Open Thread 3: Threads Are Like Parachutes

1. LW user Bakkot has written a comment highlighter for SSC that shows you which comments are new since your last visit (think the purple squares that sometimes appear around comments on LW). The installation looks pretty easy, but if anyone can figure out how to automate it by putting it on the blog itself, let me know and I’ll give you access.

2. I request that people not tell me things like “Why are you wasting your time writing about issue X? Aren’t there more important / less politicized things to talk about?” The failure mode of LW was that everyone with interesting but less than maximally virtuous and important things to say felt bad saying them in a place whose standards for discussion might be too high, and so went off somewhere with lower standards (lowering LW’s standards would also have a failure mode and not be a tradeoff-free solution; I can expand on this if it’s unclear). The potential failure mode of this blog is that I feel like my commenters’ standards for discussion are too high and so I go to Facebook or Twitter or Tumblr or instead. It’s my blog and I’ll write about what I want; if you think it’s a waste of time, don’t read it.

3. I briefly experimented with AdSense ads in the sidebar. I found that with 7000 average page views daily, I made…just under a dollar per day. This is less than I was led to believe, but probably involves readers here being web-savvy enough to tune out ads. The extra dollar doesn’t seem worth the annoyance to readers, especially if reader annoyability is a limited political-capital-style resource I can spend on (for example) posts with fifty terrible puns in them. I figure the things that might increase revenue – putting ads directly in posts, really big flashing ads everywhere – would also increase reader annoyance and aesthetic disruption, so that the overall tradeoff of annoyance for money is linear at an exchange rate I’m not willing to take.

4. Instead I’ve been trying the Amazon affiliates program. There’s a tab on the top of the blog explaining exactly how it works. There is also a link to Amazon on the sidebar. If you click on that link to make your Amazon purchases, then it will not cost you any more money, but a certain amount will go to me and help pay for site maintenance (plus hopefully some left over).

5. I continue to be terrible at answering emails and Facebook messages. The preferred way to get in contact with me about this blog (as opposed to if I know you in real life and we’re setting something up) is comments here.

6. Comment of the month is this analysis of subway building costs.


This entry was posted in Uncategorized and tagged . Bookmark the permalink.

280 Responses to Open Thread 3: Threads Are Like Parachutes

  1. Scott Alexander says:

    I have a problem I would like to crowdsource.

    A couple of weeks ago, I received a very important looking FedEx Priority Overnight envelope whose return address was Thomas Jarmyn, Veteran’s Review And Appeal Board, 1710D Woodridge Ln, Florissant MO.

    Inside was a check made out to me from the California College of the Arts for $900, and nothing else.

    California College of the Arts seems to be a college in Oakland. As far as I know I have never been associated in any way with the California College of the Arts and indeed never heard of them before.

    The Veteran’s Review and Appeal Board seems to be a Canadian military organization. Thomas Jarmyn is indeed a Canadian military officer who is in charge of it. As far as I know I am not a veteran of the Canadian military.

    Florissant, MO seems to be a suburb of St. Louis, Missouri. The particular address, as far as Google Maps knows, seems to be a private home. It has no obvious relationship to either of the two organizations above.

    There was a phone number on the return address form right below Thomas Jarmyn’s name. When I call it at any time of day or night, I get an answering machine message in what I assume is French.

    So – do I cash the $900 check or not? Everything about this screams ‘scam’, but I’m not sure how cashing a check would help someone get money from me.

    • Ialdabaoth says:


      Take it to your bank. Tell them, “I run a reasonably successful internet blog, so there’s a reasonable chance this could be an oddly anonymized donation. There’s also a reasonable chance this could be a scam. How do you recommend I proceed?”

      • James Miller says:

        If you do try to cash the check type up something saying “this might be fraudulent” keep a copy for yourself and give the other copy to the bank. This way if something bad happens you have a bit of protection.

        I personally wouldn’t take the risk and I would just rip up the check.

    • The Anonymouse says:

      Military officers have official email addresses. They are generally not hard to find, and if you can’t find it, tend to have a consistent format, so if you can find anyone’s, you can figure out the format his would be in.

      Email the official address, so you know you’re talking to him (more likely, at least). Ask him what’s up. An active military officer, presumably, would be less likely to be a scammer (professional consequences and all) than J. Random Dude.

      • The Anonymouse says:

        Also, IIRC off the top of my head, the ‘cash the check’ scam usually goes something like: give Scott a fake check, ask him to cash it for you and send the money, minus Scott’s cut, to scammer. Sometimes the bank cashes it, sometimes it doesn’t, but when something goes wrong, boom, it’s Scott who’s trying to pass the rubber checks.

    • Watercressed says:

      A similar situation was described at
      the consensus was that it’s probably a scam

      • Scott Alexander says:

        Thank you.

        I notice their check was also Fedexed, their check was also for a number just below a potential large-amount-that-makes-banks-check-for-fraud threshold (their $1970 versus my $900), their check also had a mailing address that didn’t match the name on the check, and their check also ** stars before the dollar amount. They also provided the missing information on how the scammer gets money (calls me after I deposit it). I am now totally convinced this is a scam, and possibly by the same scammer as the Redditor encountered. Thank you.

        • Deiseach says:

          I’m interested to know (if you’d care to say) why you considered cashing it in the first place?

          I mean, the scheme obviously runs on human nature – the appeal to greed (who is going to turn down free money?) – and nine hundred dollars (Canadian or no) is not a trivial sum, so most people would take the risk.

          But even setting aside the moral/ethical considerations (because it also involves us being willing to take money we know we don’t deserve from someone, putting them at a loss if it’s an honest mistake e.g a cheque was put into the wrong envelope, the address was wrong and they meant to sent it to a different “Scott Alexander”, etc. – stealing, in other words), why didn’t it trigger your suspicion enough to destroy it instantly, instead of hanging on to it in case it was legit – or rather, that you could cash it and keep the money without getting bitten in the backside later by the bank/law seeking to reclaim funds?

          I claim no superior virtue here; my only good trait in this is that I am somehow lacking in envy (or emulousness, if you want to call lit that) and one of my bad traits is stubbornness, so if someone tries to softsoap me into “But you NEED this cool new thing!” or armtwist me into “You came into our shop/showroom, you HAVE to buy this product”, I have no problem saying “No” and keeping on saying it.

          Which also means I’d go “Yes, nine hundred Canadian dollars comes to around six hundred euro and I certainly wouldn’t turn that down, but I don’t believe in the Free Money Fairy, I am not an adherent of the Prosperity Gospel, and if I bin this piece of paper I am no worse off afterwards than I was before the post came and I opened the letter”.

          It also means I can resist the blandishments of TIME magazine trying to get me to take out a subscription, with their heartstring-tugging appeal of “We have chosen only a few people in your country to enter our draw, and YOU ARE ONE OF THEM. Here are your nine numbers, subscribe to one year only of our magazine and we’ll enter you in a raffle for €€€€€€€!!!”

          Ahhhhh….very kind of you, I’m sure, but no thanks.

        • lmm says:

          Scott, unlike many of us, had reason he might genuinely be sent an anonymous-ish donation.

        • Kiboh says:


          Can’t speak for Scott, but in his shoes I totally would have held onto (but not cashed) the check, for the following reasons:

          .I’d only be *mostly* certain it was a scam or mistake. Both of my grandfathers were veterans, and there are stranger things for a charity to do than send random amounts of Canadian money to the relatives of old soldiers. Until I got outside advice, I’d be holding on to that possibility, just in case. Hope can be a wonderful thing.

          .A sufficiently bastardised version of the anthropic principle suggests that if you get a misaimed Canadian check in the mail, you’re probably not the only one. If it’s a scam, there are probably others being affected, and maybe I can do something cheap to help them. And if it’s a mistake, maybe I should try to get in touch with the relevant organisation so they can cancel all the *other* checks they probably sent to the wrong addresses.

          .I’d want to solve the mystery! And rule one of mysteries is that you don’t go burning evidence that might be useful later. Unless you’re the culprit. Or unless you’re pretending to be the culprit as part of a complicated gambit to unmask the real culprit.

    • Shmi Nux says:

      Did you talk/email to the California College of the Arts people, since they are ostensibly the ones who issued the check, to confirm ? At least some of them are bound to speak or possibly even read some English. If you estimate the odds of this being a scam at 10:1, $90 expected payout (and possibly some material for another blog post) may still warrant a few minutes of your time. At 100:1, not so much. I’m guessing that they have heard of this scam before. (My guess is that your odds are 1:1000 for this being real).

    • Shmi Nux says:

      duplicate deleted

    • Anonymous says:

      “There’s no such thing as a free lunch”.

      My own inclination would be to burn the cheque or tear it up and put in the rubbish. Secondly, stick it in an envelope and send it back to the return address.

      If it’s legit, the guy will write you out another and include an explanation. If it’s a scam, he won’t bother.

      I certainly would NOT try cashing it or lodging it to my bank account. You don’t know him, the address, the organisation allegedly involved and there’s no accompanying paperwork to explain why someone wants to give you nine hundred Canadian dollars?

      It sounds like a scam to me.

    • Eric Rall says:

      If it is a scam, this is how it would probably work:

      1. Scammer sends forged check to victim.

      2. Victim deposits the check. If it’s a certified check or cashier’s check, the bank will usually post the money to the victim’s account immediately on the assumption that it’s certain to clear.

      3. Scammer contacts victim, says the check was sent by mistake, so please send a refund. Sure, I’ll take a personal check, and I’ll even let you keep 10% for your trouble. And if you don’t send me a refund, I’ll report you for theft/fraud/etc.

      4. Victim sends scammer a check, which the scammer deposits.

      5. The bank detects the forgery (probably when bank the original check claimed to be from replies and says “Nope, we never issued that check”) and un-posts the deposit from the victim’s account. The scammer keeps the “refund” from the victim.

      I’d go with Ialdabaoth’s idea. If it is a scam, the bank will figure it out when they talk to the alleged issuing bank.

      • qsz says:

        This is very likely to be a scam, and if so, it can take surprisingly long for your bank to detect the forgery.

        Here’s a quote from a scam victim support site (source) explaining how this can happen:

        ” Account information on a legitimate bank account is stolen and forged checks are produced using that account number. A victim receives one of the forged checks and deposits it. Their bank checks with the paying bank and sees that it is indeed a legitimate account number and the funds are there to cover the check. Eventually, the account holder whose account was stolen notices the forged checks and reports it. The paying bank contacts the victim’s bank and returns the check as fraudulent. The victim’s bank notifies the victim and removes the amount for the check from the victim’s account. This entire process could take months. Remember, you are financially and usually legally responsible for all checks deposited into your account, even after the bank has made the funds available to you.”

    • Liskantope says:

      Something very similar happened to a friend of mine a couple of years ago (it was even from a “college of the arts”). Upon careful investigation, it was discovered that a scammer had gotten his address from a transaction on Craigslist. Unfortunately, I don’t remember many other details of the situation.

  2. Ialdabaoth says:

    So, as a continuation of last month’s Open Thread subtopic:

    Tell me about your gaming projects, guys.

    • The Anonymouse says:

      Not a party to the previous open thread, but I’ve become quite interested in Twine as a (somewhat interactive) fiction platform. I had previously tried to learn Inform 7, but the hurdle seemed a little much for someone who has lots of other real-life stuff to do and less desire for the “game” aspects and more of the storytelling aspects.

      Twine, fortunately or un-, is very simple, and does lend itself to more CYOA feel than “game” feel. Largely because of this, and because of the very low barrier to entry, it feels like a huge number of Twine games tend to be of the painfully-confessional-to-the-point-of-eyerolling variety (think, social justice warriors writing high school poetry).

      Anyway, that’s what I’m up to (you asked :)). Anyone else here ever use Twine?

      • Matt C says:

        I got interested in CYOA and tried Twine out a while back. I got frustrated with it pretty quickly. It didn’t interact quite the way I wanted, it had some long standing bug that was biting me, it didn’t seem to be under active development.

        There are also some CYOA websites out there, at least one of which had quite a few stories up on it. Started writing a website of my own but wandered away from the project after a while.

        Good luck with your story.

      • Erik says:

        RPG Maker has a similarly low barrier to entry, only instead of high school poetry, for me it was token combat mechanics on a prebuilt map encountering (for example) five subspecies of unicorn all represented by the same horse graphic under a different name like “Moon Unicorn”. Still painfully confessional to the point of eyerolling regarding the unicorn society depicted through badly spelled dialogue. The horses didn’t even have a palette swap.

        I nostalgically recall leveling up about five renamed variants of the default stabby class and learning the not renamed Cross Cut skill when testing these games some years ago. Then I got tired of it, wrote an angry rant about shovelware and left the community in question. Ah, memories of youth. 🙂

        • nydwracu says:

          I was vaguely aware of RPG Maker, but for me it was Game Maker: made a few incredibly lame platformers that I have long since lost.

          I also knew enough Actionscript back then to make a crappy Moon Patrol clone in Flash, but that’s also lost, as was my QuickBASIC text adventure.

          Most game-making I did on my calculator during math classes — the one thing public school was good for. But once I wrote Sokoban, I stopped writing games and started making levels for it.

          I also made a few levels for a Lemmings clone with the unfortunate name of Lix, but that was a few years ago and it looks like the community levelpack doesn’t exist yet.

      • Randy M says:

        I had a good RPGmaker start back in the day–filled up one Playstation memory card. But I lost my notes and got discouraged trying to continue from memory, and by now I don’t think my playstation works anymore.
        And tying on that thing was a pita, despite how fun it was to plot out a branching storyline and tricky dungeons.

    • Oligopsony says:

      My 5e PHB arrives tomorrow and at some point when I have much more time than I do now I want to run the heck out of it.

    • Matthew says:

      I don’t have any projects per se, but I did just have a date with a pretty girl whose reaction to a description of my absurdly nerdy favorite board game was “that sounds like fun!” rather than “what is wrong with you?” I’m sufficiently pleased about this to note it here.

