Ruby

• Ruby 15 May 2019 17:59 UTC

  4 points

  on: Ruby’s Short­form Feed

  For my own refer­ence.

  Brief timeline of no­table events for LW2:

  • 2017-09-20 LW2 Open Beta launched

  • (2017-10-13 There is No Fire Alarm pub­lished)

  • (2017-10-21 AlphaGo Zero Sig­nifi­cance post pub­lished)

  • 2017-10-28 Inad­e­quate Equil­ibria first post published

  • (2017-12-30 Good­hart Tax­on­omy Pub­lish) ← maybe part of Jan­uary spike?

  • 2018-03-23 Offi­cial LW2 launch and switch­ing of www.less­wrong.com to point to the new site.

  In paren­the­ses events are pos­si­ble draws which spiked traf­fic at those times.

• Ruby 15 May 2019 15:55 UTC

  2 points

  in reply to: jp's comment on: Data Anal­y­sis of LW: Ac­tivity Levels + Age Distri­bu­tion of User Accounts

  What’s the mechanism of ac­tion? If LW doesn’t die it will even­tu­ally achieve its aims be­cause .. ?

  I think you re­ally an­swered it, as long as you don’t die:

  you get to con­tin­u­ing swing­ing at ideas un­til you hit a home run.

  Of course, this works a lot bet­ter if you’re learn­ing as you go and your suc­ces­sive at­tempts fall closer to suc­cess than if you’re ran­domly try­ing things. I’d like to think that the LW isn’t ran­domly swing­ing. There are also hard pieces like keep­ing the team en­gaged and op­ti­mistic even as things don’t seem dra­mat­i­cally differ­ent (though you could call that part of not dy­ing).

• Ruby 15 May 2019 4:39 UTC

  5 points

  in reply to: jp's comment on: Data Anal­y­sis of LW: Ac­tivity Levels + Age Distri­bu­tion of User Accounts

  One should ex­pect a site that was pro­vid­ing a lot of value to its users to grow, even if it wasn’t ex­plic­itly try­ing to.

  Yes, I do ex­pect if you’re gen­er­at­ing enough value you will should see au­to­matic growth, from which I in­fer that LessWrong 2.0 isn’t pro­vid­ing that much value to its users right now. Though I think there’s a mix of rea­sons to not be es­pe­cially pes­simistic:

  • Suc­cess­ful com­pa­nies with worth­while prod­ucts seem to me to still have to in­vest in get­ting new users. My feel­ing (not re­ally backed by data) is that you have to be out­stand­ingly good to get full-on or­ganic growth with­out try­ing. Not be­ing there doesn’t mean you’re not pro­vid­ing value.

  • We see in the graphs that LW was not grow­ing for most of its his­tory: most of the met­rics peak around 2011 and re­main steady or de­cline slowly un­til 2015. I would ar­gue that de­spite not grow­ing, LW was still pro­vid­ing a lot of value to its users and the world dur­ing this pe­riod.

  • My out­side view and in­side view lead me be­lieve hockey stick growth to be real. Part of my model is that even if you’re do­ing many things right, it might re­quire hav­ing all the pieces click into place be­fore dra­matic growth starts. The pieces are con­nected in se­ries, not par­allel. Re­lat­edly, some­times the key to win­ning big is just not dy­ing for long enough.

  • LW2 is much more fussy about which value we provide to which users than I ex­pect most com­pa­nies are. Most com­pa­nies are try­ing to find ap­prox­i­mately any product and cor­re­spond­ing set of users such that the value pro­vided to users can be used to ex­tract money some­how. In con­trast, I care only about find­ing prod­ucts and users to whom pro­vid­ing value will gen­er­ate sig­nifi­cant value for the world at large (par­tic­u­larly through the de­vel­op­ment/​train­ing of ra­tio­nal­ity and gen­eral in­tel­lec­tual progress on im­por­tant prob­lems). I think this is a much more re­stric­tive con­straint. It leads me (and I think the team gen­er­ally) to want to forego many op­por­tu­ni­ties for user/​ac­tivity growth if we don’t think they’ll lead to greater value for the world over­all. Be­cause of this, I’m not wor­ried yet that we haven’t hit on a for­mula for pro­vid­ing value that’s or­gan­i­cally get­ting a lot of growth. We have a nar­row tar­get.

