it seems your model is
no, this is not what i said. this is not at all what i said, and i actually said something that is very close to the opposite of that.
i don’t know who to communicate if i say something and you come to the conclusion that i believe the opposite of that, so i’m bowing out.
i tried to explain that in various ways and repeating myself and doesn’t sure it useful to do that again, and i get the feeling from that replay that you simply not understand my claim, despite me already made it in previous comment, and it doesn’t feel you engage with any of my claims, so here is one last attempt:
if someone who err 1 time out of 100 make the conclusion is time was off, and change his algorithm to one that does not tell his joke, instead of updating his probability to tell this joke slightly less in a way that make him still tell the joke, he is doing the wrong update.
this is not what you do when you program simple neuron network and update in with backpropagation. the mathematically right update is “i have slightly less pobability to commit the same error, but i would have behave the same way unless i get this input a lot of more times”.
so you advocate doing the mathematically-wrong thing. we should learn what we can, but what you advocate is literally over-correction.
so i obviously prefer to get better at things, and as part of normal life have more training data and become better at things.
and yet, what you describe still look to me like common failure mode. maybe i should just describe it instead, and what is the right reaction in my opinion.
person make jokes, 100 land one not. failure mode is to apologize, and to change the algorithm to avoid the bad joke, and tell only 10 jokes.
the right reaction is to correct, but not over-correct. and in real world trade offs, it mean predictably make mistakes, because there are no known algorithm that make no mistakes, and the only way to avoid one sort of mistake is to do ridiculously bad choices. avoiding doing things because only the one who do things can err.
and this is indeed failure mode i see a lot!
basically, i see a lot of succumbing to the most outraged people, and it’s no good at all.
so i come to respect the people who can acknowledge the imperfection of their trade off. to say “this is my informed trade off, and i stand by it, and it is right and good, even if not perfect”, as the alternative is to avoid telling jokes as to not be offensive, in a way that is net-negative and make the world worse.
so i prefer the chill guy, but the one i fear is the one that apologize and promise not to do this error again, and fulfill his promise, by telling way to little jokes. and i find that group that don’t have any defense against this tendency will tend to slide toward the most sanitized and boring and saftist place, and it’s crucial to have someone who have the courage to protect the unpopular and sound-bad thing.
it’s crucial to be able accept the bad side if the chosen trade off. refusing to do that doesn’t make trade offs go away, but just mean the we, as society, lose the ability to make informed choice, and a lot of time choose the bad side that is less visible but actually worse.
so i somewhat find the so-called chill guy less trustworthy. i expect that he will be easier to join cancellation mobs and will not be one to stand against them, for example.
(i’m trying to point to pretty vague thing, here, and doesn’t think i did a good job on that. hopefully one day i will write the planned post about trade-off denying and why it’s bad.)
i know there are people that live like that, but i’m not, and when i do i find that unhealthy. i prefer not to do things just before the deadline. i’m now in time when i write every day too, and it feel good to write early, and to know i did my writing for the day. i like the feeling of accomplishment, and all the day feel like i did something useful in it.
what you describe look to me something like… adaptation to dysfunction, in a way that deny the existence of health. it’s actually possible to do the work not on the deadline, and enjoy your free time after that and not before! and on the times when i did a lot of procrastination (when i was student) it was both sign of internal conflict, and was worse job then when i did things early.
i’m in favor of Half-assing it with everything you’ve got, but there are things that worth to do not in deadliny mode, and world when all there is is that is world when something important lacking in it, i think.
why you don’t find it realistic? i find it very realist. people have so much discretion ability, but not infinity of it. like various cancer and other medical tests—you got some false positives and false negatives, and you can get better, but in lot of cases not infinity better.
my live full of places when i have imperfect algorithm and have no idea how to update. for less adversarial example—i’m not always capable to discern when my friend’s cat want to play, aka, may scratch me.
i find the idea that people always know what they did right and any absence of omniscience weaponized ignorance really weird!
i, too, hold economic-based model of apology. apologizing mean to say “i owe you”. like if i break something in store, i acknowledge i need to pay for it. which make the whole post feel like it miss the point.
What would it even mean to purportedly accept the cost “on one’s ledger” but not change one’s behavior going forward?
well, that i owe you. for example, i apologize for spilling the wine of your outfit. can i pay the cleaning price, or the price to replace it? this is not, in any way, mean i will avoid drinking wine in the future, or be more careful. maybe it’s my endorsed trade off. but part of this trade off is paying the price of the wine-spilling.