    • Icicle says:

      Over the past (long interval of time), I’ve been working on a Minecraft map. It’s coming along at a reasonable but slow pace.

      Intent: A very difficult Minecraft adventure map, like the Super Hostile series. Open-world, with lots of stuff to explore.

      Setting: Large floating island in the sky covered in snow. The interior is arranged like an onion composed of different rock layers, each one very roughly corresponding to a difficulty jump. Within that, though, there are lots of other places. Clouds, an arcology, spider nests, a frozen lake, several unfrozen lakes, Yggdrasil, caves of crystal, glaciers, a volcano, really big chasms, lava chambers, mines, shrines, a greenhouse… yeah, I’ll never finish all this.

      Causes of delay: In most Minecraft maps, the areas to be completed are planned out in advance. In this one, I’ve pretty much resolved to make a whole bunch of short dungeons and caves until the whole island looks like extra-holey swiss cheese. And there are probably several mapmaking tricks to speed the process up that I don’t know of. I need to try harder to speed it up.

      Hazards: Well, I’m trying to kill the player. There are many, so I’ll just list the more unusual ones that have been created so far. Giant sponges, Ghasts that can blow up solid rock, lightning rods, flaming rabid wolves, exploding trees, crafting tables, glacier crevasses, squid, inconveniently located lilypads, that sort of stuff.

      Plans: Since trying to finish it is something that will probably take a long time and result in me getting fed up and quitting before its done, I just work on it to funge against internet time and relax after a long day. By enjoying the process, and not the destination, I am substituting work on a creative effort that I enjoy for internet goofing off time. And in the unlikely event that it ever gets finished, I’ll put out the call for beta testers here. Next action is to finish up the End area and the Nether area, and figure out how to make good dungeons faster than I have been doing.

    • rsaarelm says:

      Programming a roguelike. In Mozilla’s Rust programming language that’s still pre-1.0 and has been changing from under me for the last year. They seem to be getting closer to stability these days though, and talk about a 1.0 release this year.

      Project’s mostly bogged down in software architecture design morass instead of actually writing a game. Rust has a really neat vaguely Haskell-like type system, but it’s also going for the C++ niche of high-performance programs that don’t use garbage collection. Trying to write a game with a complex object model in a language that doesn’t really do garbage collection might not have been the best idea for achieving a really fluid and natural mapping of ideas to code.

    • Michael Wittig says:

      Latest wacky game idea: Take the tightly-constrained base-building of XCOM: Enemy Unknown, the side-view digging glory of Terraria, a bit of the town management of Dwarf Fortress, a bit of Kingdom Rush’s Tower Defense mechanics, and then have all of that feed into a classic turn-based RPG using the heroes you recruit to your town. Code-named Dwarferria at the moment.

      • Emile says:

        Dammit. I’m also working on a side-view turn-based digging game with town/dungeon management. Next thing you’re gonna tell me you also want to use procedural generation a lot.

        Okay, could the people here who are clones of me please raise their hands?

    • suntzuanime says:

      In theory I’m writing a tactical RPG roguelike influenced by XCOM (the new one) and D&D 4th Edition.

      I was trying to put the WIP up on a git repository but was confounded by how terrible git is, and that sent me spiraling into a month-long depressive funk in which I’ve gotten nothing done.

      • lmm says:

        Urk. Bitbucket does mercurial if that makes the difference?

      • Anonymous says:

        Man! I’ve been caught into that kind of trap, it sucks! But now I got better at git 😀 (use it everyday at work) D’you want help and advice? (I am quite fond of tactical games…)

        What technology are you using? Unity?

        (this is by Emile by the way, I forgot to log in. You can ping me at flammifer at (obfuscation parenthesis for spammers) gmail)

      • DanPeverley says:

        That sounds like my sort of stuff. Care to share any further details?

        • Emile says:

          Looks like a lot of people here are into that kind of stuff!

          Who here would be up for working together on some open-source project just for fun?

          I’m thinking of something in the general area of turn-based tactical RPGs/management games/eurogames, web-based, with all the data tables in shared google docs (at least for starters); I like web-based stuff because it makes it very easy to test and show to people…

        • rsaarelm says:

          I’m thinking of something in the general area of turn-based tactical RPGs/management games/eurogames, web-based,

          Well, I have been thinking I should do some HTML5 rich client web stuff just to show that I have.

          On the other hand, I’ve seen countless internet team game projects shrivel up and die. The only way a game that actually requires nontrivial core implementation can be done is to basically have a project leader who owns the project and can bring up a skeleton themselves, and other people who contribute extra stuff. A tactical RPG is going to require some reasonably serious engine innards.

          Ways to get around this: You could find an existing RPG Maker style engine that can do tactical RPGs, and focus on the content. The project would then be more like mod authoring, which I think has a slightly better chance of working with a distributed team. I can’t think of any good web-based tactical RPG engines offhand though.

          Another trick would be to just start designing a board game instead, and use something like Roll20 or Vassal to playtest it online. Coordinating an actual computerized version might have a lot more steam behind it once there already was a playable and fun board game design in existence.

        • Emile says:

          The only way a game that actually requires nontrivial core implementation can be done is to basically have a project leader who owns the project and can bring up a skeleton themselves, and other people who contribute extra stuff. A tactical RPG is going to require some reasonably serious engine innards.

          Sure, it’s a fairly big task, though I’m thinking more of the low end in terms of complexity (think Munchkin, not Disgaea, though Munhkin’s mechanics are a bit too simple). And I have made various flavours of game engines, in Python or web-based: for example see Little Dungeons which I made a couple years ago; it’s far from being a finished game but the core mechanics works (I stopped working on it because I felt I’d need to revamp the gameplay if I wanted to be able to go in more interesting directions, and also I wasn’t happy with my choice of technology).

          Now I’m working on “spiritual descendents” of that project, but with newer design ideas and a different technology (Angular JS, and data retrieved from Google Drive). I’m not totally sure of what I want to build in the end – on whether I just want to prototype interesting mechanics, or start making a full-fledged (simple) game.

          [Edit] For anybody trying out Little Dungeons by the way – there are no explanations or tutorial (and little in terms of contextual cues, but more than in the average roguelike), if you want more details ask me, I’m not going to describe the rules here so you have a stab at figuring them out yourselves >:-D

        • suntzuanime says:

          I have about 60% of D&D 4e’s combat mechanics implemented already, as a proof of concept. An engine on the level of Disgaea does not really seem like an impossible task. I have some confidence in my ability to write code for a game engine, were I able to do anything.

          The things I might need help with would be:

          *art/graphical presentation. Right now it’s all done in ASCII, in true roguelike tradition. Obviously nice graphics would be better, but I have less than zero artistic ability.
          *game design. I’ve implemented a subset of D&D 4e as a proof of concept that I can implement complex mechanics, but for a final project I don’t want to just be ripping off D&D wholesale. I have some hopes for my game design abilities but it’s a hard task.
          *AI coding. AI in games doesn’t actually have to be good, but it shouldn’t be pathetically bad, and that’s tough in complex game systems. Again I have some confidence but it’s a hard task.
          *web programming. I know embarrassingly little about how to actually make things go on the web. Certainly a webgame would be easier to distribute than an executable you have to download. I am confident I could learn web stuff, and I probably should anyway, but it’d be nice not to have to.
          *content creation. The more content the better, so long as it’s good, and content creation is really easy to crowdsource. Crowdsourced content did great things for Dungeon Crawl Stone Soup, which I respect even though it is no longer my favorite roguelike.
          *playtesting is obviously a great thing to crowdsource because you don’t have to have any skills at all to contribute.

        • Emile says:

          *art/graphical presentation. Right now it’s all done in ASCII, in true roguelike tradition. Obviously nice graphics would be better, but I have less than zero artistic ability.

          I have *some* (I have a graphics tablet and am somewhat proficient in gimp and inkscape), but not to the point where I can quickly create a lot of graphics. Fortunately there are a fair amount of free graphics available – I like David E. Gervais’ roguelike tiles, but there is better stuff too.

          At least, there’s enough around for making a “visual” game to take full advantage of UX improvements like contextual previews of actions, feedbacks on what happened, etc. Making a game with a distinct graphical identity requires either specific artwork or a choice of theme that works well with abstract procedural graphics / simplified pixel art.

          *game design. I’ve implemented a subset of D&D 4e as a proof of concept that I can implement complex mechanics, but for a final project I don’t want to just be ripping off D&D wholesale. I have some hopes for my game design abilities but it’s a hard task.

          Agreed. Which is why I tend to be in favour of simpler mechanics that are easier to prototype and balance (have you tried Little Dungeons? What do you think of the combat mechanic?)

          *AI coding. AI in games doesn’t actually have to be good, but it shouldn’t be pathetically bad, and that’s tough in complex game systems. Again I have some confidence but it’s a hard task.

          I prefer “design solutions” to AI; for example in Little Dungeons you can pick your party’s attack, but the enemies always do the same thing – they don’t even have a concept of Special Attack. I like the way monsters in 4e work, they tend to pretty much do one thing, whereas the players have more tactical choice.

          (most tactical games out there have actual AI, but I think there’s interesting design space in doing varied “dumb” enemies (think Mario) too.)

          *web programming. I know embarrassingly little about how to actually make things go on the web. Certainly a webgame would be easier to distribute than an executable you have to download. I am confident I could learn web stuff, and I probably should anyway, but it’d be nice not to have to.

          Yep, that’s an explicit choice I made some years ago, after I was tired of making Python games that were complicated to show and distribute. Now I know how to make stuff on the web, though I still have some areas of uncertainty.

          *content creation. The more content the better, so long as it’s good, and content creation is really easy to crowdsource. Crowdsourced content did great things for Dungeon Crawl Stone Soup, which I respect even though it is no longer my favorite roguelike.

          Definitely! I have a prototype that reads game data from a google spreadsheet, which would be super convenient for creating game data – unit stats, quests, spells, whatever.

          Creating images and levels would need another tool …

        • suntzuanime says:

          That Little Dungeons prototype is maybe a bit simpler than I’m interested in. Something that stripped down is more of a puzzle game than a tactical game. Which is totally fine! There’s obviously a huge market out there for puzzle games. But it’s just not what I want to make or play.

          The problem with AI for D&D 4e is that while the monsters mostly do simple things, the players often do complex things that the monsters need to react to intelligently. E.g. the fighter marks a monster, and the monster would be very foolish to attack anyone but the fighter; the wizard makes a persistent cloud of poison gas and the monster would be very foolish to enter it; the rogue blinds the monster for a turn so it probably shouldn’t waste its powerful recharge 6 power on a turn it’s likely to miss.

        • Emile says:

          On Little Dungeons as a puzzle game: I agree! That’s one reason I stopped working on it, the system didn’t allow for enough strategic depth and character customization (and did encourage just planning everything out, which is not very fun), which is one reason I stopped working on it (the other being that some aspects are a bit overengineered, for example all the combat logic happens on the server side, which is nice to prevent cheating but makes development more complicated).

          On AI in D&D4E: I agree that even the simple ennemies in 4E would still need a bit of AI; though less than more symmetrical games like Final Fantasy Tactics. One solution is to do the AI (it’s not *that* complicated, especially if the enemies are sufficiently different / quirky that you can’t really compare their performance to yours), and another is to focus on game mechanics that don’t require AI. Tha later would be a bit of a pity for something like 4E who has a lot of nifty rules like flanking that make tactics interesting…

        • Ialdabaoth says:

          I have a decidedly non-D&D4E mechanics system specced out, if you guys are interested.

          It also has a pseudorandom AI that’s extremely simple to implement (it was originally designed to be implemented by hand, by smart 12 year olds, without a computer).

    • Emile says:

      I’m making a few javascript prototypes of game mechanics.

      I’d like to make a more complete web-based management/dungeon crawler game, I have quite a few ideas, but first I’m prototyping a few mechanics to see which ones work well and look promising, and kicking ideas around; aiming for the nice combination of simple system + rich possibilities.

      (I’m thinking of some kind of side view turn-based game where you alternate between dungeon-keeperis management and simplified dungeon crawling)

      • Ialdabaoth says:

        I have a system ready for this if you’d like; it’s designed for a card/board game rather than on the computer but that just means it’s forced to be as simple as possible.

        • Emile says:

          Why not, could be interesting! Is it published somewhere? (I’ve been following some of your design stuff on and off here ‘n on IRC, your mech game looks pretty well made)

          I’m not actively looking for new mechanics, my main focus now is implementing the ideas I have (I have many pages full of abastract analysis and brainstorming and conceptual stuff);

    • James says:

      I’m impressed at the range of responses to this query. Who knew that so many readers of this blog *would* have gaming projects? (Am I the only one who doesn’t!?!?)

    • Hardly a project, more a thing-I-thought-up-last-night which I’m never going to have the time to do in a million years.

      In the early 90s, I played a game called RoboWar, in which you programmed a tank-like robot to run around a game field shooting other robots. It was fun. Last night I was watching a lecture about HRI, and wondered how much more fun it would be to play a cooperative game alongside a robot of your own design.

      An obvious prototypical form for this would be a RoboWar-style tank game, but it’s got me thinking about what other games you could model on this pattern.

      • Ialdabaoth says:

        Oh man, you know what’d be amazing? A FarmVille clone, where you have to design and program terraforming and farming robots that are going to get dumped on Mars.

        Bonus points if you also have to design and program the skycrane delivery robots that drop them onto the map.

        Serendipity: this would segue nicely into my Unspecified Educational Game Project.

        • Emile says:

          The farmville robot thing does sound like a good idea! More fun than lightbot, works better than roborally for single player, and more visually interesting than the prisonner’s dilemma tournament.