  Gen­er­ally, I (and oth­ers on the team) don’t con­sider LW to have achieved the non­profit ana­log of product-mar­ket fit. More pre­cisely, we haven’t hit upon a definite and scal­able mechanism for gen­er­at­ing large amounts of value for the world (es­pe­cially in­tel­lec­tual progress). I have an up­com­ing post I wish I could link to which de­scribes var­i­ous ideas we’re try­ing or think­ing about as mechanisms. Open Ques­tions is one such at­tempt.

  Per­haps most of the value of the site is in the fact that it has posts, com­ments and votes. Beyond that it’s the value of the con­tent, and that is mod­est and static.

  I’m un­sure of your mean­ing here. Are you say­ing there’s con­tent sep­a­rate from posts and com­ments? I con­sider all our con­tent to fall into those cat­e­gories. Some of it is ar­guably static, but I’m not sure I’d say mod­est? Can you say more what you meant by that?

Data Anal­y­sis of LW: Ac­tivity Levels + Age Distri­bu­tion of User Accounts

• Ruby 13 May 2019 20:21 UTC

  4 points

  in reply to: Ruby's comment on: Why is it valuable to know whether space coloniza­tion is fea­si­ble?

  Ad­den­dum:

  In an Open Philan­thropy Pro­ject blog post, The Mo­ral Value of the Far Fu­ture, Holden Karnofsky men­tions Nick Bostrom’s Astro­nom­i­cal Waste ar­gu­ment to say that he does not con­sider it ro­bust enough to play an over­whelming role in his be­lief sys­tems and ac­tions.

  In Astro­nom­i­cal Waste, Nick Bostrom makes a more ex­treme and more spe­cific claim: that the num­ber of hu­man lives pos­si­ble un­der space coloniza­tion is so great that the mere pos­si­bil­ity of a hugely pop­u­lated fu­ture, when con­sid­ered in an “ex­pected value” frame­work, dwarfs all other moral con­sid­er­a­tions. I see no ob­vi­ous an­a­lyt­i­cal flaw in this claim, and give it some weight. How­ever, be­cause the ar­gu­ment re­lies heav­ily on spe­cific pre­dic­tions about a dis­tant fu­ture, seem­ingly (as far as I can tell) backed by lit­tle other than spec­u­la­tion, I do not con­sider it “ro­bust,” and so I do not con­sider it ra­tio­nal to let it play an over­whelming role in my be­lief sys­tem and ac­tions [em­pha­sis added].

  Ad­mit­tedly, Karnofsky pro­ceeds to say that even if he fully ac­cepted the rea­son­ing, he isn’t sure what im­pli­ca­tions it would have.

  In ad­di­tion, if I did fully ac­cept the rea­son­ing of “Astro­nom­i­cal Waste” and eval­u­ate all ac­tions by their far fu­ture con­se­quences, it isn’t clear what im­pli­ca­tions this would have. As dis­cussed be­low, given our un­cer­tainty about the speci­fics of the far fu­ture and our rea­sons to be­lieve that do­ing good in the pre­sent day can have sub­stan­tial im­pacts on the fu­ture as well, it seems pos­si­ble that “see­ing a large amount of value in fu­ture gen­er­a­tions” and “see­ing an over­whelming amount of value in fu­ture gen­er­a­tions” [em­pha­sis added] lead to similar con­se­quences for our ac­tions.

  Nonethe­less, I sus­pect that were we to have a non-spec­u­la­tive, ro­bust case about what is pos­si­ble that this well might push our be­hav­ior in par­tic­u­lar di­rec­tions. For in­stance, per­haps we find that Bostrom’s 10^38 hu­mans lost per cen­tury of de­lay is ex­tremely spec­u­la­tive, yet 10^20 is em­i­nently at­tain­able. I be­lieve that if did have a ro­bust case for the lat­ter, this would shift the pri­ori­ti­za­tion of some and likely bolster the al­tru­is­tic mo­ti­va­tion of those who right now are pri­mar­ily sus­tained by the spec­u­la­tive plau­si­bil­ity of Bostrom’s ex­treme case.

