Aiyen

• Aiyen 20 Mar 2019 23:17 UTC

  3 points

  in reply to: Unreal's comment on: Rest Days vs Re­cov­ery Days

  That make a fair bit of sense. And what are your thoughts on work days? I get my work for my job done, but ad­vice on im­prov­ing pro­duc­tivity on chores and fu­ture plan­ning would be ap­pre­ci­ated. Also good point on pica!

• Aiyen 20 Mar 2019 18:16 UTC

  6 points

  on: Rest Days vs Re­cov­ery Days

  Very in­ter­est­ing di­chotomy! Definitely seems worth try­ing. I’m con­fused about the read­ing/​screen time/​video games dis­tinc­tion though. Why would read­ing seem ap­peal­ing but be­ing in front of a screen not? Watch­ing TV is es­sen­tially iden­ti­cal to read­ing right? You’re tak­ing in a pre­set story ei­ther way. Ad­mit­tedly you can read faster than TV char­ac­ters can talk, so maybe that makes it more re­ward­ing?

  Also, while play­ing more video games while re­cov­er­ing and fewer while rest­ing makes sense (they’re an easy ac­tivity while low on en­ergy, and thus will take up much of a re­cov­ery day, but less of a rest day), “just fol­low­ing my gut” can still lead to plenty of gam­ing. Does this mean that I should still play some on a rest day, just less? That I al­most never have enough en­ergy to rest in­stead of re­cover? That I’m too into gam­ing and this is skew­ing my gut such that a good rest day rule would be “fol­low your gut, ex­cept play­ing fewer/​no games to­day”?

• Aiyen 8 Mar 2019 22:34 UTC

  10 points

  on: [NeedAd­vice]How to stay Fo­cused on a long-term goal?

  First off, you prob­a­bly want to figure out if your nihilism is due to philos­o­phy or de­pres­sion. Would you nor­mally en­joy and value things, but idea of finite life gets in the way? Or would you have difficulty see­ing a point to things even if you were sud­denly granted im­mor­tal­ity and the heat death of the uni­verse was averted?

  Either way, it’s difficult to give a defini­tive solu­tion, as differ­ent things work for differ­ent peo­ple. That said, if the prob­lem seems to be philos­o­phy, it might be worth not­ing that the satis­fac­tion found in a good mo­ment isn’t a func­tion of any­thing that comes af­ter it. If you en­joy some­thing, or you help some­one you love, or you do any­thing else that seems valuable to you, the fact of that mo­ment is un­change­able. If the stars die in heaven, that can­not change the fact that you en­joyed some­thing. Another pos­si­ble solu­tion would be try­ing to sim­ply not think about it. I know that sounds hor­ribly dis­mis­sive, but it’s not meant to. In my own life there have been philo­soph­i­cal (and in my case re­li­gious) is­sues that I never man­aged to think my way out of… but when I stopped fo­cus­ing on the prob­lem it went away. I man­aged this only af­ter get­ting a job that let my brain say “okay, worry about God later, we need to get this task done first!” If you think it would help, find­ing an ac­tivity that de­mands at­ten­tion might help (if you feel that your brain will let you shift your at­ten­tion; if not this might just be overly stress­ful).

  If the prob­lem seems to be de­pres­sion, adrafinil and/​or modafinil are ex­tremely helpful for some peo­ple. Con­ven­tional treat­ments ex­ist too of course (ther­apy and/​or anti-de­pres­sants); I don’t know any­one who has benefited from ther­apy (at least not that they’ve told me), but one of my friends had night and day im­prove­ment with an anti-de­pres­sant (sadly I don’t re­mem­ber which one; if you like I can check with her). Another as­pect of over­com­ing de­pres­sion is hav­ing friends in the mo­ment and a plan for the fu­ture, not a plan you feel you should fol­low, but one you ac­tively want to. I don’t know your cir­cum­stances, but in­so­far as you can pri­ori­tize so­cial­iza­tion and work for the fu­ture, that might help.

  As for the ac­tual ques­tion of self-im­prove­ment, peo­ple vary wildly. An old friend of mine found huge im­prove­ments in her life due to schedul­ing; I do markedly bet­ter with­out it. The best ad­vice I can offer (and this very well might not help; drop it if it seems use­less or harm­ful) is three things:

  Don’t do what you think you should do, do what you ac­tu­ally want to (if there isn’t any­thing that you want, maybe don’t force try­ing to find some­thing too quickly ei­ther). Peo­ple find mo­ti­va­tion in pur­su­ing goals they ac­tu­ally find worth­while, but fol­low­ing a goal that sounds good but doesn’t ac­tu­ally ex­cite you is a recipe for burnout.

