Good post; you should put it in Main. It’s nice to see more stuff here on some of the aspects of instrumental rationality from HPMoR. Similarly, the two linked posts on your blog are brilliant; I’d been vaguely intending to write about them myself at some point, but you might have saved me the work.
KnaveOfAllTrades
Confused as to usefulness of ‘consciousness’ as a concept
Overly convenient clusters, or: Beware sour grapes
Welcome to Less Wrong! (6th thread, July 2013)
Yes; apology is an underrated consequentialist tool among nerds.
Some of the social function of apology can be understood game theoretically: Apology explicitly disavows a past action, allowing the one to whom the apology was made to leverage that confession in future: If someone apologises for something then does it again, then response can escalate because we have evidence that they are doing it even knowing that it’s ‘wrong’. The person who apologised knows this, and often the implicit threat of escalation if they do the same thing checks their future behaviour. Therefore apology is (possibly among other things) a signal, where the cost to apologising is the greater susceptibility to escalation in future cases.
Apology falls into a class—along with other things such as forgiving misdeeds, forgetting misdeeds, retribution, punishing an agent against its will, compensation for misdeeds—of things that would make no sense among sufficiently advanced and cooperative rationalists. Some things in that class (e.g. forgiveness) might already have been transcended by LW, and others (e.g. apology) are probably not possible to transcend even on LW, because the knowledge of other participants (e.g. confidence of their cooperativeness) required to transcend apology is probably too high for an online community of this size.
I would guess that the Bay Area rationalist set and its associates—which as far as I can tell is by far the most advanced community in the world in terms of how consummately instrumental x-rationality is forged into their swords—apologizes way, way, way more than the average LW’er, just like they talk about/express their feelings way more than people on LW typically do, and win because they’re willing to confront that prospect of ‘being vulnerable’.
“Well,” said the boy. His eyes had not wavered from the Defense Professor’s. “I certainly regret hurting you, Professor. But I do not think the situation calls for me to submit to you. I never really did understand the concept of apology, still less as it applies to a situation like this; if you have my regrets, but not my submission, does that count as saying sorry?”
Again that cold, cold laugh, darker than the void between the stars.
“I wouldn’t know,” said the Defense Professor, “I, too, never understood the concept of apology. That ploy would be futile between us, it seems, with both of us knowing it for a lie. Let us speak no more of it, then. Debts will be settled between us in time.”
Two mistakes in thinking that my past self made a lot and others might also:
(1) Refusing to apologize if another party was ‘more wrong’. Even if you’re 99.9% right/innocent/blameless, you still have to make a choice between apologizing and not apologizing to the other person. If you refuse to apologize, things will probably get worse, because the other person thinks you’re more wrong than you think you are, and they will see you not apologizing as defecting. If you apologize in a smart way, you can give an apology (which shouldn’t make a difference but has the actual consequence where the other person is more probable to also apologise) without tying yourself down with too broad a commitment on your future behaviour, and without lying that you thought something was a mistake that wasn’t.
(2) Using the fact that, in the limit as rationality and cooperation become arbitrarily great, apology is meaningless, as a rationalization for not apologising, when in fact you just feel embarrassed/are generally untrained and therefore not fit enough to apologise, and you’re therefore avoiding the exertion of doing so.
I want to point out the difference between completely fake apologies for things one does not think were mistakes, and apologising for things that were mistakes even if the other person’s mistakes were much greater. The former is less often the smart thing to do, and the latter is a lot more often than one might think. Once you get fairly strong, you can sometimes even win free points by apologising in front of a big group of people for something that everyone but the other disputant think is completely outweighed by the other disputant’s actions.
E.g. ‘I’m sorry I used such an abrupt tone in asking you to desist from stealing my food; it probably put you on the defensive.’ If you really mean it (and you should, because you’re almost certainly not a perfect communicator and there were probably things you could have done better), then often onlookers will think you’re awesome and think the other person sucks for ‘making you’ apologise when you’d ‘done nothing wrong’. Sometimes even the other disputant will be so disarmed by your unwavering ‘politeness’ that they will realise the ridiculousness of the situation and realise that you’re being genuine and that they made a mistake, whereas when they thought you were a hostile opponent, it was much easier for them to rationalise that mistake.