When someone wrongs me, it seems like the smallest ask I could reasonably make in exchange for my forgiveness is that they not do that again
I have to admit I’m skeptical. If I can’t ask not to be hurt again, what can I ask for? Money? Chocolate? Their car?
well, yes, actually. asking me to pay for the ruined outfit is much smaller ask then to change behaviour pattern foe example, avoiding drinking wine. but the part of paying for the ruined outfit is critical in making good trade offs, ones the does not just externalize costs on others.
the idea person must change behavior of harm occur look to me like part of general pattern of inability to accept trade off and bad things, that sort of thing that lead to Fabricated Options.
it’s fine to bump into someone if i need to do that to save my cat, and also, it incur cost that i should internalize, to align the incentives. so i will bump, and own you metaphorical 5 dollars. and yes, you can think 5 dollars is to little and refuse to accept the apology. this is fine, it;s still part of the game.
but the idea that i can’t do something with some downside, and then offset the downside somehow, it’s just lead to people lying to themselves a lot. there are no ways to live your live with no harms done. you must chose the lesser one and internalize the cost. because attempting to never cause harm just lead to Fabricated Options.
if the joke land in 9 times out of 10, and Ben unable to discern when it lands and when not, it’s still positive in expectation to tell the joke, and he don’t know who to update the algorithm.
if it’s 99 out of 100 that land, then it’s obvious that the right thing to do is tell the jokes. and try to always learn something from this sound like the road to overfitting. i prefer people who accept that ridiculously good sucess rate of 99%, and apologize on the downside of the endorsed trade-off. it’s sure better then… stop doing things that good in expectation because they good only 99% of the time? i really not sure what is the algorithm you advocate here.
while i dislike the word naughtiness and its connotations, i don’t see the contradiction. as the post state int he start “They delight in breaking rules, but not rules that matter. ”. all Decision Theory rules are rules that matter.
this doesn’t answer the question of what rules to break and what not. but you cant be Lawful by following all laws in our way-to bureaucratic society.
my mother think that going outside while wearing pajamas is unthinkable. it can probably be described as naughty. it’s also totally harmless.
but also, now that i thought about you comment it become clear that i actually don’t know what is even the meaning of “Be Naughty”.
i think this is the kind of post that would benefit greatly from three examples at least.
did you run it? what did you found out?
the thing this post missing is the connection between the Process Crime and Actual Crime. the child have no reason to use the stepstool, beside to eat the cookies, in the vast majority of situations. is it true about lying to banks, or is in situation when Everyone guilty of some crime and the state get to lawfully punish whoever it want, by some criteria that are not written in the law?
this is empirical question with empirical answer, and when i saw disagreements about Process Crimes and Pedantic Rules in real life, it was of the form of “there are legitimate, non-criminal reasons to break this Process Rule” or “this Pedantic Rule is pretty costly to this set or rule-abiding citizens”.
interesting, i came to the opposite conclusion. or, more concretely, i found that there are rules that i can follow 95% of the time, and it’s actually better then follow it 100% of the time, and there are rules when i simply can’t, and i slip all the way down the slippery slope. and a lot of time i can search for differentiator, or more complicated solution.
simple rules need for coordination, when you have about five words. but i have much wider channel to coordinate with future-me, and i should use it to my advantage.
so my lesson was that having rules that are in the form of “never do X” or “always do Y” are almost always bad rules that lack granularity.
instead of “never go sleep after midnight” i can go for “don’t go sleep past midnight, except once a month you can read a book for as long as you want, if you commit to wake up on time at the morning, and go sleep early the next day”.
there are patterns when “breaking” the rule occasionally does not break the habit, and get me more utility (washing the dishes before going to sleep 95% of the time is better then 100% of the time - i get the utility in the rare cases when I’m honestly too tired, and i do not slip on this slope).
now, rules in the form of (1), that lack granularity, lack build-in exceptions, look to me surprisingly stupid. do you really can have only this simple rule? are you sure? did you tried?
you say ” humans tend to heavily biased towards believing that their future selves will make the decisions that they want it to make. ” but it look to me there are symmetrical and opposing biases -toward optimism and toward pessimism. and you just chose bias in the other direction. i prefer calibration.
and claims like (4) look to me like self-fulfilling prophesies. why do you believe that? my experience show that it’s mostly wrong. sometimes it is true, and then i get one of the hard line rules. but most of the time, more granularity is possible and sderiable. and having it, having better rules, is skill that is possible to develop. as is the ability to have exceptions without dismantling the rule.