  Make ac­tion­able plans-if there’s some­thing you want to do, try to break it down into steps that are small enough, fa­mil­iar enough or straight­for­wards enough that you can ex­e­cute the plan with­out feel­ing out of your depth. Per­son­ally, at least, I find there’s a strik­ing “oh, that’s how I do that” feel­ing when a plan is made suffi­ciently ex­plicit, a sense that I’m no longer blun­der­ing around in a fog.

  Fi­nally, and per­haps most im­por­tantly, don’t elimi­nate your­self. That is, don’t aban­don a goal be­cause it looks difficult; make some­one else elimi­nate you. This is es­sen­tial be­cause many tasks look im­pos­si­ble from the out­side, es­pe­cially if you are de­pressed. It’s al­most the mir­ror image of the plan­ning fal­lacy-when peo­ple com­mit to do­ing some­thing, it’s all too easy to en­vi­sion ev­ery­thing go­ing right and not ac­count for set­backs. But be­fore you ac­tu­ally take the plunge, so to speak, it’s easy to just as­sume you can’t do any­thing, which is sim­ply not true.

• Aiyen 3 Mar 2019 17:43 UTC

  4 points

  on: To un­der­stand, study edge cases

  “To un­der­stand anatomy, dis­sect ca­dav­ers.” That’s less a de­liber­ate study of an edge case, and more due to the fact that we can’t eth­i­cally dis­sect liv­ing peo­ple!

• Aiyen 11 Feb 2019 19:35 UTC

  8 points

  on: Min­i­mize Use of Stan­dard In­ter­net Food Delivery

  At the risk of ap­pear­ing defec­tive, isn’t this the sort of ac­tion one would only want to take in a co­or­di­nated man­ner? If it turns out that use of such de­liv­ery ser­vices tends to force restau­rants out of busi­ness, then cer­tainly one would pre­fer a world where we don’t use those ser­vices and still have the restau­rants-you can’t or­der take out from a place that doesn’t ex­ist any­more! But de­cid­ing unilat­er­ally to boy­cott de­liv­ery im­poses a cost with­out any benefit-whether I choose to use de­liv­ery or not will not make the differ­ence. This looks like a clas­sic tragedy of the com­mons, where it is best to co­or­di­nate co­op­er­a­tion, but co­op­er­at­ing with­out that co­or­di­na­tion is a pure loss.

• Aiyen 29 Jan 2019 1:15 UTC

  1 point

  on: [Link] Did AlphaS­tar just click faster?

  In­ter­est­ing ar­ti­cle. It ar­gues that the AI learned spam click­ing from hu­man re­plays, then needed its APM cap raised to pre­vent spam click­ing from eat­ing up all of its APM bud­get and in­hibit­ing learn­ing. There­fore, it was per­mit­ted to use in­hu­manly high burst APM, and with all its clicks po­ten­tially effec­tive ac­tions in­stead of spam, its effec­tive ac­tions per minute (EPM, ac­tions not count­ing spam clicks) are go­ing to out­class hu­man pros to the point of break­ing the game and ren­der­ing ac­tual strat­egy re­dun­dant.

  Ex­cept that if it’s spam­ming, those clicks aren’t effec­tive ac­tions, and if those clicks are effec­tive ac­tions, it’s not spam­ming. To the ex­tent Alphas­tar spams, its su­per­hu­man APM is mis­lead­ing, and the match is fairer than it might oth­er­wise ap­pear. To the ex­tent that it’s us­ing high burst EPM in­stead, that can po­ten­tially turn the game into a micro match rather than the strat­egy match that peo­ple are more in­ter­ested in. But that isn’t a ques­tion of spam click­ing.

  Of course, if it started spam click­ing, needed the APM cap raised, con­verted its spam into ac­tual high EPM and Deep­mind didn’t lower the cap af­ter­wards, then the ar­ti­cle’s ob­jec­tion holds true. But that didn’t sound like what it was ar­gu­ing (though per­haps I mi­s­un­der­stood it). In­deed, it seems to ar­gue the re­verse, that spam click­ing was so in­grained that the AI never broke the habit.

• Aiyen 25 Jan 2019 22:17 UTC

  2 points

  in reply to: avturchin's comment on: For what do we need Su­per­in­tel­li­gent AI?

  It de­pends on the goal. We can prob­a­bly defeat ag­ing with­out need­ing much more so­phis­ti­cated AI than Alphafold (a re­cent Google AI that par­tially cracked the pro­tein fold­ing prob­lem). We might be able to pre­vent the cre­ation of dan­ger­ous su­per­in­tel­li­gences with­out AI at all, just with suffi­cient surveillance and reg­u­la­tion. We very well might not need very high-level AI to avoid the worst im­me­di­ately un­ac­cept­able out­comes, such as death or X-risk.