          Based on a slightly random terrain, you could have nice choices between a simple and safe strategy, and strategies that try to “optimize” your growth but can Go Horribly Wrong™

          You could even have a nice setup where everyday everybody’s robot is dropped on a planet, and compare their high scores; you can see what others’ robots did but not their programming…

    • ckp says:

      I’m making a KOTH-gamemode map for Team Fortress 2. Alpha version will be released within about a week so I can get feedback on the layout. I chose KOTH because it just needs a single, compact, central fighting arena, and thus is much easier for a newbie mapper to balance than more complicated modes.

      I plan to use what I learn for a more ambitious project I’ve had churning in my mind, which is a Dustbowl-style Attack/Defend map where BLU must scale a RED skyscraper.

    • Ialdabaoth says:

      So I’m working on:

      This is my card game about building and fighting giant robots. Right now, I’m doing some last-minute card balance tweaking, and deciding on a final art style and art budget (if any).

      Current card list is here (dropbox gallery).

      MechBrawl is kind of cool, in that it uses nothing but cards to simulate a fully tactical 1v1 combat between two giant mecha, which each player can customize to their play style. Here’s how it works:

      Deck Creation
      Deck creation happens before a game.

      1. Each player chooses a Pilot card. Your Pilot will determine your hand size, and your base accuracy with each type of weapon. Additionally, each Pilot has a special ability that they can use once per turn.

      2. Each player chooses a Chassis card. Your Chassis will determine how much damage you can take before exploding (or losing weapons), how many Maneuver cards you can play per turn, and how many weapons you can mount.

      3. Each player chooses the Weapon cards that they will mount on their Chassis. There are four kinds of weapon: Energy, Ballistic, Missile, and Physical. Energy weapons are typically longer-ranged and more damaging; Ballistic weapons are more flexible and can fire faster; Missile weapons can only fire once per battle but are extremely powerful and compact; Physical weapons are also very powerful but very short-ranged and take up more space. Certain weapons have secondary capabilities beyond dealing damage – for example, Shields and Reinforced Plating can reduce damage dealt to you, and Jet Packs and Servo Boosters let you move faster when you play maneuvers.

      4. Finally, each player builds an actual shuffled deck of 15 Maneuver cards. There are three kinds of Maneuver cards: Movement Maneuvers (that let you change the range between you and your opponent, and often have stay-in-play effects), Strike Maneuvers (that let you shoot your weapons at your opponent), and Response Maneuvers (that change the effects of your opponent’s maneuvers in your favor).

      Players play back-and-forth by playing Movement or Strike maneuvers to attempt to get in range with their weapons and shoot them. Their opponent can then play Response maneuvers to attempt to affect those Movements and Strikes. Weapons automatically hit during a Strike unless your opponent can get out of range of them, or reduce their Accuracy to zero (or less) via Responses and other tricks.

      If you’re hit by a weapon, you decide where to assign its damage (and can split that damage up) unless some card effect says otherwise. Your Chassis card has four sections, each with its own armor: Left Arm, Right Arm, Torso, and Legs. If your Legs armor drops to 0, you can only play one maneuver per turn. If an Arm’s armor drops to 0, all the weapons on that arm stop working. If your Torso’s armor drops to 0, you lose.

      The game is taking about five to fifteen minutes to play at this point.

      Eclipse: New Dawn for the Galaxy is an amazing board game that I’ve picked up recently. It’s a 4X galaxy exploration and conquest game with everything that made Masters of Orion amazing twenty years ago, but as a board game instead of a computer game.

      I’m working on designing and printing up an alternative Research tech tree for the game, and maybe some alternate combat rules.

      Unspecified Educational Game Project
      My work is dangling an $80,000 grant in front of me and asking if I can do something cool with it relating to educational computer games. More on this as it develops.

    • Nornagest says:

      In theory I’m working on the same MUD I’ve been working on for years. We’re aiming at a simulation-heavy New Weird-influenced low fantasy setting: think two parts Vinland Saga to one part Viriconium or Zothique and one part old-school pulp. Even after this long, though, it’s very much a work in progress; right now I’m (theoretically) focusing on improving the player economy and persistence features, in order to give the playerbase something to build off of while I start putting in cruel survival-oriented features.

      In practice, though, I’m mostly doing other things.

    • Multiheaded says:

      I want to one day run a Dungeonpunk Communist Star Trek collective game (“quest”) on 4chan’s /tg/. I have the fluff, but desperately need a good world map and to lurk more in their “quest general” threads for advice and such; I’ve only ever had some abortive attempts at this sort of thing before.

      Basic fluff: a century ago, the kind of apocalyptic Heaven vs. The Other Side conflict that much D&D-inspired fantasy features was won by the “good” side, who immediately set about genociding the “evil” races, suppressing forbidden ways of thinking which “evil” spellcasters used to tap into the Void, and building a (shamelessly NrX-inspired) authoritarian utopia. The world was also torn up a great deal, magitek made enormous strides in The Last War, and the magical Industrial Revolution arrived.

      The orcs, trolls, “dark” elves (those who lose connection to the traditional pantheon) and mutated/demon-touched humans founded a Republic that claimed the two previously uncharted/”evil”-held continents to fight back against the “good” Coalition of Light-worshipping humans and High Elves. The Coalition pacified and transformed the “civilized”/”core” continent. The Republic spent a hundred years exploring and settling and industrializing and resisting the Coalition’s slow crawl along one of the continents. They built a more or less functioning democratic socialism and things were tolerable. But as of now, a Messiah has descended from the Upper Planes to complete the Divine vision for the world, and the Coalition is blitzing across the second continent while the Republic tries to deny it routes over the magically-and-demonically-messed-up Great Chasm into the third continent. Magitek airships are the biggest new military thing and they work just like spaceships in ye olde space opera.

      The protagonist is a young airship officer who has recently defected from the Coalition for and had to assume command of a prototype flight-deck battlecruiser with a Diverse crew, and help it lead a caravan of refugees along the edge of the war-ravaged second continent, then across the Chasm and back home after a strategic encirclement by the Coalition makes the situation desperate. The ship works like in ST: Voyager, with different tasks for different departments, exploring the Chasm’s depths, skirmishes with the enemy, jury-rigged magitek upgrades, etc. Also action sequences with lighter aircraft and leading “away teams”/stealth missions. Play up the Diversity in the crew (cool old orc Tactical guy, hotheaded mutant human lady in charge of Engineering, nerdy and withdrawn dark elf Science head) and the need for Teamwork and Communist Values.

      • Emile says:

        So your players are going to be a band of trolls led by Captain Obvious?

        • Multiheaded says:

          Not quite… the players vote on what the captain does and how she reacts to things socially, either by picking out of several options proposed by the DM, or by favouring one write-in option. (I’m not too worried about trouble with maintaining the theme, /tg/ is generally good with this sort of thing once primed… if you don’t anger them with railroading and indulge them with carefully measured humour and pandering).

          Most options would result in me making some rolls and calculations based on the 40k RPG system under the hood – in particular, a streamlined and modified ship combat model from Rogue Trader, which I’m working on at the moment.

          The rest of the characters are essentially DMPCs – they have stats which figure in rolls, and the players have the captain give orders to them, but ultimately they are part of the experience and not whom the players identify with.

          So, uh, it’s more like trying to be troll enough for a band of Obvious Trolls.

    • ThrustVectoring says:

      I’m playing a significant amount of Kerbal Space Program.

      It turns out the rocket science is !!FUN!!

      • Ialdabaoth says:

        Yay! You should definitely grab some of my mods.

        (Well, most of them aren’t “my” mods anymore)

    • von Kalifornen says:

      I’ve been working on story and stuff for a text-based/art/point-and-click game (Low labor input…) that would be a parody of the Dead Space series, in which the player character is a mechanical engineering student at a university where there are lots of interesting projects (Solar car race, build-a-race-car team, labs with powered exoskeleton experiments, etc).

    • lambdaphage says:

      I really like turn-based tactical wargames. My ideal game would be 40k for linux with 1-2 orders of magnitude more realism. Frozen Synapse comes close (and is great in its own right) but I prefer platoon- or company-level engagements to the SWAT team feel of FS. Every few months I go on a google-binge, trying to find this ideal form. Finally, I decided to be the videogame I wanted to play in the world.

      Currently I’ve got the world’s most boring program that could technically be called a videogame, but this is my first foray into game development and I’m only one day in. Without indulging too many details, I want to write a game that emphasizes uncertainty, communication, cooperation, intelligence in the military sense, the difficulty of acting effectively under stress and (conversely) the value of suppressive fire. Hopefully something like ASL as imagined by John Boyd.

      If anyone knows of any good tbt games that will run on a linux box, on the other hand, I’m all ears.

      • Emile says:

        About suppressive fire, The Dawn of War series used it pretty well.

        On Linux games, have you tried Battle For Wesnoth, The Only Good Game On Linux? (Though last time I looked must have been like ten years ago, maybe new better games came out)

        • lambdaphage says:

          Dawn of War looks cool, though I wonder if you lose something in translating 40k to RTS. I’m an embarrassingly big sucker for the space goth stuff, though. A galactic civilization nearly as indifferent to human life as the forces it opposes, technologies bequeathed unto generations that have long since forgotten their workings, war at scales that ensure the brute anonymity of the dead. Good stuff, especially considering that most of it must have been dreamt up by three dudes in a basement huddled over a pizza box.

          I have played Battle for Wesnoth. It’s well done, but it doesn’t really scratch the itch of simulating a platoon commander’s decision-making process. It’s the closest thing I’ve seen to the game I want to play, though. Or maybe European Escalation, though I’ve never played it.

        • Andy says:

          Dan Abnett’s 40K fiction is really really really good, too.
          Dawn of War I used to like, but what I missed most – especially in the sequel – was the feeling of deploying a customizable standing army. The first one had a great deal of base-building, and the second didn’t have the amount of customizability that I wanted. That’s the feeling I want to capture with Elemental Brigades (my project described below)

    • Andy says:

      Mine is a Dungeons and Dragons crossed with 40K, or more accurately, GW’s large-scale fantasy battle system Warmaster. Each player plays a sorcerer-general with an army of soldiers and magical creatures, and players have to work together as a party to go through a campaign and achieve an overall goal. Also vaguely inspired by Magic, though I hate collectible card games with a special passion. Hex-based battlefield and combat mechanics, where each piece represents ~1000 soldiers or an equivalent number of magical critters. Most armies would be the Sorcerer, bodyguard, and 3-6 units.
      The army lists are broken into six Elements, each with different tactical specializations: Earth and Plant are stubborn defensive armies, Air and Fire are fast but fragile offense, and Water and Death are flexible and can fit in multiple situations.
      This will (hopefully) require players to work together like a DnD party – a Plant army taking a Tank-y role defense role with Water support, while a Fire force hits the enemy on the flank, for example. But campaign-scale mechanics, weather and supplies and all that, not to mention worldbuilding, are the pain in my butt right now. So I’ve mostly been working on projects that don’t require that level of commitment.

    • Ins Mietermlm says:

      I’m working on designing ~7 card/board games, all at different stages of development. In order of finishedness,

      1 microgame which is finished, but I’m not sure how fun it is. The main mechaninc is working with different card backs in the same deck. I’m still deciding between making it animal-based or monster-based.

      1 tactics-based card game designed to feel like the epic showdowns at the end of action movies or some books. I’ve managed to make it pretty balanced, but I’m not getting the feeling I was aiming for. Could just be because it’s me, though.

      1 asymmetrical CCG where one player’s deck is a dungeon or evil plan that the other player works through. The mechanics seem solid, I just need more cards for it.

      2 ideas for games focused around one concept of existing games

      2-3 games based on already existing material (games mentioned in books, pokemon …)

      I’m just having a really hard time playtesting them all enough.

    • Anthony says:

      Not my project; not even a “real gaming project”, but the Castalia House blog has had a number of interesting posts on the early history of D&D and similar. Most recent is RETROSPECTIVE: Warriors of Mars by Gary Gygax and Brian Blume; there’s also this post which only looks like a book review, and some other book reviews that veer off into discussions of their influence on early RPGs.

    • Auroch says:

      I have a couple card game prototypes.

      The older one is an auction game where each card represent a company which either produces some resources straight up or turns several resources into a better resource. Companies which need no input or which you supply with all the input they need ‘run’, and score points. The idea is to complete a supply chain where you go from the 1-point mining companies which need no inputs up to the 10-point companies that use advanced resources and don’t actually produce anything. It’s coming along really well, and I’m currently looking for people to playtest it blind.

      The second one is summarized as ‘a deckbuilder, only played on the board from Labyrinth’. What cards you can buy isn’t limited by your money, only by your position on the board, which is mutable. Also you build a hand rather than a deck (no shuffling, just draw everything when you run out of cards), and it’s a lot like Eminent Domain with some overtones of SolForge. This one’s pretty rough, but I’ve found a local designer’s meetup to bounce ideas off, so I expect I’ll make progress quickly. (Or find out that it sucks. Or both.)

  3. Alice Monday says:

    Automating that sounds pretty easy. I would like to try doing this.

    • Scott Alexander says:

      What do you need, admin access to the wordpress panel? Give me your email and I will send you some info.

        • Bakkot says:

          As LW user Bakkot, I know nothing about WordPress but would be happy to help if you have questions or difficulties regarding the script itself. You can reach me at [my handle] at gmail.

        • DavidS says:

          First of all, thank you very much to Alice Monday and Bakkot!

          A suggestion for a useful feature: Could the plug in be made to insert an unusual text string at the end of the new comments (say NEWNEW)? I could then navigate through all the new comments by CTRL-F.
          Thanks if it is easy, and please don’t feel obliged if it isn’t.

          • Alice Monday says:


            $=jQuery; $(document).on(‘keydown’,ι => {if (ι.which === 70) {t=$(‘.new-comment’).eq(0); location.hash = t.closest(‘li’).attr(‘id’); t.removeClass(‘new-comment’)}})

            in the console. the .which property of the ‘f’ key is 70, so if you’ve pasted this in the console, it’ll go to the next new comment and de-highlight it whenever you press the ‘f’ key

            (the quotation marks are messed up; replace the ‘’ characters with the single quotation mark character)

  4. Aris Katsaris says:

    How do you feel about directly asking for donations via Patreon?