And once you have a message you actually need to share, you’ll actually be able to express it.
that doesn’t look like good advice to me. or, rather, to make sense, one should assume, a priori, that the day will come when one will have “a message you actually need to share”. this way of thinking proving too much—why shouldn’t I read those trip reports, assuming one day i will need this knowledge, instead?
there are a lot of things one can do with one’s time, and most people don’t come to situations like that. learning to write better because one day i will want to write trip report is bad algorithm to use.
with science, the problem is not that people bad at writing, bad that the system have a standard of how to write, and it’s a bad standard. if good writer tried to have good writing in it’s science paper, they would have been told to rewrite it, to be more “scientific”, so learning to write wouldn’t help.
better advice will be to see what skills one predictably may need in the future, and practice those skills. but, claim, this will not include “writing” for most people.
there is a high correlation between what voters as a whole want on any single narrow issue, and what the outcome of the black box produces.
I’m way to late to the party, but I’m still reading old posts, and this is evidence that other people may too.
this statement is not true, in the country that I live. there are some narrow issues in which there is large public majority, and the black box produce not-correlated output. and this happen for a long period of time.
For the rest of this comment, I’m going to restrict “fox” to “good-scoring generalist forecaster”
why? this is very much not how I understand it, and I find it hard to even generate hypothesis to how you came to this conclusion. which make the rest of the comment seem irrelevant and missing the point.
foxes are people who go to people that have some big theory, and telling them they are wrong, everyone who try to have one theory is wrong, they should have many theories in their toolbox and use the more appropriate.
basically I understood it as Toolbox-thinking and Law-thinking.
restricting foxes to good-scoring generalist forecaster is even more weird. like… why? bad scoring are not foxes? it’s look to me like trying to rig the game.
I opened a tab with this comment as something I want to answer too, when an opportunity arise for that. it was two years now, and the tab remain open. I wrote two post about that in my blog, and clarified my thinking about the issue a lot. and yet, I still believe what I believed then, although hopefully I’m more capable to express this. not sure if that, English is hard. but, it’s worth trying.
here is my opinion, in its most simple form: one should not be cooperation bot.
if I have repeated opportunities to pay 1 unit to get someone 10 units, and they have the same option, it’s better for us both to make those trades. but if the other people refuse to do that, they take happily the provided 10 units and then not give up 1 unit to give me 1, I should stop.
that is the Morality as “Coordination”, vs “Do-Gooding” distinction.
and this is what I remember I want to write to you, 2 years ago, and didn’t succeeded on the first try. you call for Civic Duty look to me as call to cooperate even if the other side defect, call to create cooperation-cooperation equilibrium, while ignoring that other people not cooperating.
there is no such thing as acceptable exchange rate, that is the wrong category. there is an option to enter agreement to cooperate, to exchange 1 unit of mine for 10 or 100 or 1000000 of yours if the opportunity arise. and the right thing to do is tit-for-tat with forgiveness. and it’s definitely not being cooperation bot. and deciding that if the exchange rate is high enough you must take it, it’s you civic duty, is just calling people to be cooperation bots, and punish those who don’t do that, while ignoring the important difference between those who enter the agreement and those who don’t.
***
I don’t really engage with this comment. I don’t think it react to the concept I tried, and failed, to communicate, so I tried again.
but I can say that from my current ontology, when I see Morality-as-Coordination and Morality-as-Do-Gooding as two very different things, the claims about local governments look confused to me. there is “charity” that is coordination, and there is charity that is do-gooding, and those are different things.
most of that governments do is not do-gooding, it’s coordination, with elements on insurance.
small and very belated comment to comment (6): I did, actually believe in things and do things because of the decision theory. I sometimes had the impulse myself, but considered it childish and stupid and basically irrational, as in the past i saw people implement “spite” in really bad way.
then I encountered the decision theory on that, and evidence that people react to incentives more then I thought (I’m not sure if it’s a crux—I endorse doing the correct thing even when the relevant people don’t react to incentives).
so I changed my mind. I endorse the feeling you describe when it arise, but i still naturally pretty low on it. I still behave accordingly, though.