  On the other hand, true su­per­in­tel­li­gence offers both the abil­ity to be far more se­cure in our en­deav­ors (even if hu­man-level AI can mostly se­cure us against X-risk, it can­not do so any­where nearly as re­li­ably as a stronger mind), and the abil­ity to flour­ish up to our po­ten­tial. You list high-speed space travel as “nei­ther ur­gent nor nec­es­sary”, and that’s true-a world with­out near light­speed travel can still be a very good world. But even­tu­ally we want to max­i­mize our val­ues, not merely avoid the worst ways they can fall apart.

  As for truly ur­gent tasks, those would pre­sum­ably re­volve around avoid­ing death by var­i­ous means. So anti-ag­ing re­search, anti-dis­ease/​trauma re­search, gain­ing se­cu­rity against hos­tile ac­tors, en­sur­ing ac­cess to food/​wa­ter/​shelter, de­tect­ing and avoid­ing X-risks. The last three may well benefit greatly from su­per­in­tel­li­gence, as com­pre­hen­sively deal­ing with hos­tiles is ex­tremely com­pli­cated and also likely nec­es­sary for food dis­tri­bu­tion, and there may well be X-risks a hu­man-level mind can’t de­tect.

• Aiyen 25 Jan 2019 22:01 UTC

  1 point

  in reply to: avturchin's comment on: For what do we need Su­per­in­tel­li­gent AI?

  Most peo­ple seem to need some­thing to do to avoid bore­dom and po­ten­tially out­right de­pres­sion. How­ever, it is far from clear that work as we know it (which is op­ti­mized for our cur­rent pro­duc­tion needs, and in no way for the benefit of the work­ers as such) is the best way to solve this prob­lem. There is likely a need to de­velop other things for peo­ple to do alongside alle­vi­at­ing the need for work, but sim­ply say­ing “un­em­ploy­ment is bad” would seem to miss that there may be bet­ter op­tions than ei­ther con­ven­tional work or idle­ness.

• Aiyen 25 Jan 2019 21:57 UTC

  2 points

  in reply to: Chris_Leong's comment on: For what do we need Su­per­in­tel­li­gent AI?

  Where gov­er­nance is the bar­rier to hu­man flour­ish­ing, doesn’t that mean that us­ing AI to im­prove gov­er­nance is use­ful? A tran­shu­man mind might well be able to figure out not only bet­ter poli­cies but how to get those poli­cies en­acted (per­sua­sion, force, mind con­trol, in­cen­tives, some­thing else we haven’t thought of yet). After all, if we’re wor­ried about a po­ten­tially un­friendly mind with the power to defeat the hu­man race, the flip side is that if it’s friendly, it can defeat harm­ful parts of the hu­man race, like poorly-run gov­ern­ments.

• Aiyen 25 Jan 2019 21:52 UTC

  1 point

  in reply to: avturchin's comment on: For what do we need Su­per­in­tel­li­gent AI?

  Safer for the uni­verse maybe, per­haps not for the old per­son them­selves. Cry­on­ics is highly spec­u­la­tive-it *should* work, given that if your in­for­ma­tion is pre­served it should be pos­si­ble to re­con­struct you, and cool­ing a sys­tem enough should re­duce ther­mal noise and re­ac­tivity enough to pre­serve in­for­ma­tion… but we just don’t know. From the per­spec­tive of some­one near death, count­ing on cry­on­ics might be as risky or more so than a quick AI.

• Aiyen 21 Jan 2019 15:32 UTC

  2 points

  in reply to: mr-hire's comment on: Do the best ideas float to the top?

  This. Also, poli­ti­cal fac­tors-ideas that boost the sta­tus of your tribe are likely to be very com­pet­i­tive in­de­pen­dently of truth and nearly so of com­plex­ity (though if they’re too com­plex one would ex­pect to see sim­plified ver­sions prop­a­gat­ing as well).

• Aiyen 21 Jan 2019 15:14 UTC

  4 points

  in reply to: ChristianKl's comment on: Life can be bet­ter than you think

  “Emo­tions have their role in pro­vid­ing mean­ing.”

  Even if true, is mean­ing ac­tu­ally valuable? I would far rather be happy than mean­ingful, and a uni­verse of truth, beauty, love and joy seems much more worth­while than a uni­verse of mean­ing.

  Caveat-I feel much the same dis­con­nect in hear­ing about mean­ing that Gal­ton’s non-imagers ap­peared to feel about men­tal imag­ing, so there’s a pretty good chance I sim­ply don’t have the men­tal cir­cuitry needed to ap­pre­ci­ate or care about mean­ing. You might be gen­uinely pur­su­ing some­thing very im­por­tant to you in seek­ing mean­ing. On the other hand, even if that’s true, it’s worth not­ing that there are some peo­ple who don’t need it.