    • CaptainBooshi says:

      He said last time this subject was brought up that he strongly dislikes directly asking for donations like that.

      • Michael Wittig says:

        There is only one path left, then: the selling of tshirts.

        I mean, hey, who wouldn’t want a shirt that has “‘It’s the Leviathan!’ Tom said superficially” on it?

        • Rachael says:


          While I love your serious posts, I’d actively pay for the puns.

        • David says:

          I dunno, I think I’d prefer the one with ‘I broke my back lifting Moloch to Heaven and all I got was this lousy Disneyland with no children’ myself 🙂

        • Vulture says:

          @David: Actually, I think he tried to sell that one, but the print-on-demand people shut it down because everyone’s afraid of Disney’s trademark enforcement.

        • David says:

          everyone’s afraid of Disney’s trademark enforcement

          Damn you, Moloch! … I mean, Damn you, Disney!

        • Paul Torek says:

          Replies to this show that posts with fifty puns in them are not an annoyance at all, to plenty of us. “Going commercial with Amazon,” Scott adverted.

      • Braden says:

        Then he doesn’t need to ask. He can explicitly say, with his link to Patreon, that he discourages donations, especially from people not making a lot of money, and requests donations to a Givewell charity instead. But realistically, I will not bother to increase my charitable giving, and I would be happy to make a recurring donation to subsidize this blog. I would appreciate it if he would allow that.

  5. Joe says:

    Did Ozy ever find a job?

    • Scott Alexander says:

      Ozy is working on applying for App Academy. They have gotten two levels through what is apparently at least a three level process, and my fingers are crossed.

      • Ialdabaoth says:

        I strongly approve of this and would like to offer algorithm and data structure design tutelage, SQL schema design tutelage, and/or HTML/CSS tutelage if it would be helpful.

  6. AMac says:

    “Scott, you said, ‘especially if reader annoyability is a limited political-capital-style resource I can spend on (for example) posts with fifty terrible puns in them,’ but I enjoyed discovering Tom Swift’s preferences,” AMac noted wordily.

  7. jaimeastorga2000 says:

    Dr. Alexander, are you planning on creating an account at some point? Perhaps before editing RationalWiki but after joining

    • Tumblr is actually quite good for this purpose, if you open your askbox, which Scott has not done.

    • Scott Alexander says:

      I have enough trouble finding excuses not to answer the questions people already ask me! I don’t need people asking me even more questions I need to find excuses not to answer!

      • potatoe says:

        This is the perfect use case for – simply stipulate that when people submit questions, they must also provide a ready-made excuse to ignore their question. Now you’ll be even further ahead, because you got a question which you can answer, or which you can choose to ignore based both on the question’s goodness and the additional information of how convenient the excuse is.

        • Auroch says:

          And if they don’t provide an excuse not to answer it, then you can refuse to answer it because it didn’t fit the criteria!

  8. syllepsis says:

    Do you keep a list of your blog headers? (the thing where it currently says ‘not enough hubris not to try to ill god’)
    I find them clever and wonder if I missed any. Also might show something mildly interesting about the trends to your blogging

    • CalmCanary says:

      To save other people the trouble of digging through the archive:

      17 Feb 2013 – 11 Nov 2013: “With malice toward none, charity for all, confusion about most, secret crushes on a few, and miniature flags for others!”

      12 Nov 2013 – 4 Dec 2013: “The proud exception to ‘Don’t read the comment section'”

      28 Dec 2013 – 14 Feb 2014: “Previously an exception to ‘Don’t read the comment section'”

      24 Feb 2014 – 25 Feb 2014: “Tell one lie, and the truth is forever after your enemy”

      28 Feb 2014 – 27 Mar 2014: “Peripherally associated with the Gwernosphere”

      28 Apr 2014: “Less [adjective] than Zeus”

      2 May 2014 – 18 May 2014: “Still confused, but on a higher level, and about more important things”

      30 May 2014 – 8 Jun 2014: “Relatively non-swashbuckling!”

      25 Jun 2014: “As cited in a published academic paper! ( )”

      8 Jul 2014 – 18 Jul 2014: “Not a social justice blog! Stop only reading my posts about social justice!”

      1 Aug 2014 – 19 Aug 2014: “Not enough hubris not to try to kill God”

  9. RCF says:

    Do you just not reply to emails, or do you not read them?

  10. Alexander Stanislaw says:

    I’m very surprised and sorry to hear 2. I can’t recall more a few commenters telling you that you shouldn’t be writing X, but I suppose it sticks out whenever it happens. I get the impression that most commenters are extremely grateful for you putting so much effort into this blog and creating a relaxed atmosphere as opposed to the occasionally finger wagging feel of LessWrong. But I do realize that criticism hurts more than praise feels good.

    • The Anonymouse says:

      I initially came here because I like reading Scott. I didn’t realize at first how much I would enjoy the commenters, particularly the fact that it remains civil (by commenter standards) in the face of such a diversity of opinion. (I even secretly like Multiheaded!)

  11. Could you name the curtains that block light, the industrial-strength earmuffs, and dream diary that looks like a sinister 16th century magical grimoire which you mentioned in your post on how Amazon reviews usually actually work? The first two would benefit my sleep and concentration greatly, and I’m just fascinated by the grimoire.

  12. Nobody Really says:

    So I’ve been working on a new idea which I call “Too-Effective Altruism”, working off of facts like this. TEA occurs when Effective Altruism achieves the goal of saving the maximum amount of lives per dollar. Using lives saved as a measure of The Good is short-sighted because the carrying capacity of the Earth for the human population is limited. Gnon/Moloch demands a trade-off between number of lives and quality of lives, and this is ignored at great peril.

    Western medicine and agricultural technology, combined with the capital of Western altruists, removes the limits to population growth in the world’s poorest areas allowing population to go exponential. But these new populations are dependent on Western donors – a fragile system.

    High-tech human society can be conceived of as a race with catastrophe. Exponential growth in the human economy will hit a limit eventually given fixed planetary resources. Technology pushes that limit further away, increasing resource consumption brings us closer to it. Population growth accelerates the crisis: there are more minds to think up new technologies but there are also more mouths to feed. To use a Poker analogy, it is raising all-in.

    Effective altruism tends to add lives in the poorest areas of the world since that is where it is cheapest to save lives. Unfortunately, these areas are the least capable of pitching in to the technological struggle of enabling eternal economic growth with fixed planetary resources. The technologically advanced sections of the world aren’t growing, but are aging and shrinking. Ameliorating the situation somewhat is that poor economies use fewer resources per person, though they also have very low standards of environmental stewardship.

    Effective Altruism needs a dose of long-term thinking. Female education is probably a better investment for long-term utility since female education both increases economic growth and lowers birth rate. This is better than lowering deaths and raising births in the present through food and medical aid.

    Lastly, I am skeptical of the implicit moral foundations of the Effective Altruism movement. There is a implicit egalitarian/utilitarian assumption that creating more human minds is better than not making them. But quality of life matters, including the nobility and stability that comes from a culture being capable of sustaining itself without constant outside charity. Maximizing the number of dependent humans kept alive with the money of altruists is a questionable moral goal.

    • The Gates Foundation has written a pretty good takedown of the idea that saving lives in low-income countries leads to overpopulation. In particular, anything we can do to help increase life expectancy, income, etc. in such countries will help move them closer to the demographic transition, at which point their populations will stop increasing so quickly.

      GiveWell has also written about these kinds of deep value judgments. They believe that saving or improving lives can have what they call flow-through effects that cause direct short-term benefits to lead to indirect long-term ones. They are also actively looking for charitable opportunities designed to work over the long term.

      I believe GiveWell’s own staff members have differing views on whether it’s morally better to save lives or to improve them; there isn’t a consensus on this point in the EA movement.

      • anon says:

        I don’t think we have enough evidence to say that population increases will always slow as countries become wealthier. Seems like there are a lot of potential confounds, like religiousity, income inequality, etc.

        In addition, economic development doesn’t occur without a tradeoff cost. If eg India gets a standard of living near the US’, I think that necessitates bad things happening to the environment. Might not be a good idea, overall.

        I’m not saying that overpopulation is a clear danger and Effective Altruism is certainly misguided. I’m only saying that I am uncertain what the best way to spend money is. Seems worth a more detailed analysis, at least, because both sides of the argument seem to have a point.

    • David Barry says:

      I’ve seen or heard multiple people argue that the long-term consequences of more children in the developing world surviving into adulthood are bad, and that therefore trying to save lives in the developing world is misguided. But I’ve never seen anyone bite the bullet and say that it would therefore be a net positive if people switched to campaigning against vaccinating poor people, so that more of them die.

      There is a implicit egalitarian/utilitarian assumption that creating more human minds is better than not making them.

      There are certainly lots of utilitarians in the EA movement, but we come in various stripes, and I at least reject the idea that creating new humans is necessarily better than not. I’d consider it overwhelmingly good news if the fertility rate in countries like Nigeria fell to 2 after women became more empowered in matters of family planning and a growing economy and so on. I don’t think this would even be a controversial opinion within EA — a common (somewhat tenuous) justification for preventing children’s deaths is that in the medium- to long-term, parents will respond by not having so many children.

      Effective Altruism needs a dose of long-term thinking. Female education is probably a better investment for long-term utility since female education both increases economic growth and lowers birth rate.

      I don’t think that education (either in general or for girls in particular) is largely absent in EA out of short-sighted thinking, but rather because GiveWell studied the topic as closely as GiveWell usually do, and they don’t think that any education charity has the sort of large, robust positive impacts that they want to see before making a recommendation.

      Relatedly, Population Services International was one of GiveWell’s recommended charities for a few years, and if the effective altruism movement so-named had existed in 2007, then reproductive health and distribution of contraceptives etc. would have been one of the main “headline causes” of the movement. But PSI dropped off the GiveWell recommendations a few years ago as GW became even more rigorously sceptical about everything, and so now PSI’s work is little more than a footnote in an occasional EA page. For better or worse, the EA movement (at least the one that cares most about human poverty reduction) relies heavily on GiveWell’s research and decisions.

      • Nobody Really says:

        We live in a world where all the population growth is coming from the crappiest places to live. The trend continues to the limits of crappiness. Africa is on average a pretty crappy continent to live on and it’s slated to add 3 billion people this century, but the crappiest pieces of Africa are expected to add more population than the relatively nicer places.

        I’m not sure how much Western altruists have to do with this population explosion, and I wish I knew more. Whether or not their $billions are having some effect, hopefully one of the following is true:

        1) The (aging, shrinking) population of the world that does science continues to perform technological miracles in agricultural/environmental science

        2) The population growth curves in poor countries start to bend down, for reasons that Taymon A. Beal states above, and soon start to look asymptotic.

        If (1) or (2) isn’t true, then it is probably best if Givewell gets things wildly wrong and ends up recommending extremely ineffective charities.

        But I’ve never seen anyone bite the bullet and say that it would therefore be a net positive if people switched to campaigning against vaccinating poor people, so that more of them die.

        Thinking wildly outside the box for a second, would it be better or worse if the technologically advanced world had a variant of Star Trek’s “Prime Directive” for dealing with technologically backward countries? Maybe donors from the West should be cautious about 4x-ing the population of a country that didn’t have the technology to 4x itself. This gives it time to develop the institutions needed to deal with the side effects of high technology.

        Unfortunately, I can’t think of any examples of charity making a poor country become not-poor.

        I’ve seen Effective Altruists criticize people who invest in charities that save lives or improve quality of life for the poor/sick/disadvantaged in the West since they do so at a cost which can save a hundred lives in a very poor country. But it seems that this kind of more local giving is less problematic and more likely to have a strictly positive outcome. For example, I’ve seen ALS charities criticized, which have been publicized recently due to the ice bucket challenge. Enabling Stephen Hawking to be Stephen Hawking is a very different thing from saving 1,000 third-world village children. The benefits of the two are hard to compare from a theoretical perspective, but it doesn’t seem to me that the latter is the obviously the morally correct thing to do if you have to choose where to invest your resources.

        On a similar note, it is not obvious to me that it is wrong to give to charity that helps residents of rich countries to live fulfilling lives (perhaps by paying for their education) even though it is very expensive and the same price could add hundreds or thousands of lives to a third-world country. Ayn Rand talked of “maximax” and “maximin” social philosophies. Helping humans reach higher than any human has ever reached is “maximax”, alleviating suffering or inequality is “maximin”. I definitely see value in maximax thinking whereas effective altruism seems mostly focused on maximin – and may not even do a good job of that if they are fueling a Malthusian trap.

        If you discover the next Beethoven or Einstein living in a poor neighborhood and help him pay for the college education that enables him to do amazing things, would that money have been better used on 10,000 mosquito nets? The theoretical framework of Effective Altruism doesn’t have the complexity to begin analyzing that dilemma.

        • David Barry says:

          1) The (aging, shrinking) population of the world that does science continues to perform technological miracles in agricultural/environmental science

          The number of people who don’t have enough to eat is falling in absolute terms, despite the rapid increase in population in the parts of the world with traditionally the greatest food problems. [WFP; see also the more dramatic famine deaths chart from The Economist.] So I think I am far more confident than you of the world being able to cope with more people.

          Aside from that more empirical point, I think we diverge more fundamentally in two places.

          Firstly, the sort of egalitarian instinct you’ve mentioned/alluded to is much stronger in me than in you. A life lived on less than a dollar a day is a very terrible life compared to how I live. But I still think that those sorts of lives are worth living. The death of a child is, at least, not much less of a tragedy simply because they were growing up in poverty.

          (We can play some sort of moral thought experiment game: if you were actually present in, say, Malawi, and had the opportunity and knowhow to intervene in an emergency, you’d be unlikely to start considering the carrying capacity of the land and the consequences of the future marginal children for the economy and then leave the people to die. Because they’re humans, and we’re also humans.