  Per­haps more im­por­tantly, if we are un­able to even es­tab­lish a firm lower bound much above what the Earth alone could sus­tain long-term, then those who have made Astro­nom­i­cal Waste ar­gu­ments part of their be­lief sys­tems and ac­tions have rea­son to pause and re­con­sider how they should up­date given that the po­ten­tial of space coloniza­tion might be much weaker than pre­vi­ously hoped.

• Ruby 13 May 2019 1:04 UTC

  2 points

  in reply to: Said Achmiz's comment on: Co­her­ent de­ci­sions im­ply con­sis­tent utilities

  A week or so ago Ar­bital was work­ing but had load times of sev­eral min­utes.

• Ruby 12 May 2019 19:00 UTC

  4 points

  in reply to: James_Miller's comment on: Claims & As­sump­tions made in Eter­nity in Six Hours

  For some des­ti­na­tions, but not for most of them (I’m pretty sure). At least Eter­nity in Six Hours spends a great de­tail of time dis­cussing de­cel­er­a­tion.

• Ruby 11 May 2019 1:33 UTC

  7 points

  on: Space coloniza­tion: what can we definitely do and how do we know that?

  The an­swer to this ques­tion likely de­pends heav­ily on what we con­sider to be ad­e­quate coloniza­tion:

  1) Run­ning com­pu­ta­tion in oth­ers star sys­tems, i.e. Run­ning digi­tal minds on com­put­ers or other com­pu­ta­tional pro­cesses. (this is what Eter­nity in Six Hours as­sumes)

  or

  2) Hav­ing ac­tual, or­di­nary biolog­i­cal hu­mans colonize the stars.

  There are challenges com­mon and sep­a­rate to each.

  Last­ing the Journey

  In ei­ther case, you must be able to cre­ate a probe (to use the lan­guage of Eter­nity in Six Hours) which can last the du­ra­tion for a trip which lasts thou­sands to mil­lions of years. Is it at all fea­si­ble to have hu­mans last long in some form? (Per­haps only as em­bryos which can be “grown” upon ar­rival, but even then, can we safely pre­serve biolog­i­cal ma­te­rial for mil­le­nia?) Could cry­on­ics some­how be a solu­tion? Even if you were only send­ing com­put­ers/​robots, can we build elec­tri­cal and me­chan­i­cal de­vices which won’t break down af­ter such ex­tremely long time pe­ri­ods?

  Challenges for Humans

  Nick Beck­stead’s pre­li­me­nary notes men­tion micro­grav­ity, cos­mic ra­di­a­tion, health and re­pro­duc­tion in space, and ge­netic di­ver­sity as con­sid­er­a­tions which come into play when send­ing live hu­mans through space.

  Challenges for Computers

  Can we build ma­chines (as­sume non-AGI) we can solve all the prob­lems they will en­counter in differ­ent sys­tems?

[Question] How do the differ­ent star-types in the uni­verse (red dwarf, etc.) re­lated to hab­it­a­bil­ity for hu­man-like life?

[Question] How many “hu­man” hab­it­able planets/​stars are in the uni­verse?

• Ruby 11 May 2019 0:50 UTC

  4 points

  on: What speeds do you need to achieve to colonize the Milky Way?

  How fast you need to go un­sur­pris­ingly de­pends on quickly you need to get there. I’ve es­ti­mated that 100kly is larger than the dis­tance to most places within the Milky Way.

  • Trav­el­ling at 99%c, you can cover that in ~100,000 years.

  • Trav­el­ling at 50%c, you can cover that in 200,000 years.

  • Trav­el­ling at 10%c, you can cover that dis­tance in 1,000,000 years.

