If you managed to read the comment I posted and removed yesterday, I'm sorry. I shouldn't post at all in the evening.

I have experience with depression both personally and professionally, so you don't have to explain to me what it is. This doesn't mean I know the optimal way to handle it, or that all forms of it can be handled the same. If you have a strong bias against antidepressants, which is quite common, you should acknowledge that before reading further, not because I'm going to recommend them to you, but because reading those linked comments might cause a negative halo effect on me and the rest of this reply.

As learning agents, our algorithms for dealing with the present are necessarily path-dependent. If my path through experience-space has shown me that most social interactions were negative-sum games at some point in the past, and that repeated attempts to behave as if they might NOT be negative-sum games result in losing, and losing badly, then it might not be worth the perceived risk to take a chance on new people, unless those new people go to extraordinary efforts to demonstrate that they aren't playing a negative-sum game.

I wasn't communicating clearly, sorry about that. I didn't mean you shouldn't calibrate your expectations according to your experience. I meant you shouldn't necessarily calibrate how you treat people according to your experience. Expecting the worst from people, and telling people you're expecting the worst from them are two entirely different things. The latter is going to make it more probable that people treat you badly, whether the probability is low or high to begin with. Worse than that, it's going to make you miscalibrate your expectations.

Now, posit that in the past, people have gone to extraordinary efforts to demonstrate that they weren't playing a negative-sum game with me, only to turn around and spring elaborate traps, because they thought it was hilarious and worth the cost of the effort just to trip me up. Now what are my expectations primed to? What should I rationally expect from the world, given those priors?

I see. I've had a few such experiences too, no doubt damaging. What do you think about treating your expectations and your emotional investment in people as separate things? Acting like you trust people on default doesn't necessarily mean you need to get emotionally invested in them, but it will almost certainly make them treat you better. Take this interaction for example. I'm not expecting you to get emotionally invested in me, but I'm expecting you're not constantly acting like I'm attacking you. In fact, you said it quite well yourself:

my usual pattern of assumption is NOT that people in general are evil assholes; it's that I'm caught in a loop of behaviors that provokes them into questioning my veracity, I overreact to their questioning, and they become primed to act assholeish towards me, thus reinforcing the pattern

--

I'm not sitting here accusing you of malevolent intent just because I'm a depressive curmudgeon. I'm also attempting to use this as an explanation for why people become depressive curmudgeons, and describe actual steps that I believe could be taken to break the cycle.

I'm quite certain people become depressive curmudgeons in various different ways, to various different degrees, and benefit from different kinds of treatment approaches. Don't generalize from one example, or think your mind is typical for all depressive people.

When a system gets into a feedback loop, you don't tell it that it's being a bad system and it should feel bad; you change its inputs so the loop can be broken. If hundreds of other people are telling it that it's being a bad system and it should feel bad, and those inputs are strengthening the loop, then if you want the loop to break you have even more work to do. Or you can acknowledge that the loop isn't worth the effort of breaking.

When someone corrects you, it hurts. When you're depressed it hurts more. I think you're instinctively jumping from this to the false conclusion that when people correct you, they want to hurt you. You're not a bad system and you should not feel bad. Those are not my reasons for correcting you. LessWrong is all about people correcting each other and improving that way. People are biased in general ways and depressed people are biased in more specific ways.

If I hadn't talked about splitting but confirmation bias for example, would you have taken it as badly? Do you think people shouldn't correct you in general, or that they shouldn't correct you when your behaviour is clearly caused by depression? Does pointing out that you have confirmation bias cause you to feel that you're a bad system?

Keeping in mind this is LessWrong: If you don't want people correcting you about something, don't bring it up.

Drilling down a level, you're having trouble acknowledging that the loop isn't worth the effort you would need to expend to break it, because YOU believe that that would make you an "evil asshole". I made no such value judgment.

You implied that people who correct you do so to get an excuse to call you a parasite, a bad system, whatever. If I really believed I did that, then yes I would believe I'm an evil asshole. It's true you didn't make that value judgement, but you're incorrect about why I made it.

In fact, I have complete empathy for people who realize that the effort that it would take to fix people with my level of psychological problems isn't worth what they'd get out of it.

I have no illusions about fixing you. If I nudge you in the right direction, great, if not, at least I've learned something about people.

But because YOU continue to believe that it would be evil for you to stop trying to help

I don't believe that. I'm mostly an egoist, although I do have altruistic tendencies like most people.

you continue to perform weak half-measures that only serve to agitate the problematic mind-states further, and then turn it around to being my fault when you do so.

Some people take advice, some don't. This applies to depressed people too. For people who have depressive tendencies, we're both highly atypical, in the sense that we're interested in X-rationality. If I expected you to be a typical depressed person, I wouldn't talk to you in this particular manner, and would have indeed expected that correcting you is utterly useless to begin with.

Concerning your metaphor, not all depressed people are drowning, although some are. Some didn't know how to swim in the first place, some forgot how to swim in specific ways and talking to them without giving them practical swimming lessons could be sufficient for making them better swimmers in relatively calm waters.

This thread has become painstaking enough that it's time for me to eject.