          If you want to get picky with my thought experiment, you could say that an emergency intervention doesn’t pose the same sort of incentive problems that supporting a developing world country’s health system does, but is that really what’s driving your moral instincts?)

          If it’s accepted that the moral worth of preventing a poor person’s death is at least not much less than the moral worth of preventing a rich person’s death, then it immediately follows that, e.g., we don’t need to spend much time comparing hundreds of lives in the developing world with a life in the developed world. It’s overwhelmingly obvious that the hundreds of lives outweigh the one.

          (And while an extra great scientist could potentially do amazing things for the world, it is highly unlikely that the person whose college education you fund is actually going to become the next Einstein or Borlaug. Overwhelmingly, the more typical case is that you help someone become a more productive member of society.)

          The second fundamental disagreement, more minor I think, is that I’m less risk-averse than you. It may be the case that contributing I’m contributing to a population explosion that in time will cause greater suffering than a baseline scenario. But I think it’s unlikely, and that the definite benefits that come from many extra people not dying outweighs the (small) increased risk of future disaster.

        • Nobody Really says:

          David Barry,

          I think you are correct in naming our areas of disagreement. From a risk-aversion perspective, adding 3 or 4 billion consumers to the world accelerates our confrontation with natural limits (I’m thinking of climate change), while these consumers are incapable of contributing to the science needed to address these limits. Even middle-income countries add basically nothing to science, it remains almost an entirely a rich Western pursuit.

          You are correct that if I lived in a third-world country, I might be more empathetic to the plight of poor individuals. If I had a pet deer, I might be more in favor of plans to eliminate the wolf population. I only note that empathy doesn’t necessarily lead to sound decisions.

          I guess I largely lack the egalitarian moral intuition common to the Less Wrong crowd. There seems to be an assumption that the utility of one person is as good as the utility of another. Keeping billions of humans alive as pets of Western altruists is seen as obvious good as long as their lives have slightly positive utility.

          My criticism of Effective Altruism comes down to (1) this moral framework doesn’t motivate me. I don’t see the “definite benefits” that you do of a more crowded globe. I do care about the irreplaceable natural beauty of Africa which will be lost as the population doubles every 30 years. And (2) it seems like EA could accelerate crises which are disasters no matter what moral framework you subscribe to.

          It’s natural to care more about people who are closer to you physically, genetically, ideologically, and in way of life. Universal egalitarian altruism may be more “rational” and pure in that it comes cleaner deduction from fewer axiomatic premises, but I think it only appeals to people who are already mentally biased to like clean, rational systems. It would be hard to convince me to care more about strange, incompetent people far away because I can help more of them per dollar than my neighbors, and I suspect these differences in moral intuition are somewhat unchangeable.

      • AspiringRationalist says:

        I’ve never seen anyone bite the bullet and say that it would therefore be a net positive if people switched to campaigning against vaccinating poor people, so that more of them die.

        Higher disease prevalence increases religious conservatism, which leads to a higher birth rate and less technological advancement, so vaccination probably decreases third-world population growth.

    • jaimeastorga2000 says:

      I recommend Garret Harding’s “Living on a Lifeboat” (and, to a lesser extent, his more well-known “Lifeboat Ethics”).

      • Nobody Really says:

        This is wonderful and very relevant, thank you.

        ”There’s nothing more dangerous than a shallow-thinking compassionate person,” he said in a 1987 interview with The New York Times. ”God, he can cause a lot of trouble.”

        from his NY Times obituary

  13. Multiheaded says:

    SSC commenters, what is your favourite race? Are there any races that you especially dislike?

    • George says:

      Personally, I think Scott should just make another post titled “Open Thread 4: Race War” and let you and Jim go at it.

      • lmm says:

        I wish there were such a thing. There are places where I’ve realised I’m confused about what one side or the other meant in an old thread here, but I can’t ask about it.

      • Vulture says:

        I would actually really appreciate this. Especially if gender was in-scope as well. It would let people blow off some steam, and I remember feeling frustrated in the past that I couldn’t comment on those topics in these threads.

        On the other hand, I’m sure it would be hell for Scott as a moderator.

        • Anthony says:

          Unless Scott decided to just not get emails from the thread, and ignored it for a week then closed it.

    • Took says:

      For me, definitely The Dwarves. I like their no nonsense attitude and love of technology.

      • Bugmaster says:

        I prefer the Kalashtar. A psionic blending of a humanoid mind and an eldritch abomination who fell in love with humanity and abandoned its extra-dimensional form to join it. Today, they wage a secret war against their former brethren (or technically half-brethren), wielding their psionic powers like a laser against the encroaching dark.

    • Sam Rosen says:

      Favorite Race: Wife Carrying World Championship

      • Multiheaded says:

        Truly, the Nordic races are superior and it would be a shame if they gradually disappeared.

    • lmm says:

      This is not the first time you’ve done this. It’s not funny. Stop it.

    • Deiseach says:

      Am probably contractually obligated or some such to nominate this as my favourite race.

      Now, can I ask if anyone has any opinions on Sir Thomas More versus Sir Thomas Cromwell? I am being recommended to read Hilary Mantel’s “Wolf Hall” and “Bring Up the Bodies” as good, well-written, historical fiction.

      My main obstacle to this is that I hate Sir Thomas Cromwell, and I am given to understand that Dame Hilary writes him like Ayn Rand writes John Galt or Howard Roark (i.e. the sum and acme of human perfection mmm mmmm he’s so dreamy).

      Me: The Tudor Court was a snake pit, and a bear pit, and a scorpion pit, and a tiger pit all in one.

      Them: I love subverting historical narratives of prigs like Thomas More
      Me: I am a prig myself, you are not discouraging me re: More or encouraging me re: Cromwell

      Them: She shows the private side of Cromwell; he loves his family and those who work for him and his hawks and small dogs
      Me: Yes, he probably would like small, yappy, annoying dogs that lie in wait to bite your ankles and piss and poop everywhere and God help you if you’re mean to poor little Teddy-Tiddles even if you fall over the obnoxious little beast.

      My cold, stony heart doesn’t care if he’s nice to his family or his pets, it cares about what did he do when he was in the seat of power?

      We’re both agreed that Cromwell was a hard man in a hard age, and the recommendation is that Mantel doesn’t try to soften his character to make him appealing to modern era sensibilities, but that she “does the trick that Renault does where she fits him into his context and makes him the most sympathetic figure within it”.

      My insuperable objection still remains that I heartily dislike and detest Cromwell. So – anybody read these? Any opinions?

      • Anonymous says:

        Read them! I don’t think she whitewashes Cromwell, and you can probably still continue in your hatred of them if you read them, and they’re just so good. (Although I had no strong opinions on More vs. Cromwell before I read the books and now absolutely love Cromwell in the way one loves morally problematic antiheroes, so my advice may not be worth much to you.)

        • AMac says:

          Yes, to Anonymous’s 1:08pm opinion. Mantel’s Cromwell is wonderful in certain respects, and monstrous in others. She tries to portray him as a person, and as an ambitious person of his place and time who chooses to ride a tiger. I think she succeeds.

          In a funny way, some of Mantel’s writing reminded me of certain Elmore Leonard novels, where the plot decrees that Something Very Bad happen to a minor character. A speed reader might skim right over the paragraph where the violence is done. Leonard seems to shrug, “Well, if you’re not paying attention…”

      • Gareth Rees says:

        Yes, read them. Mantel’s a subtle writer and the novels contain an (invented but) plausible psychological study of Cromwell. What could have motivated him? Why did he undertake such cruel actions? To be convincing this has to be sympathetic, at least initially. But as Cromwell carries out his monstrous projects: annulling the marriage with Katherine, destroying Thomas More, and fitting up Anne for adultery, no reader’s sympathy could possibly survive. Nonetheless, you start to understand the social and political framework in which these events took place. (Also, look out for the way Mantel hints at the role of Henry.)

    • whatever says:

      In my experience races with more than one head are usually duplicitous and generally untrustworthy. I’m not saying they’re all like that but there is correlation.

    • Hainish says:

      IDK about a favorite race, but my favorite gender is vegetable-transportation.

      • Multiheaded says:

        Whoa! Now that’s the reason I troll! Can you see my utility now, SSC commenters?

      • Douglas Knight says:

        If you want to promote that gender, you need a better name. How about “plant-plane”? For example, Lakoff turned “women, water, fire, and fighting” into “women, fire, and dangerous things,” which scans better.

        • Hainish says:

          Veggie-plane? Trans-veg? Orgo-mobiles?

          What I really want to promote, though, is the book, which is Through the Language: Why the World Looks Different in Other Languages, by Guy Deutscher.

        • Multiheaded says:


          Indeed, I like it! OK, doing something a little illegal but highly productive: here’s a full version.


          What are your views on high seas activities in the comments here? If it’s not OK, please notify me.

    • Scott Alexander says:

      I think I’ll ban you for a week, just because I’ve been meaning to for a while and this finally reminded me.

      • Alexander Stanislaw says:

        The ban hammer is no longer effective?

        I’m not sure if/why you think this will cause Multiheaded’s behavior to change.

        • Scott Alexander says:

          Yeah, I tried to ban IP, but I guess he has many of them.

          Now I’ve banned him by name, which should work, but also means any comment that speaks his name doesn’t go through. Oh well.

  14. Brian Potter says:

    The trolley problem gets a little closer to not being a thought experiment: Link

    Regarding the trolley problem, is anyone familiar with a construction that doesn’t rely on something that seems obviously untrue (pushing a man in front of a trolley will stop it)? As it is it seems to say a lot more about our intuitions for trusting third party information than it does about morality.

    • Alexander Stanislaw says:

      Have you heard of the organ donor one? Kill an organ donor to save many others who need the organs. You can make it a homeless person in a very remote hospital where the incident can be kept a secret to increase the plausibility.

    • anon says:

      Lifeboats with limited room, or cannibalism with limited food.

      • Those don’t *quite* map to the trolley problem because they are of the form “subset of group A dies as opposed to all of group A” whereas the trolley problem is “group B dies instead of group A,” which is the part that makes the thought experiment feel problematic.

    • Gavin says:

      You might be able to do something with blood transfusions. Sacrificing one O Negative patient could provide life-saving transfusions to the other ten passengers in the bus that crashed in remote appalachia.

      Now I need to go take a shower and nuzzle a cute bunny.

    • Anonymous says:

      Shooting down an airplane that is about to hit a building. See

      • Ken Arromdee says:

        That doesn’t count unless the people in the airplane would survive if you don’t shoot down. Since the airplane is about to hit a building, that probably isn’t true.

        • Douglas Knight says:

          You say it doesn’t count because consequentialists find it a very easy problem. If you want a problem to produce purity contests between consequentialists, it fails. But not everyone is a consequentialist! A better project is to produce tension between consequentialism and deontology. In particular, the Constitutional Court did not find this an easy consequentialism exercise, but rejected the law.

  15. I took a page from your book and designed an unrealistic political system. The basic idea is to use clustering algorithms to pick the most representative citizens. I’m eager to hear SSC readers poke holes in it.

    • anon says:

      I like your idea because it’s creative. I especially like the way you laid out your idea, it was very systematic and step-by-step.

      However, I do see it some problems. But, don’t let this criticism discourage you, because your idea definitely got me thinking and learning new things. Keep making more ideas, or try to fix the flaws in this one. Not all of what I have to say is critical, also, I have some neutral comments and questions, and positive compliments as well.

      I’m going to call your system Statsism rather than Statism, because although I like the pun I don’t like the confusion it might cause.

      Arrow’s Impossibility Theorem has never seemed important to me. If all candidates have an equal number of citizens who like, dislike, and are neutral towards them, who cares? It’s a three way tie, so a nonresult is exactly what the system should yield. Since ties never happen anyway, it’s not important, just a dumb theoretical nitpick. I don’t think your system avoids the problems with AIT – the clustering algorithm would deal with similar ones when selecting a representative human, I think.

      What if a moderate is objectively worse than an extremist of any type? Or, what if a pure moderate is worse than a slight moderate of any type? These would be plausible situations in which statsism would be worse than democracy. I’m not sure whether or not it’s true that being moderate is bad, but if it is then statsism is definitely bad.

      I’m not sure you’ve exhaustively listed all benefits of democracy. Specifically, I think that civic engagement deserves more attention. Statsism seems likely to reduce civic engagement, which is bad in itself but also indirectly – since if there’s a 99.999% chance I’ll never be influencing power, I probably won’t bother paying attention to politics or economics, or bother participating in them. That implies statsism reduces the quality of its candidate pool. Conversely, if people under statsism do pay attention to politics or economics, then they’ll be just as focused on signalling as before, won’t they? Maybe moreso, because their signalling won’t actually cause any bad policies, they won’t feel responsible for their rhetoric if they no longer are voting.

      There’s a difference between having a representative person elect a qualified official and having a representative person serve as the official themselves. The representative person probably isn’t capable of running a country or state. Democracy in America was set up by Founding Fathers who were terrified of the stupidity of the common person, it’s a very error averse system, and your system eliminates a lot of the safety checks in place by putting a representative person directly in power, rather than having the representative person elect someone more qualified. Being good enough to run a country is different than being good enough to choose which of two preselected candidates should run a country. Hypothetically, representative people could be good enough to do both tasks. But I’d like to see what makes you think so.

      (You could hypothetically have the representative person be the sole voter, but it’s unclear who their candidates would be. If their candidates campaign for the spot or do something similar, the same problems in current democracy just reappear. One possibility is to have no official candidates, the representative voter simply elects the nicest/smartest/richest/prettiest/best person they know. This has some potential, it’s very intriguing. However, what if people just elect their loved ones or bosses into the position, in return for emotional or monetary compensation? Or they just elect someone who they can control like a puppet? Seems ultimately unworkable.)