  • Trav­el­ling at 1%c, you can cover that dis­tance in 10,000,000 years.

  Re­call that there are at least tens of mil­lions of stars in the Milky Way. There are prob­a­bly many stars within 50kly or even 25kly of Earth.

  Nonethe­less, these dis­tances mean that even at ex­tremely fast speeds it would still take tens of thou­sands of years to mil­lions of years. This may or may not be a prob­lem. The uni­verse will prob­a­bly last for at least an­other few billion years, com­pared to which a mil­lion years is not much at all. The ques­tion is whether your ex­pe­di­tion can sur­vive that long be­tween stars. (It might make a big differ­ence whether you are send­ing only digi­tal ma­chines or hu­mans too.)

  What are the dis­tances?

  Tak­ing stats from its Wikipe­dia en­try, the Milky Way has a di­ame­ter of 150-200 kly (kilo­lightyears), how­ever:

  The disk of stars in the Milky Way does not have a sharp edge be­yond which there are no stars. Rather, the con­cen­tra­tion of stars de­creases with dis­tance from the cen­ter of the Milky Way. For rea­sons that are not un­der­stood, be­yond a ra­dius of roughly 40,000 ly (13 kpc) from the cen­ter, the num­ber of stars per cu­bic par­sec drops much faster with ra­dius.[68] - Wikipedia

  How­ever, I will as­sume that the up­per bound given of 200kly cap­tures most of the 100-400 billion stars.

  Our sun is 26.4 ± 1.0 kly from the Galac­tic Cen­ter (see image). It might be difficult to travel through the cen­ter of the galaxy, but let’s as­sume that the dis­tance you travel to get any­where in Milk Way from our sun is no more than trav­el­ing to the Galac­tic Cen­ter (~25ky) plus the up­per bound of the ra­dius (~100kly), so ap­prox­i­mately 125kly. That’s the dis­tance to the outer edge so ac­tu­ally the vast ma­jor­ity of des­ti­na­tions should be less than that. One could do some fancier tri­gonom­e­try to get ex­act num­bers and nice av­er­ages, but this gives us the or­der of mag­ni­tude: ~100kly to travel al­most any­where in the Milky Way.

  That is prob­a­bly still well above av­er­age since the den­sity of stars is much higher to­wards the core. Likely there are a lot of stars within 50kly.

  

• Ruby 10 May 2019 23:12 UTC

  4 points

  on: Why is it valuable to know whether space coloniza­tion is fea­si­ble?

  The kinds of num­bers thrown around in the as­tro­nom­i­cal waste ar­gu­ment are some­times ac­cused of be­ing a Pas­cal’s Mug­ging. Even if one has doubts about whether to work on ex­is­ten­tial risk re­duc­tion, it could be ar­gued that be­cause the Far Fu­ture has such over­whelming and im­mense value that the ex­pected value of work­ing on ex­is­ten­tial risk out­weighs all other op­por­tu­ni­ties, e.g. near-term al­tru­is­tic pro­jects like global poverty, global health, and an­i­mal welfare.

  Hav­ing sharper es­ti­mates of the po­ten­tial of the Far Fu­ture, bounded by how much of the uni­verse we can ac­tu­ally reach, could help us re­late to as­tro­nom­i­cal waste ar­gu­ments with far more prin­ci­ple than “aahhh, these are such big num­bers!!”

  They’re big num­bers, but not all num­bers are equally big.

• Ruby 10 May 2019 23:02 UTC

  4 points

  on: Why is it valuable to know whether space coloniza­tion is fea­si­ble?

  The as­sump­tion that we can colonize the stars is core to the Astro­nom­i­cal Waste Ar­gu­ment made in fa­vor of work­ing on ex­is­ten­tial risk re­duc­tion. If this as­sump­tion is weak­ened, so is the case for pri­ori­tiz­ing work ex­is­ten­tial risk re­duc­tion.