Notice than in that example, your apology has not even constrained your future actions; everyone was so distracted by the ridiculousness of you apologising when you were innocent and the contrast it made between yourself and your opponent, that nobody will think to escalate against you in future the next time somebody steals your food.
That’s why it’s so important to know how to lose—so that you can win! Just like how the best things you could do to decrease your personal risk from fights are things like practising conflict defusion techniques, learning how to walk away from conflict, being less tempestuous, being situationally aware, or even just learning how to play dead/fake a seizure/panic attack, rather than something that just looks like winning, like practising flashy kicks.
Did the whole thing. Cheers to all involved. :)
Donated $50.
MIRI 2014 Summer Matching Challenge and one-off opportunity to donate *for free*
Thanks to Luke for his exceptional stewardship during his tenure! You’ll be awesome at GiveWell!
And Nate you’re amazing for taking a level and stepping up to the plate in such a short period of time. It always sounded to me like Luke’s shoes would be hard for a successor to fill, but seeing him hand over to you I mysteriously find that worry is distinctly absent! :)
I always enjoy your contributions and it makes me sad that you use the site less than you would because of a/the downvote stalker.* I’m interested in applying microeconomics to these sorts of domestic or between-friends situations in which it might conventionally be taboo to do so, and this type of bidding in particular is something I’ve wondered about, so thanks for posting this!
*(I am not sure how easy it would be for an admin to look at the database and figure out what to do once someone’s been identified as a downvote stalker, but I’m skeptical it would not be worthwhile at least trying to sort it out given how destructive it is, and at the very least I’d like to know why it seems nothing’s been done. (Not to assign blame, but to try to figure out how we can get faster response on such issues in future.) If there really is one or a small number of people behind most of the downvote stalking, then checking the downvotes against victims should take (<)<15 minutes. I’d put >80% probability on this turning up one or more stalkers even just doing a naive database query using the set of victims in this thread.)
I sometimes find that telling my Inner Lazy that it can decide—after I’ve done the first one—between whether to continue a series of tasks or to stop and be Lazy gets me to do the whole series of tasks. Despite having noticed explicitly that in practice this ‘decision delay strategy’ leads to the whole series getting done, it still works, and rather seems like tricking my Inner Lazy to transition into/hand the reins over to into my Inner Agent.
I am really very pleasantly surprised with how this comment tree turned out and these are useful warnings. The level of internal insight was higher than I would have expected even if our first two comments hadn’t been vaguely confrontational. Thank you!
Rationalist households: What can London learn from its predecessors?
Yes. If we change “We shouldn’t eat chickens because they are conscious” to “We shouldn’t eat chickens because they want to not be eaten,” then this becomes another example where, once we cashed out what was meant, the term ‘consciousness’ could be circumvented entirely and be replaced with a less philosophically murky concept. In this particular case, how clear the concept of ‘wanting’ (as relating to chickens) is, might be disputed, but it seems like clearly a lesser mystery or lack of clarity than the monolith of ‘consciousness’.
Hi, jackal_esq. As someone involved in criminal justice, you might find the following interesting, if you haven’t seen them already:
Evidence under Bayes theorem, Wikipedia
R v Adams, Wikipedia
Sally Clark, Wikipedia
Amanda Knox case, Less Wrong (followup post linked at bottom)
A formula for justice, Guardian
Bayesian analysis under threat in British courts, Less WrongAside from that, welcome to Less Wrong!
Anthropics doesn’t explain why the Cold War stayed Cold
Begin here and read up to part 5 inclusive. On the margin, getting a basic day-in, day-out wardrobe of nice well-fitting jeans/chinos (maybe chino or cargo shorts if you live in a hot place) and t-shirts is far more valuable when you start approaching fashion than hats. Hats are a flair that come after everything else in the outfit you’re wearing them with. Maybe you want to just spend a few hours one-off choosing a hat and don’t want to think about all the precursors. But that can actually make you backslide. If you look at their advice about hats, you’ll see that pork pies and fedoras are recommended, but it’s well-known how badly a fedora can backfire if you aren’t very careful.
(For example, I’m still in the ‘trying new t-shirts/shirts/jeans/chinos/shoes with an occasional jumper purchase’ phase after about a year, 18 months. Still haven’t even got to shorts. You might progress faster if you do shops more often or have a higher shopping budget. But suffice to say hats are a long way in.)