      What stops the representative person from treating themself and their friends like a king once they win the statsism lottery? They cease to be representative once they’re in office, really. “Accountability” is our current democracy’s check on this kind of selfish behavior, but that doesn’t apply if no individual will ever be in office more than once. Your system would collapse into a series of pseudo-dictatorships, wouldn’t it? (Or am I simply confused? – I think I might be, since I’m currently uncertain why normal presidents don’t behave in such a blatantly corrupt manner during their second term. Please help me here, by explaining what I’m doing wrong.)

      I don’t think that a computationally average person will make the same decisions as an interacting group of distinct individuals. If democracy benefits from swarm intelligence or wisdom of the crowd or similar processes, which it plausibly does, then your proposal will suffer for it.

      You don’t discuss how local politics would function. Would they function similarly? Or would local politics work differently than national ones?

      You know more than most social scientists about computers. Please continue to apply insights from programming to political or economic questions – the basic metaphor of information transfer is insightful, accurate, and brings a unique pragmatic perspective along with it. Thanks for the ideas.

      • I appreciate the kind words and the thoughtful response.

        I see AIT as a (not the) problem because the only equilibrium is two parties competing for the median voter in action and extremists in their words. If a third option tries to replace one of them, it helps the party it is most unlike. That’s not the worst thing, but it seems suboptimal. I’m being admittedly Americacentric here because I don’t know much about how this problem is reflected in other countries, but it definitely seems suboptimal in a way statsism isn’t.

        If moderates are bad, then statsism is bad. I would think statsism would prime people to signal less, but I might be wrong there.

        Statsism seems likely to reduce civic engagement, which is bad in itself but also indirectly – since if there’s a 99.999% chance I’ll never be influencing power, I probably won’t bother paying attention to politics or economics, or bother participating in them.

        One could make the same argument against Democracy. There’s a 99.999% chance you’ll never be the median voter, so there’s no reason to bother paying attention to politics or economics. I think that’s a legitimate concern in both systems. In the status quo, the void is filled with signaling. In statsism, the void might be filled with inaction. As you say, this is likely bad.

        I didn’t realize until I read responses to my paper that most people see accountability as a big deal to democracy. Perhaps it’s because I grew up in a state which treats political corruption as inevitable and expected, but I don’t model politicians as caring as much about reprisals from voters as they do reprisals from federal corruption investigations. Why fight corruption through the vague threat of “accountability” when you can just make corruption illegal and treat those who act corruptly as criminals? To some extent, we already do this (topically, see the indictment of lame duck Rick Perry), and we could do this under statsism as well.

        I don’t think that a computationally average person will make the same decisions as an interacting group of distinct individuals. If democracy benefits from swarm intelligence or wisdom of the crowd or similar processes, which it plausibly does, then your proposal will suffer for it.

        I’m going to think carefully about this and not respond off the cuff.

        After reading your response and others, I’m wondering if there is a more modest application of this idea. Perhaps both parties could be convinced to use a consistent clustering algorithms to fairly allocate district areas when they have the ability to do so, and if either party deviated from this, they would be publicly shamed for doing so. I’m not sure what would constitute a fair district algorithm, but I’d enjoy designing one that met criteria.

        • Alexander Stanislaw says:

          Its a very cool idea, I do love ideas that aren’t bound by the shackles of practicality. The core of the idea (if I understand right) is a to achieve a better representation of the public interest in government. The government would make policies as a true moderate would.

          I do think, though, that a true moderate without outside influences might be much worse than what we have. A true moderate in the US is a creationist who thinks that anti-American speech by Muslims should be banned.

    • Toby Bartels says:

      I am somewhere above 95% sure that I read a science-fiction short story with this premise.

      In this story, polling techniques had advanced so far that pollsters could predict the outcome of any election before it happened, so that there was no need to actually run the election … ha, no, that scenario makes too much sense! Actually, polling techiques had advanced so far that pollsters could identify a single person who would vote whichever way the majority would vote, even though for some reason they could not predict how this person would vote. So instead of having an election, the authorities just asked that one person. This is the variation of your idea in which the Exemplar votes for the office-holder (in the story, it was the President of the World or some other big executive position) rather than being the office-holder. And somehow there were still just two candidates.

      The story recorded the thoughts of the main character as he (above 90% confidence that it was he) followed the news, watched the demographics centre upon his area, discovered that he was the Voter, dealt with the subsequent celebrity, and agonized over his decision. In the course of the story, it became clear (without any need to say so) that the process was irredeemably corrupt, but the main character refused to be corrupted and took his job very seriously. Nothing in the story made us care about the candidates, however, and I believe that it ended before he had decided.

      • lmm says:

        I remember that story. It was by Asimov. I did not think the process was obviously corrupt.

        • Toby Bartels says:

          Well, you probably remember it better than I do, since at least you remember who wrote it. But I recall something about people giving him (after his selection and before his vote) congratulatory gifts that he ignored. Maybe they weren’t thinly disguised bribes, and maybe he ignored them only because of his depressed mood borne from feelings of inadequacy (which is my chief emotional memory of this story), but I remember thinking that allowing people to communicate with the chosen voter after the selection and before the vote was a big flaw in their electoral system.

      • Niall says:

        The Asimov story is Franchise:

        It’s similar enough that it rung the same bell for me as for lmm, but different enough that there might be another story closer to yours.

        • Toby Bartels says:

          Yes, that’s it!

          One difference, then, is that the selected Voter of the Year doesn’t directly choose candidates but actually gets to vote on the issues, so that’s nice.

    • Toby Bartels says:

      Also you should read about sortition (Wikipedia). This is how most offices were filled in classical Athens. It’s basically a randomized version of your system, where we give up on trying to measure how typical people are and instead trust everything to work out correctly on average over time.

      • Paul Torek says:

        Seconded. Sortition gives extremists a little more hope than statsism, because there is always a chance that this year’s legislature will be biased in the direction you favor. And when it is, the legislation you favor will pass and everyone will see what a good idea it is.

    • Rob says:

      Very interesting! The main problem I see is the assumption that people will fill out the questionnaire honestly. Seems like they’d be incentivised to lie.

      If you want to maximise your chance of election, lie to aim for the very middle of a known cluster, then if you get in, implement whatever your actual preferred policy is.

      If you want to change policy without being elected, take the most extreme position that will still associate with a cluster, to drag that cluster’s centre towards policy you like.

      Maybe not those specific examples, but it looks to me like citizens are not incentivised to be honest at all, and will distort the system with tactical voting.

    • lmm says:

      What are you comparing to? The clustering algorithms for papers have a large corpus as baseline and cost the sentences that make this paper different from that, right? So would you end up just selecting the “most American” citizens, those who were most different from people in other countries? That might not be ideal.

      I share Rob’s concern.

      • Toby Bartels says:

        There’s one part that I found interesting (as did the author, apparently, since it led to the title): the idea that selecting an average person as leader instead of the best person (with all previous political debate being about how to choose the best) is unheard of in the history of thought going back to ancient Greece. In direct analogy with non-Aristotelian logic, he terms it non-Platonic politics.

        But in fact, Plato’s insistence on choosing the best rulers was deliberately subversive against the Athenian democracy, which filled most offices by lot. It’s not as self-evident as it appears.

  16. Andy says:

    This is perilously close to gender, and so I fervently hope this doesn’t spiral into a This Is Why We Can’t Have Nice Things flamewar, but here goes:
    I wrote an explicitly porny fem-slashfic that mentioned safe sex, specifically dental dams. Several readers from various parts of the world commented that they’d never heard of dental dams, and thanked me for indirectly teaching them. Given the horrifying (if you’re a liberal) state of sex education in much of the world these days, it’s probably safe to say that the Internet is a better resource than the schools. And fandom has a very broad reach, connecting people from many different countries and cultures, reaching into places and people that more overt Here Is How To Have Sex And Not Die campaigners might not reach. And people do have their heroes – if Captain America insists that Iron Man use a condom, it might make it easier for some scared kid from an Indiana cornfield to insist that his partner use one.
    So yesterday, I created a Tumblr called The Safe Slash Project to disseminate information across fandoms, and highlight slashfic and fanart that depicts safe sex practices. If any of you are active on the NSFW side of General Internet Fandom, I’d appreciate a signal boost. If you write or draw explicit works depicting safe sex practices, I can reblog them and (maybe, someday) get them to a wider audience.
    Thanks, and sorry Scott if I have sinned against your comment thread!

    • Not My Normal Name says:

      If you’re trying to increase fic writer’s knowledge, it might also do to reiterate firmly that cervical penetration is both NOT NORMAL and A SERIOUS MEDICAL ISSUE, GO SEE A DAMN DOCTOR.

      I have no idea how often I’ve seen it but every time I do it kills part of my faith in the future of humanity. People talk about porn giving unrealistic expectations but that’s one of the few I’ve seen with the potential to seriously injure someone.

      • Andy says:

        That I was aware of, but I am no kind of medical professional (dammit Jim, I’m a mapmaker, not a doctor!) and I’ve heard similar horror stories about anal sex. Do you have an Internet resource mentioning this and giving tips for how to notice/avoid cervical penetration that I can put up?

        • anon1 says:

          This isn’t a thing that happens, right? Just a mistake that appears in fanfic?

        • Andy says:

          From a cursory Google search, with very little desire to go deeper, I’d say it’s a thing that’s shown in porn (somehow) and then people try to imitate in meatspace.

  17. Vamair says:

    Scott, I wanted to ask if you’d allow me (or for that matter any other person) to translate some of the posts (or something like Consequentialism FAQ) and then post them somewhere, with all the proper credits and links. The posts are great, but most people I’d like to show them can’t read English.
    If you would, then would you let me change some minor detailes (mostly historical examples to a sufficiently similar ones) iff I’m reasonably sure the changes don’t affect the idea of a post and I believe people are going to get mindkilled about these examples and are going to miss the big idea because of that?

  18. Shmi Nux says:

    Please consider installing a comment upvote plugin… This one should take maybe 5 min to get going:

    • Anon says:

      Allow me to voice strong disagreement. Comment voting is a tool for establishing community consensus. Discussion is more diverse and therefore (to me and I think a lot of people here) more valuable without that.

      • Zathille says:


        No, the irony of doing so is not lost on me. But I feel strongly on the matter.

      • The Anonymouse says:

        It may be a tool for creating community consensus, but it is also an immensely valuable tool for those of us who sometimes don’t have time to read threads that spring into the hundreds of comments in a day. One of the reasons I come here is I trust the taste of the commenters, and if the community consistently upvotes a comment, that’s a strong signal that it’s worth reading when I don’t have a free hour+ to read all of them.

      • Darcey Riley says:

        Thirded. The possibility of my comments getting downvoted would substantially decrease my probability of posting them. (An upvote-only comment system might be ok; for some reason the lack of approval isn’t nearly as bad as explicit disapproval would be.)

        • Ialdabaoth says:

          Rather than an upvote-only or upvote-and-downvote system, I might approve of some kind of “like” system. This also has the advantage of lending itself to clustering analysis.

        • Nornagest says:

          I violently dislike upvote-only systems, especially if they have any algorithmic (i.e. sorting) significance. They do the job of providing positive feedback, but the incentives they create ultimately point in the direction of what’s essentially clickbait. And I really don’t like clickbait.

          It works a little better if the upvotes are limited, though: if you only have a small number of blessings to bestow, most people won’t waste them on One Weird Trick (Invented By A Mom). Political/identify-based clique formation is still an issue, though.

      • nydwracu says:

        One of SSC’s major advantages right now is that it’s not LW. It’s probably a bad idea for it to become more like LW.

        The clustering failure mode seems like it would be much more likely here than LW, since there’s no politics taboo. (One major risk is that the comments section ends up giving off even stronger NRx signals and triggering memetic immune reactions even more frequently.)

        • Multiheaded says:

          Hey, an idea. Why don’t you, with all your interesting thoughts about thede and elthede, make a design proposal for a system with two entirely different kinds of upvotes, ingroup and outgroup.

          The way I see it is like this: before voting, everyone has to register and pick two or three, no more, tribal components for their identity (and migration is limited in some clever way) out of some list hashed out in a grand big thread. For example, “LessWrong + Islam”, or “Transhumanism + Rad-Fem”, or “Islam + Deep Green + Neo-Confucian Technocracy”. And then everyone, no matter their thede, can tag their posts as open for upvotes from any combination of tribes – and people who are registered as part of those tribes (can’t have bots, must limit alts) can contribute to those specific karma meters. And in addition, there’s the outgroup karma, which can be given to any posts not tagged with any of your tribes, to show appreciation for a different way while signalling something different from in-group approval.

          So you can see how this is intended to turn clustering from a flaw into a mechanism. This probably has all sorts of horrible bugs and exploits. But you seem to think about such things a lot. Any comments?

        • Matthew says:

          Really, Multi? You want to reify people’s differences? What happened to keeping one’s identity small?

        • Multiheaded says:


          I just happen to agree with some of the neoreactionaries on this one (although, like with everything else, they didn’t invent it) – as well as note that the general principle – “It’s here to stay, either regulate it, or deny it and suffer” – is applicable in many spheres (although sometimes invoked mistakenly or with ulterior/malicious motives).

        • nydwracu says:

          Moldbug proposed a wiki along those lines. (But maybe that’s what you’re referring to…?)

          That might be worthwhile for some (much larger) sites, but probably not here — it would be best to keep divisions in the background, to avoid the breakdown of the overarching identity that fosters civility and inter-group exchange of ideas and so on here.

          Reifying divisions reifies divisions, and thedes like the currently-nameless-and-only-approximated-by-labels-like-‘rationalist’ LW/SSC-sphere are rare enough that it would be best not to fuel their fracturing.

      • Nick says:

        Fourthed. My reasons for not liking voting systems are Nornagest’s, with one more complaint: time and time again I see an interesting discussion in which both or all people get something out of it and which clarifies a bunch of things, except that one person or side of it is getting the vast majority of upvotes. It’s extremely demotivating to participate in a discussion that is clearly fruitful but in which only one side is being silently praised.