• Aiyen 10 Jan 2019 2:15 UTC

  1 point

  on: What are ques­tions?

  It’s a no­ticed gap in your knowl­edge.

• Aiyen 2 Jan 2019 23:24 UTC

  1 point

  on: Con­se­quen­tial­ism FAQ

  Link doesn’t seem to work.

• Aiyen 30 Dec 2018 19:40 UTC

  8 points

  on: What makes peo­ple in­tel­lec­tu­ally ac­tive?

  My best guess: There’s a differ­ence be­tween re­view­ing ideas and ex­plor­ing them.
  Re­view­ing ideas al­lows you to un­der­stand con­cepts, think about them and talk about them, but you’re look­ing at ma­te­rial you already have. Con­sider some­one prepar­ing a lec­ture well-they’ll make sure that they have no con­fu­sion about what they’re cov­er­ing, and write elo­quently on the topic at hand.

  On the other hand, this is think­ing along pre-set path­ways. It can be very use­ful for both learn­ing and teach­ing, but you aren’t likely to dis­cover some­thing new. Ex­plor­ing ideas, by con­trast, is look­ing at a part of idea space and then see­ing what you can find. It’s think­ing about the im­pli­ca­tions of things you know, and look­ing to see if an un­ex­pected re­sult shows up, or sim­ply con­sid­er­ing a topic and hop­ing that some­thing new on the sub­ject oc­curs to you.

• Aiyen 19 Dec 2018 19:19 UTC

  2 points

  on: Fif­teen Things I Learned From Watch­ing a Game of Se­cret Hitler

  “The more liberal poli­cies you pass, the more likely it is any fu­ture policy will be fas­cist.”

  Sadly this one is likely true irl. When you have a gov­ern­ment that passes more and more laws, and does not re­peal old laws, then the de­gree of re­stric­tion of peo­ple’s lives in­creases mono­ton­i­cally. This cre­ates a prece­dent for ever more con­trol, un­til the end is ei­ther a back­lash or tyranny.

• Aiyen 19 Dec 2018 19:12 UTC

  3 points

  in reply to: lionhearted's comment on: 18-month fol­low-up on my self-con­cept work

  Not Kaj, but shame and self-con­cept (dam­ag­ing or oth­er­wise) are thoughts (or self-con­cept is a thought and shame is an emo­tion pro­duced by cer­tain thoughts). It seems ob­vi­ous that peo­ple with a greater ten­dency to think will be at greater risk of harm­ful thoughts. Of course, they’ll also have a bet­ter chance of com­ing up with some­thing benefi­cial as well, but that doesn’t strike me as likely to can­cel out the harm. Hu­mans are fairly well adapted for our in­tel­lec­tual and so­cial niche; there are a lot more ways for in­tro­spec­tion to break things than to im­prove them.

• Aiyen 27 Sep 2018 0:30 UTC

  6 points

  on: Open Thread Septem­ber 2018

  Happy Petrov Day!

• Aiyen 10 Aug 2018 16:20 UTC

  1 point

  in reply to: Elo's comment on: A Ra­tion­al­ist’s Guide to...

  ...? “Win­ning” isn’t just an ab­strac­tion, ac­tu­ally win­ning means get­ting some­thing you value. Now, maybe many ra­tio­nal­ists are in fact win­ning, but if so, there are spe­cific val­ues we’re at­tain­ing. It shouldn’t be hard to delineate them.

  It should look like, “This per­son got a new job that makes them much hap­pier, that per­son lost weight on an ev­i­dence-based diet af­ter failing to do so on a string of other diets, this other per­son found a sig­nifi­cant other once they started prac­tic­ing Ali­corn’s self-aware­ness tech­niques and learned to ac­cept their ner­vous­ness on a first date...” It might even look like, “This per­son de­vel­oped a new tech­nol­ogy and is cur­rently work­ing on a startup to build more pro­to­types.”

  In none of these cases should it be hard to ex­plain how we’re win­ning, nor should Tim’s “not look­ing care­fully enough” be an is­sue. Even if the wins are limited to sub­jec­tive well-be­ing, you should at least be able to ex­plain that! Do you be­lieve that we’re win­ning, or do you merely be­lieve you be­lieve it?

• Aiyen 22 Jul 2018 16:29 UTC

  1 point

  on: Who Wants The Job?

  This is si­mul­ta­neously hor­rify­ing and in­cred­ibly com­fort­ing. One would hope that peo­ple would be or­ders of mag­ni­tude bet­ter than this. But it also bodes very well for the fu­ture prospects of any­one re­motely com­pe­tent (un­less your boss is like this...)