There is a known phenomenon of guys walking around with a fedora or brimmed hat or whatever with a poorly coordinated outfit, dirty clothes, odour, bad fit, etc. basically not having the basics down before going intermediate. In these cases you will lose points with a lot of people because they will cringe or think you’re trying to compensate. You may or may not have been engaging in similar thinking when making this thread, but watch out for that failure mode.
Supplementary reading and good to get a yay or nay before buying something, or to get recommendations within a type of garment: /r/malefashionadvice/
Fashionability and going for safety helmets/caps might be divergent strategies though. If you were purely optimizing the former, what I say above might be relevant. If the latter, just getting some Crasches and calling it a day might be enough.
This premise sounds interesting, but I feel like concrete examples would really help me be sure I understand
Yep, I find the world a much less confusing place since I learned capitals and location on map. I had (and to some extent still do have) a mental block on geography which was ameliorated by it.
Rundown of positive and negative results:
In a similar but lesser way, I found learning English counties (and to an even lesser extent, Scottish counties) made UK geography a bit less intimidating. I used this deck because it’s the only one on the Anki website I found that worked on my old-ass phone; it has a few howlers and throws some cities in there to fuck with you, but I learned to love it.
I suspect that learning the dates of monarchs and Prime Ministers (e.g. of England/UK) would have a similar benefit in contextualising and de-intimidating historical facts, but I never finished those decks and haven’t touched them in a while, so never reached the critical mass of knowledge that allowed me to have a good handle on periods of British history. I found it pretty difficult to (for example) keep track of six different Georges and map each to dates, so slow progress put me off. Let me know if you’re interested and want to set up a pact, e.g. ‘We’ll both do at least ten cards from each deck a day and report back to the other regularly’ or something. In fact that offer probably stands for any readers.
I installed some decks for learning definitions in areas of math that I didn’t know, but found memorising decontextualised definitions hard enough that I wasn’t motivated to do it, given everything else I was doing and Anki-ing at the time. I still think repeat exposure to definitions might be a useful developmental strategy for math that nobody seems to be using deliberately and systematically, but I’m not sure Anki is a right way to do it. Or if it is, that shooting so far ahead of my current knowledge was the best way to do it. Similarly a LaTeX deck I got having pretty much never used LaTeX and not practising it while learning the deck.
Canadian provinces/territories I have not yet found useful beyond feeling good for ticking off learning the deck, which was enough for me since I did them in a session or two.
Languages Spoken in Each Country of the World (I was trying to do not just country-->languages but country-->languages with proportions of population speaking the languages) was so difficult and unrewarding in the short term that I lost motivation extremely quickly (this was months ago). The mental association between ‘Berber’ and ‘North Africa’ has come up a surprising number of times, though. Most recently yesterday night.
Periodic table (symbol<--->name, name<-->number) took lots of time and hasn’t been very useful for me personally (I pretty much just learned it in preparation for a quiz). Learning just which elements are in which groups/sections of the Periodic table might be more useful and a lot quicker (since by far the main difficulty was name<--->number).
I am relatively often wanting for demographic and economic data, e.g. population of countries, population of major world cities, population of UK places, GDP’s. Ideally I’d not just do this for major places since I want to get a good intuitive sense of these figures for very large or major places on down to tiny places.
Similarly if one has a hobby horse it could be useful. Examples off the top of my head (not necessarily my hobby horse): Memorising the results from the LessWrong surveys. Memorising the results from the PhilPapers survey. Memorising data about resource costs of meat production vs. other food production. Memorising failed AGI timeline predictions. Etc.
I found starting to learn Booker Prize winners on Memrise has let me have a few ‘Ah, I recognise that name and literature seems less opaque to me, yay!’ moments, but there’s probably higher-priority decks for you to learn unless that’s more your area.
I’ve been ever-more-excitedly watching you post your training and head off to workshop over these past few months. I teared up a little when I got to that standalone sentence, “On Saturday I was invited to become a MIRI research associate,” because now I know your origin story, I understand how much that invitation must have meant to you.
I haven’t really felt qualified to comment on many of your other posts (sometimes the level of the material, sometimes feeling too shy to commend your efforts), so I shall say now:
Thank you. We’re rooting for you. Keep on saving the world!
‘It goes without saying’ that I’m hella looking forward to your next posts.
Salute