      • Matt C says:

        Please nth the preference for NO voting/karma system.

        I am morally certain (i.e. asserting without proof) that introducing a karma system will change the type of discussion here, with a slide toward comments written just to gather upvotes. Even comments that are intended as actual conversation will be written with half a mind toward how they’ll get scored.

        A voting system means SSC acquires a new layer of stupid meta complaints and discussions about how other people voted.

        A voting system will encourage voting factions where people aggressively vote for “their side” regardless of actual content quality.

        Let’s just skip all that, please.

    • The Anonymouse says:

      This would make me very happy.

    • The major advantage of a comment voting system, in my view, is that it lets people find the really good cream-of-the-crop comments that are especially worth reading in their own right. The absence of anything that does this on SSC is somewhat unfortunate, but as others have observed a voting system carries serious costs. Can anyone think of anything else that might serve this function?

      • anon says:

        Memorize the names of certain quality commenters, and control+f. Or, scroll through the page very quickly, assuming that the most interesting discussions are those with the longest comment chains. Seems workable enough.

        • Multiheaded says:

          For that to be so, Scott would need to come through on his promise of banning the less worthwhile half of reactionaries.

          I think we can all agree that people like Troy or Nydwracu are very cool and should be enticed to contribute more with their enlightened perspective. However, many of us might also say that people like pwyll or Steve Johnson add nothing to the discussion (as their ideological views, unlike, say, mine, already have excellent representation here) while detracting from the quality of life and posting a great deal.

          P.S.: that last part, “posting a great deal”, may be read both ways.

        • Andy says:

          This is roughly what I do.

        • Douglas Knight says:

          The longest discussions are the result of revealed preference, revealed interest. But are they ones a reader should seek out? Upon reflection, they seem to me to be the result of addiction, not interest. I am in favor of the current strict limit on threading for this reason. Maybe LW’s hiding of deeply nested threads is good for this reason.

          At first I found Multiheaded’s comment a non sequitur. But maybe it is supposed to be close to mine: eliminate flamers and long threads will no longer be flamewars, but actually interesting.

        • I just realized where I got my idea about the purpose of an upvote system: from this comment. Yes, the parent of that comment is a pun rather than a keen political insight, but it was still so incredibly awesome that I wanted to give the author some kind of award. More importantly, I very nearly didn’t see that comment at all.

          I haven’t found really long digressive comment threads here to provide me with all that much insight compared to a really good single post or comment.

    • Anonymous says:

      I often find myself wishing I could upvote or downvote comments … but honestly, I feel like the current “report” system is better, much to my surprise.

      Perhaps a “gold” or “favourite” system, that promotes the comment, and that’s all?

      • Nornagest says:

        Honestly, the “report” system feels a little too much like casting opinions into the void for my taste. Especially since all the moderation is being routed through one very busy guy.

        • potatoe says:

          You say “casting opinions into the void” in a way that suggests this is a negative thing. But what if you are sending them on a magical space adventure? It’s all in the phrasing.

        • Alexander Stanislaw says:

          Wait what? The report system is for commenters that repeatedly violate the comment policy (or egregiously violate it once). Its not a disagree button.

        • Nornagest says:

          I’m not saying I report for disagreement. I’m saying there’s some subjectivity in what to report (“opinion”), and no feedback after reporting (“void”). Does that clarify things for you?

          That said, comments that egregiously violate the comment policy (which I operationalize as “not true or useful, and over a certain threshold of rude”) seem surprisingly common.

      • jaimeastorga2000 says:

        @Alexander Stanislaw: Something tells me Moloch is at work here.

        If some people report other people who disagree with them, and others don’t, and each reported comment carries a tiny probability of banning…

    • Toby Bartels says:

      I want a system in which there is a Like button (or +1, thumbs-up, whatever) on every comment.

      And that’s it. No running total, no record of who’s voted, no sorting comments, in fact no consequences whatsoever. Just the button.

    • Tenoke says:

      Relevant reply from Scott on the last OT

      Maybe you can ask him for a (simplified?) version of his wordpress installation, so you can test the plugin and see if there are any problems. I suspect that if you try something like that, he will be okay with trying it out.

  19. Ergoemos says:

    Anyone know of any good podcasts in the same vein as this blog? I am a very relaxed rationalist (finished only part of the legendary Sequences) but I love reading stuff like this blog. At work, while playing around with spreadsheets, I like listening to podcasta too.

    Any good rationalism themed podcasts out there?

    • Jai says:

      I’m a big fan of the Methods of Rationality Podcast, which started off as an HPMOR audiobook but has since branched out into readings of other stories and essays. There’s no original content (yet), and I’m not sure if Eneasz plans to continue after HPMOR is done, but I hope he does.

      Rationally Speaking is co-hosted by Julia Galef, president of CFAR, and produces some pretty good discussion. It’s not SSC-level, and I am not a huge fan of Massimo, but it’s about as good as you can get for a rational podcast right now (as far as I know).

      Not quite in the same vein, Planet Money is an economics podcast from NPR which I think the SSCosphere would enjoy as light listening. Their series on the creation of a t-shirt is, and I mean this sincerely, fascinating.

  20. Kiboh says:

    Just putting this out there: I would be *very* interested in seeing a self-identified rationalist do a blind liveblog of Umineko No Naku Koro Ni. I think a lot of other people would as well.

    • Aleph says:

      Gwern has a review somewhere on his site. My general impression (not having played the game) is that it deliberately tries to throw off any attempt to reason your way through things, which makes it useless as a way of practicising/testing thinking skills. (It’s like me asking you “guess what’s in my pocket”, and calling this a “riddle”, and then telling you you’re bad at riddles because you guessed wrong.)

      • Douglas Knight says:

        Reading your comment made me wonder if Kiboh agreed and that was the whole point. If the purpose is to practice or test, there is little point in a record of the blind live reaction.

        • Kiboh says:

          I’m not *quite* that mean. My motivation is that when trying to figure out the truth in Umineko, rationalist techniques turned out to alternate between amazingly useful and amazingly counterproductive, so I think that someone playing Umineko from a rationalist perspective would end up being alternately impressive and hilarious in their attempts to deduce the solution(s).

          Also, Umineko is generally a fun thing to read liveblogs of. I’ve read two so far and it’s fascinating how their approaches diverged from my own: my “never try to beat the protagonist at his own specialities, focus on what he neglects and what you know that he doesn’t” vs Blog 1’s “always use the protagonist’s reasoning as a starting point, but try to extend it further” vs Blog 2’s “disregard the protagonist’s reasoning, just do your own thing: also, never end a chapter without SOME kind of theory about the new events”. And the outrageous plot twists are a lot of fun to watch someone get hit by.

    • MicaiahC says:

      I would also be highly amused if someone bought, played and liveblogged a session of the Umineko Board game:

      Incidentally, did you know there was an Umineko Board game available for sale?

    • lmm says:

      You should make it clear whether you mean visual novel or animé.

    • Ken Arromdee says:

      I’m currently playing through Higurashi (actually, I stopped after the first 4 chapters, someday I’ll do the rest). Trying to figure out things in Higurashi is generally useless for a number of reasons, including
      — characters hallucinating
      — sudden genre shifts making it unclear what hypotheses are reasonable in context
      — unclear relationship between the various chapters, and in particular, what we are permitted to assume about whether events in one chapter happen in another
      — a couple of what seem to be outright mistakes (though I can’t prove it until I finish the whole thing)

      Also, some characters seem to act too dumb to live at times, especially the main character.

      I did take some notes while playing, so maybe I’ll post them somewhere someday.

      • Fadeway says:

        Funny, I stopped Higu after four chapters as well. A year later I just marathoned the anime.

        There’s two ways to doing parallel routes in VNs, one where the worlds are entirely the same with a small change in events at the beginning of the route (FSN, Ever17), and one where the worlds are different – or at least the changes are numerous and in too many places to keep track – but made to look the same on the surface level. Higurashi uses the second type, which makes trying to combine information from different parts an exercise in frustration. It also had one or two parts which made little sense even after all the reveals, although I’ve long forgotten which.

        In thirty years when my memory is hazy, I hope to do a full reread with theories on a blackboard and all.

    • a person says:

      I might do this, it looks like a kind of interesting game.

  21. Gareth Rees says:

    I request that people not tell me things like “Why are you wasting your time writing about issue X? Aren’t there more important / less politicized things to talk about?”

    Daniel Davies says, “I have always worried that people who make this kind of argument will end up going to the toilet and then thinking ‘people are being killed in conflicts all over the world every day, and all I’m doing is wiping my bum’, and then not wipe their bum.”

  22. Jordan D. says:

    Just a few things I feel like saying:

    1) I want to thank you for your diligent blogging- your posts are always well-considered, fun to read and feature interesting takes on novel topics.

    2) I dunno if last post’s link to Popehat means that you’ve been doing an archive-binge or came across that post on Reddit or something, but I’m always happy to see one of my favorite blogs mentioned on another.

    3) You’re probably already aware of it, but Lowering The Bar is an indispensable resource for hilarious things, like the current story about the Ohio man who remains legally dead. I think you’d also like the short parodies of classic stories which Kevin Underhill wrote, here.

    4) Anyone who happens to be looking for a really good older-style RPG should take a look at Divinity: Original Sin. The over-arching plot is a bit predictable, but the foundations are solid.

    5) This.doctrine tickles me for some reason.

    6) Last year I decided to start a Pathfinder Mythic Gestalt game, and my players are at level 15 now. The mechanical soundness of the game broke down months ago, but it has left some truly strange and beautiful things behind. I recommend the experience to anyone who can withstand seeing a character with a class/archetype listing longer than this paragraph.

    I hope all is going well with everyone!

    • Scott F says:

      With the gaming mega-thread firmly convincing me that many people will benefit from this, I’m going to also recommend Divinity: Original Sin despite only being halfway through it.

      As mentioned, the plot is standard, though well-executed. But the characterisation, the world-building! The world is tastefully crammed with oddities, the NPCs and your conversations with them are on par with Baldur’s Gate II and Planescape Torment, and the sense of humour is consistent and very true to that peculiar RPG style.

      The combat is where it really shines, though. They make a big deal out of “using the environment”, and they deliver: environmental manipulation is available to most character builds and is something you can use to your advantage without being a cheap and easy victory, and carelessness can put you at a disadvantage without being immediate defeat. Working to minimise the effects of the environment is just as effective as working to maximise them.

      It’s good! You should play it!

  23. Shmi Nux says:

    Further regarding voting: both those who oppose to and those who like it have excellent points, but these points are generally independent, not mutually exclusive, which makes me think that it ought to be possible to design a feedback system which satisfies all or most of these.

    Arguments for:
    – show (at least to the author) appreciation of a comment you like
    – make the best comments easier to find

    Arguments against:
    – reduces diversity by explicating most popular opinions
    – a possibility of being downvoted is a strong deterrent for many
    – a possibility of gaming the votes given that this blog allows anonymous replies
    – other expected and unexpected ways any metric can be hijacked (Goodhart’s law)

    In the anti-rational spirit of offering solutions, here is one way to satisfy those:
    – “likes” only, no downvotes
    – keep them invisible (except possibly to the liker and likee).
    – show, say, top 5 comments (without the detailed vote count) once, say, the post has more than 50 comments and only if the comment has, say, 10 likes or more.
    – use cookies to limit vote gaming by anon upvotes (imperfect, but I am not sure what else can be done without adding trivial inconveniences).

    Most of these are solved problems, but I am not aware of a single WordPress plugin which does most of that, so there is little chance of this blog gaining a meaningful feedback system.

    • Multiheaded says:

      See my proposal for a proposal, above.

    • Nornagest says:

      This doesn’t seem to solve the clickbait problem; it just delays it until posters can hill-climb their way into it rather than directly deriving what gets votes from observation of others.

    • Matt C says:

      > make the best comments easier to find

      If you are daunted by a 300 comment thread, I have some sympathy. But I don’t agree that setting up a Like system will actually find the best comments.

      > there is little chance of this blog gaining a meaningful feedback system.

      It already has “Reply”, which is meaningful feedback for people having a conversation.

    • Auroch says:

      If there was a liking system, I would want to show these things and no others:

      Show the liker which things they have liked. (Possibly in a way that let them find comments from the same people quickly, possibly on on the comments themselves.)

      Show the likee that they have been liked. (Possibly including who by, but probably not.)

      Show a point of comparison for how many likes are typical (Possibly including several measures of average, possibly a full distribution.)

      This would be optimized to grant warm fuzzies to respected commenters, satisfy the desire to reward commenters you respect, and minimize the degree to which it would encourage signalling and behavior distortion.

  24. Adam Casey says:

    There’s a very annoying trivial inconvenience that can be fixed without a voting system. The conversations here would be a lot more pleasant to read if you could collapse a particular thread rather than scrolling over a five-mile-long discussion to get the next topic.

    5 seconds of google failed. Anyone know if there’s a wordpress plugin or similar for this?

    • Vulture says:


    • Bakkot says:

      Dunno about wordpress, but I just updated my script to add that feature, so you could use that in the interim. If you already had it installed, and are not using the Chrome extension, you will probably need to reinstall.

    • princess Stargirl says:

      This would have made reading the comments on Scott’s more popular popular/controverisal posts much easier. Great idea.

  25. Sophronius says:

    Ooh, open thread. Good opportunity for me to express my annoyance at reading comments such as the ones in this old thread:

    On why people think rationalists are less emotional: “Okay, he is more like a computer than I am, because he can come up with associations and knows more of things, I’ll bite that. What can I do that a machine can’t? Okay: feel, have emotions. Hence, emotions are good and whoever does not feel them is a loser.”

    So it is a very common trope that people who are more logical are less emotional. That would be an actual weakness of rationalists, alarming if true. So how do we respond to this? Make fun of it of course! And for good measure, speculate on the motives of whomsoever would dare to criticize rationalists. Anything other than, you know, consider for 5 goddamned seconds whether or not the criticism might actually be valid.

