bullies, if you stretch the definition or I suppose steelman it to mean do actions where you can leave the bully behind and it's costly for the bully to follow unless they stalk or resort to means or risks that are too troublesome. 

e.g. Quit a workplace full of harassers or leave a club or an organization full of jerks. In some settings if enough victims "vote with their feet", the bully runs out of victims. Obviously totally unreasonable in settings like schools, prisons etc. 

Also moving away from situations with bullies is easier for adults or people with more agency or choice than kids.

But it doesn't work in settings where locally "walking away" doesn't put you out of the bullies reach and there is a low barrier for the bully to "walk back to you and resume bullying" -- moving to a different part of the playground, the bully will just follow. Changing office spaces, while a harasser still remains your co-worker and will still harass in regular encounter.

(Sorry Scott Alexander, you totally fell down as a rationalist when you saw one of the biggest effect sizes ever, in a study that controlled for so many things, and still thought "hmm, i doubt being tortured regularly for a decade has a long term bad effect.")

Controlling for things isn't a good way to go about researching the effects of this. Instead, you should ask, what factors lead to variation in who gets bullied?

Eh, I think this is no innate evil, but people optimizing their behavior around the incentive of getting dopamine jumps, when they show their superiority to people around them while acting as a bully, or self pleasure of knowing they managed to troll someone on the internet forums and thus are better than them.

You gave no examples of evilness but just things that some predisposed people do to get their rewards, because their prehistoric ancestors evolved that way over the course of millions of years. Real evilness would be to do things that give nothing in return, therefore would be irrational to do. There is no real evil in the world, because even a madman doing irrational subjective evil is still rational to himself.

Want to eliminate evil? Just give people some Soma.

Don't think this study is big enough to be representative? Neither do I. 

I created an economics.stackexchange question for whether that finding generalizes. If you are also interested in the question, go and upvote/star it. 

Consider a game theory study by J. Sayer Minas, Alvin Scodel, David Marlowe, and Harve Rawson. The study itself is behind a paywall

I think we are talking about Some descriptive aspects of two-person non-zero-sum games which is openly available at https://journals.sagepub.com/doi/pdf/10.1177/002200276000400204?casa_token=2Tu4lDl7sxoAAAAA:hZAPPiOyqZmfCINGNs8a-zu1O-1vA4oXu4JJ3MD1XTwywyNaeNuPs6AJyXIeToTNePHyczLcfBVsLQ

Coincidentally, I started re-reading Worm about a month ago, and am almost done.  I'd forgotten how LONG it is, but I'm pleasantly surprised that it holds up well to a re-read.

I don't understand the point of this post, though.  You seem to be making a claim that some people cause pain to others, intentionally and for no obvious gain.  That claim seems so amazingly obvious and uncontroversial that I suspect I'm missing the point.  What work is "innately" and "evil" doing in your analysis, and what difference does it make whether someone's pointlessly cruel or pointedly cruel because they see a benefit in it?

What should I do once I classify an act or person as "evil"? Apparently "wait it out" isn't something you recommend, but you don't offer alternatives.  Is it even something you have to address, or can you just wait it out on a longer timescale?  They probably won't stop just because you're stoic, but they'll stop regardless of anything you do in less than a hundred years, because they and you will be dead.  

Don't think this study is big enough to be representative?

How big is the study?

While I find much of the post not good evidence of a large share of adult people being mean, I did experience multiple times adults being arbitrarily mean in situations with no apparent reason at all - while I believe most people are innately somewhat good, I agree there is a significant share of them being evil to random strangers for really no apparent reason at all.

Different point, probably not the most crucial point here but: In these prisoners-dilemma/public-good behavioral games, I always wonder whether the results are affected by the fact that participants might, half-consciously or consciously, understand that it's zero-sum: no real economic value gets lost. It's merely a transfer between the leader of the experiment and the participants. No social value gets destroyed from defecting, only money gets transferred between different agents in response to my behavior in the game.

Inherent evil is rooted in 2 sources.

1 Palatability is main source of inherent evil.

A fish inside water breath outside suffocating , a man the opposite.

Favorite food vs hated. Nutritional vs poisonous or allergic.

2 competition.

The enemy is enemy because what they do.

That is why inherent evil is a matter of perspective.

The sadist perspective is that the victim's suffering is a good thing. If you like peanuts everyone (those who hate or allergic) ought to eat them. Jocker of mean jokes don't see them as mean just the sharp point of the funny. Cats see the reaction of the pray as what they are doing right when practicing hunting. Falling from what someone judge as incompetent is funny , both as a show of dominance and not having in mind the suffering.

Homogeneity and competing for the same side , can take care of inherent evil.

No bullying I or anyone else I know has experienced was that bad, but the point is, bullies can go far beyond name-calling or even hitting.

Selection bias much?