It’s not exactly supportive evidence for your claim, but it feel to me like it close enough, so I wanted to write here—that it’s possible and desirable to change one’s behaviour to do the right decision-theoretical thing. especially in the direction of not negotiating with terrorists.
so I did actually change my behaviour.
One of the best forums I knew was homeschooling place, that was officially anti-vax. it was really bad at true-seeking, but I learned there so much. it turned out there are a lot of things that doesn’t get said in places that are no default-collaborative. my models of people and societies much better because I was there. and also I leaved when I couldn’t endure the idea that I’m bad and wrong for pointing out obvious falsehoods.
I don’t think you can seek true without pointing out at false thing and saying they are false. but also, this way of thinking tend to be blind to all of the things that can exist only in places where there is good faith assumption. there is a lot of sharing of information, of ability to say or write socially forbidden truths.
and I think most places are far far away from the Pareto front.
the way to collaboratively say “I don’t understand” include demonstrating Good Faith—trying to come up with two hypotheses, trying to understand in a way legible to the people reading it.
you can disagree without blocking. more effort needed to be invested, and it’s coordination problem, without norms it can lead to bad places when the less busy person win without regard to truth.
but when it works, it works better then the bad-faith-allowing place. disagreements with good faith are harder, but are more rational.
I’m thinking a lot about the way that my good faith heuristics, that I got from place that was atrocious at truth-seeking, is so good at promoting it. this is because it forbid very frequent failure mode of irrationality. people who decided the other side is enemy tend to interpret things not in the most correct way, but in the most strawman-y way. it’s very very bad, and it’s dynamic that prevent thinking and understanding.
the “it will be used i status based asymmetry” claim look to me like general counterargument to any norms at all. in the same time, it pessimistic in way tat look unjustified to me. there are rules that enforced more or less neutrally, and i see no reason to assume the worst here. LessWrong is not actually the sort of places when the popular kids never get called out.
while the emoticons suggestion is interesting, it’s missing the point of building mental habits. like asking myself if I’m dreaming is practice for lucid dreaming. having other people do that loss that benefit. it’s also loss the saving throw of avoiding escalating by apologizing,
I actually disagree, it may be true about the other parts of the prefix, but reading something that is tagged as uncharitable rant would make me less inclined to get mad and react uncharitably. the point of the tagging things as rants is to prevent the dynamic when emotions mindkill people and discussions, and saying something is unfair at the beginning produce different emotional reaction.
also, the last time i encountered that sort of prefix it was at part of long post, and was much less hard to parse, as i had context. and it prevented the need to going back to understand what was intended, thing that is cognitively costly, too.
the question is if something is worth it. reading long posts is more cognitively costly then watching tiktok videos. and yet, here we are. something being costly is not enough to avoid it.
here is my main counterarguments, and i will elaborate on them if asked:
The Light World allow to gather information that the Dark World disallow. it’s in high-trust environments that i gathered examples of personal experiences i didn’t know exist, and so learned more about the world, and have better Map of it.
people in the Dark World often fail to listen to other people. in the Autumn framework, they have filters that prevent them from see certain evidence.
i don’t think all distrust id bad, but it’s very common failure mode. and if you try to pass Large Inferential Distance, then there is certain kind of Dark Worldism that make it impossible, and certain kind of intellectual respect that needed for it to work.
my opinion about the need of Good Faith in discussions to not be almost completely futile didn’t come from theory, but from Empiricism. in my observations about the quality of discussion in different environments.
there is sort of mental move, especially stark in political contexts, when i can see how some person totally fail to understand the thing someone said, and assume something more stupid, and it’s both not-truth-seeking and wrong, and low trust.
i learned some interesting things from reading people i respect and disagree with, and i simply wouldn’t invest the time needed to cross the Inferential Distance if i didn’t have high prior on them telling something meaningful.
people VERY OFTEN underestimate the Inferential Distance, and in low trust place it mean people attack ideas they fail to understand, and miss the chance to understand them.
in my opinion, most people fail at the Autumn-update on the part when they fail to listen to the claim or evidence, as most things are complicated, reality have surprising amount of details, and they don’t wait enough, don’t invest enough time and energy modeling the other persons, opinion (of they try at all to model it). they fail at the “exchange evidence” part.
(there is a symmetrical error when people assume someone said something more truth then they did, and sanewash it.)