    No seriously. If everyone observes that rationalists are less emotional, isn’t the most obvious explanation that it is simply TRUE? Because I very much notice that being able to control your emotions is essential to good rationality. People who get angry at hearing certain views are almost never rational. And conversely, people who are very smart and rational tend to be really bad at reading other people, in my experience. Just think about aspies. There definitely seems to be a trade-of here.

    • Vulture says:

      Agreed on most counts, although I do think I’ve seen some other people in the rationalist community notice this. Is it really a problem, though? Unless of course we hold emotional experience as a terminal value.

      On a probably related note, I’ve seen it observed before (I think on Less Wrong) that people in the rationalist community tend to be very low-empathy. This one seems like a much more serious tradeoff, since I would bet that low empathy is a big risk factor for doing (and being able to rationalize) various evil things. Since consequentialist-thought-experiment scenarios are probably less common in real life than regular ol machiavellian opportunities, I think I would accept a pure general-empathy boost if I was offered one.

      • Sophronius says:

        I see it as much the same thing. Emotion/empathy/intuition on the one hand vs. logic/cleverness/rationality on the other. It’s not that rationalists don’t have emotions, they are just bad at dealing with them.

        The main problems I see with the failure to recognize this trade-of are:
        1) People fail to realize that rationality requires being able to tell your emotions to STFU for a moment. This is why people on Less Wrong cannot discuss politics rationally, for example.
        2) Focusing far too much on ‘clever’ solutions to problems while failing to take ‘common sense’ into account.

        I definitely would accept a general-empathy boost, on account of more understanding being pretty much always better. That said I very much favor consequentialism.

        • Anonymous says:

          I’m not so sure about wanting an empathy boost. I think a lot of effective altruists are low-empathy, probably because high-empathy people are more likely to just give to the Charity for Cute Puppies with Terminal Diseases or whatever triggers their empathy most.

        • ozymandias says:

          Counteranecdote: when my affective empathy increases I am more effective altruist, because I am like “gosh! People are suffering! I should figure out how to fix that!” When it is low, I am more like “trying to fix things sounds haaaaard.”

          That said, lack of affective empathy is good for some things. I’ve noticed that I’m better at dealing with sad people when my affective empathy is low (if I have some other reason to try to comfort them, such as sense of duty), because not being sad when other people are sad makes it easier for me to note what their needs are and try to fulfill them.

        • Liskantope says:

          I think the ideal rationalist would experience as much emotion as the average non-rationalist, but would know how to manage them by letting their reasoning processes direct their emotions, rather than letting their emotions direct their reasoning processes. I agree that the ability to feel more empathy should enhance one’s capability to reach rational conclusions. That said, I can relate to what Ozy is saying in that I’ve found that my generally high capacity for empathy sometimes seems to inhibit me from acting in the most rational manner. Even when I’m able to attain the most rational viewpoint concerning a social situation when dealing with it in the abstract, that doesn’t mean that I’ll find it in me to act on it.

          I do notice a correlation between rationalism and lack of empathy. I feel that only a few of my the people I know in real life come across as particularly rationalistic in social affairs, but that those same people are overall much less compassionate than my friends who are less rationalist. One of the main reasons I like reading Scott’s blog is that I feel that he strives for a very high level of rationalism while at the time making a point of being very sensitive to other people’s feelings.

        • Auroch says:

          Liskantope: For pure personal anecdote, I found that actually thinking through the consequences of empathy beyond a limited circle and what that meant — something like the skill of ‘being convinced by empathy’ — is enough to make a thinking person seriously doubt the value of empathy and try to avoid having it.

          Also, time spent learning cool things as a child which lead to desire to be rational is usually time not spent learning social skills or developing empathy or empathy controls. Which is less conclusive but certainly more general.

    • Nornagest says:

      The usual failure mode of nerds dealing with feelings (and the one being alluded to in that XKCD) isn’t being too unemotional, it’s being more emotional than they think, or act like, they are. The nerd sees others openly motivated by emotion and (correctly) observes that this involves a lot of drama, bias, and hardship. They then cross-reference against their subcultural norms and touchstones, especially a respect for intelligence and logical argument above all else, and (wrongly) conclude that the expression of emotion is causing all the trouble and it’s got to go. Over time they’re likely to get pretty good at appearing hyperrational, and to be pretty smug about it. But the underlying feelings are still there, and they tend to come out in covert and inevitably more sinister ways.

      A related issue is that being unable to recognize feelings looks just the same, from the inside, as not having them in the first place, and through inexperience or contempt or various organic problems nerds are often bad at emotional modeling. This can combine with the above in some pretty pernicious ways, and the two reinforce each other. It’s common, for example, to meet people in the poly scene (which tends to be pretty nerdy) who claim not to have the jealousy patch installed, and get very upset at the suggestion — but nonetheless act if so provoked in ways which, to be charitable, are nearly indistinguishable from bog-standard jealousy.

      tl;dr: Acting like a computer is totally peachy if you are, in fact, a computer. Very few people are.

    • nydwracu says:

      Bottom-up stereotypes vs. top-down stereotypes.

      “Stereotypes are usually true”: if this is about accumulated cultural wisdom, ordinary people consistently noticing things and crystallizing the patterns they notice over time, this should only apply to bottom-up stereotypes.

      This would predict that, if top-down stereotypes are true, it’s because they were made to become true after they were called up — or because they originated through the same process as bottom-up stereotypes. (This is possible: how many terms did Tom Wolfe coin? Tom Wolfe is pretty observant.)

      Is “rationalists are unemotional” a bottom-up or a top-down stereotype? Probably top-down: unemotional robots in science fiction, then Spock. (Vulcans are complicated in interesting ways here, but TNG has Data instead.)

      Wouldn’t expect it to have originated through bottom-up-like processes either: SF doesn’t need to tap them.

      • Douglas Knight says:

        I don’t know how important it is to your argument, but the image was well-established before Star Trek. Feynman always complained about people emphasizing his drumming. In particular, his famous quote (“…perpetual desire to prove that people who do it are human…”) is from 1966, the year of the show’s debut.

    • Liskantope says:

      I don’t think the fact that people with Asperger’s are both very rational and bad at reading emotions is part of an overall trend of rationalists being less emotional. In my experience, aspies have emotions every bit as strong and sometimes irrational as those of non-aspies; their level of emotion has little to do with their difficulties in interpreting the emotions of those around them. I’m not even sure that aspies are overall less empathetic; again, this is not implied by impairment at interpreting others’ feelings accurately.

  26. Robert Liguori says:

    Hi, everyone. Long-time reader, first-time poster.

    So, I’m in the middle of reading Smartest Guys In The Room, a book about the rise and fall of Enron, and it reminded me heavily of the Living By The Sword post from a few months ago. One thing that positively leaped out at me was that the exact same negative traits that fueled Enron’s growth (a focus on short-term versus long-term rewards and a deliberate cultivation of ruthlessness and disdain for cooperation) was also responsible for its downfall; once the decline started, it was vastly accelerated by the presence of bad actions within Enron, who proved just as willing to defraud the company as the company had been willing to defraud the world.

    I don’t know of any cases of actual whale tumor biopsies to discuss the point biologically, but I did want to bring it up as a data point, and an open thread seemed the best place as any to do it.

  27. I am working on something very exciting and of potential interest in the crowd here, but I’m very conflicted about revealing what it is, since that would entail revealing my True Name, which will give all of you the power to do sorcery to me.

    Alternately, I would contact SA and ask him to help me out with something, but above it’s been suggested that he doesn’t read emails, so…

  28. Darcey Riley says:

    Hey Scott, I have a psychiatry question for you. Is there any psychiatric disorder where people have trouble distinguishing between reality and dreams? In particular, is there a disorder where people recall events, and then can’t figure out whether those events happened in reality or in a dream?

    I ask because this happens to me occasionally. Like, the other day, I dreamed that a friend had updated his blog, and when I visited his blog in real life, I was surprised to find that it hadn’t actually updated after all.

    Assuming this is a fairly normal experience, I would expect there to be a pathological version where people can almost never distinguish whether their memories come from reality or dreams. But I’ve never heard of such a disorder, and so I assume that if it exists, it’s extremely uncommon.

    Am I the only one who finds this surprising? For some reason, I would have expected a disorder like this to be one of the most common psychiatric problems.

    Edit: I just looked up your very convenient post on sleep disorders, and it says that this happens to narcoleptics. My new question, then, is why does no one ever talk about this? I mean, it would make such good science fiction and epistemological horror! Why is this not part of our culture’s standard conception of mental illness, the way that something like schizophrenia is?

    • Emile says:

      FWIW the closest thing I had was a few times where I would be confused about whether something happened up to a few minutes after waking up – so that during breakfest I’d think back “of course I didn’t get yelled at by Batman yesterday, how could I have been confused about whether that was a dream or not?”.

      Never had it persist through the day tho.

    • Said Achmiz says:

      I’ve had this happen to me a number of times, but am not (to my knowledge??) a narcoleptic.

    • Scott Alexander says:

      I can’t think of anything except narcolepsy right now. Possibly some dissociative experiences might be kind of dreamlike, but they wouldn’t be dreams per se.

      I’ve had this problem with very minor things, especially shortly after I wake up, but nothing major.

    • Sniffnoy says:

      Just going by anecdote, I get the impression this (or this to some extent) is pretty common.

    • nydwracu says:

      I’ve had this happen a few times before — but only with the layouts of buildings that I haven’t been to in a long time.

    • Paul Torek says:

      Somewhat related, I once had the experience of a dream sorta-continuing after I woke up. The dream became a lot less vivid than they usually are, but still I was relieved when it went away. I agree about the sci fi and epistemological horror; I immediately thought that someone (with more talent than me) should write a fiction like this. Only, obviously, the dream doesn’t go away, at all, ever. Or until the end of the story. Or reality goes away. Or they both go away; movie fades to black. Or…

  29. Carinthium says:

    I’m doing a Philosophy of Science course right now, so I want to ask people’s opinions on a question.

    Obviously the science we have at present is flawed, but what are the major factors in making Science distinguishable from Non-Science? Given what Science and Non-Science are, what is the best way to make a definition of scientific behaviour?

    • jaimeastorga2000 says:

      Have you tried reading The Sequences?

      • Carinthium says:

        Science and Rationality are not the same time. The question is more about what it is about science which makes it superior to pseudoscience or non-science. That is not the same question as how to maximise one’s rationality.

        • Auroch says:

          Certain of the Sequences spend significant time talking about the successes and failures (mostly failures) of Science as She Is Practiced.

  30. Oligopsony says:

    Loving the new comment markers – I’m not going to embarrass myself with an estimation of how much time they’re likely to save per day, but I will hint that they are embarrassing.

  31. Matthew says:

    [Stipulating before continuing that Ozy and the other cluster B people in these parts appear to be good people.]

    I’ve mostly managed to suppress the reaction now, but I thought it worth mentioning that in the past, whenever someone would blithely mention in the comments that they have Borderline or another Cluster B personality disorder, this was actually a trigger for me for a mild panic reaction.

    I lost about a decade of my life to an emotionally (and toward the end, a bit physically) abusive marriage to a borderline, culminating in absconding with my children and lodging a false abuse claim against me. The connotation of borderline for me is “manipulative sociopath minus the high self-esteem.” It was absolutely bizarre to me to see people mentioning it and garnering sympathy from other commenters. To me it was hard not to read as, “Oh, by the way, I’m a monster.” “Oh, you poor thing.” I’m aware that many borderlines had abusive childhoods; it’s still weird for me that the prevailing association here is “victim” rather than “predator.” But I’m working on it.

  32. jivyfadni says:

    An interesting “traditionalist” Tumblr:

  33. Alexander Stanislaw says:

    Regarding genetic influences on behavior, IQ and everything else.
    there are two kinds of people – people who think that genes don’t matter except in extreme cases like genetic diseases, savantism, or athletic genes that small tribes have (like Kenyan runners), and people like myself who think that genes account for about 40% of the variance of most traits (less if there is environmental depression such as malnutrition). All of the theoretical arguments would take me far too long carefully process and it still might not resolve the issues. But there is something that would cause me to shift my view towards the first. A battery of tests showing the existence of an HQ (Health Quotient) that obeys all of the usual properties that are supposed to indicate a strong gene contribution (heritability, correlations are greater between biological parents and child than adoptive parents and child), differences between populations are stable over time even as it changes, differences between populations still exist controlling for socioeconomic status, and the value becomes more stable during adulthood.

    I think it is pretty clear that health is largely a function of lifestyle and diet except in extreme cases. So if it mimics the properties of a supposedly genetically influenced trait then perhaps genes weren’t that important after all.

  34. Wulfrickson says:

    Scott, a few things:

    1. Thanks for the selection as Comment of the Month. (I knew that obscure knowledge about tunnel construction would be useful sometime!)

    2. You said:

    (lowering LW’s standards would also have a failure mode and not be a tradeoff-free solution; I can expand on this if it’s unclear)

    Could you expand on this? I’m curious. I lurk on LW sometimes; I was there to witness the Eugine_Nier fracas and wonder if this had any bearing on your thoughts about LW’s standards.

    3. As other commenters have pointed out, the colored outlines for new comments are wonderful. Would it be possible, however, also to mark such comments textually, e.g. by putting “Pseudonym says (new comment):” as the first line? This would make it possible to Ctrl-F “new comment” to find new comments, which is a good deal faster than scrolling through a thread and looking for the outlines. (Edit: DavidS made a similar suggestion upthread.)

  35. Fadeway says:

    Scott, the new “recent comments” plugin is great.

  36. Matt C says:

    Also wanted to say how nice the new comments tracker is. I usually don’t get much out of enhancements like these. Often I hate them. This one is genuinely useful.

  37. blacktrance says:

    Did the font on SSC just change, or is it my browser?