  Most things are im­pos­si­ble. Per­haps our be­lief that we could pos­si­ble colonize the stars is based only our ig­no­rance. If we ac­tu­ally tried to colonize the stars (or sim­ply tried to ac­tu­ally look into the pos­si­bil­ity), we would find that we shouldn’t take it for granted at all that space coloniza­tion is a re­al­is­tic pos­si­bil­ity.

  Sum­mary of the Astro­nom­i­cal Waste Argument

  Nick Bock­strom’s 2003 pa­per, Astro­nom­i­cal Waste: The Op­por­tu­nity Cost of De­layed Tech­nolog­i­cal Devel­op­ment:

  With very ad­vanced tech­nol­ogy, a very large pop­u­la­tion of peo­ple liv­ing happy lives could be sus­tained in the ac­cessible re­gion of the uni­verse. For ev­ery year that de­vel­op­ment of such tech­nolo­gies and coloniza­tion of the uni­verse is de­layed, there is there­fore a cor­re­spond­ing op­por­tu­nity cost: a po­ten­tial good, lives worth liv­ing, is not be­ing re­al­ized. Given some plau­si­ble as­sump­tions, this cost is ex­tremely large.

  Bostrom ar­rives at differ­ent es­ti­mates of the po­ten­tial num­ber of hu­man minds de­pend­ing on whether we are satis­fied with run­ning “hu­man” minds on com­put­ers or wish to stick with biolog­i­cal in­stan­ti­a­tion.

  Us­ing digi­tal in­stan­ti­a­tion:

  As a rough ap­prox­i­ma­tion, let us say the Virgo Su­per­cluster con­tains 10^13 stars. One es­ti­mate of the com­put­ing power ex­tractable from a star and with an as­so­ci­ated planet-sized com­pu­ta­tional struc­ture, us­ing ad­vanced molec­u­lar nan­otech­nol­ogy, is 10^42 op­er­a­tions per sec­ond. A typ­i­cal es­ti­mate of the hu­man brain’s pro­cess­ing power is roughly 10^17 op­er­a­tions per sec­ond or less. Not much more seems to be needed to simu­late the rele­vant parts of the en­vi­ron­ment in suffi­cient de­tail to en­able the simu­lated minds to have ex­pe­riences in­dis­t­in­guish­able from typ­i­cal cur­rent hu­man ex­pe­riences. Given these es­ti­mates, it fol­lows that the po­ten­tial for ap­prox­i­mately 10^38 hu­man lives is lost ev­ery cen­tury that coloniza­tion of our lo­cal su­per­cluster is de­layed; or equiv­a­lently, about 10^29 po­ten­tial hu­man lives per sec­ond.

  Us­ing biolog­i­cal in­stan­ti­a­tion:

  Sup­pose that about 10^10 biolog­i­cal hu­mans could be sus­tained around an av­er­age star. Then the Virgo Su­per­cluster could con­tain 10^23 biolog­i­cal hu­mans. This cor­re­sponds to a loss of po­ten­tial of over 10^13 po­ten­tial hu­man lives per sec­ond of de­layed coloniza­tion.

  Bostrom clar­ifies that not only util­i­tar­i­ans should care about this im­mense po­ten­tial value which might be reached:

  Utili­tar­i­ans are not the only ones who should strongly op­pose as­tro­nom­i­cal waste. There are many views about what has value that would con­cur with the as­sess­ment that the cur­rent rate of wastage con­sti­tutes an enor­mous loss of po­ten­tial value. For ex­am­ple, we can take a thicker con­cep­tion of hu­man welfare than com­monly sup­posed by util­i­tar­i­ans (whether of a he­do­nis­tic, ex­pe­ri­en­tial­ist, or de­sire-satis­fac­tion­ist bent), such as a con­cep­tion that lo­cates value also in hu­man flour­ish­ing, mean­ingful re­la­tion­ships, no­ble char­ac­ter, in­di­vi­d­ual ex­pres­sion, aes­thetic ap­pre­ci­a­tion, and so forth. So long as the eval­u­a­tion func­tion is ag­grega­tive (does not count one per­son’s welfare for less just be­cause there are many other per­sons in ex­is­tence who also en­joy happy lives) and is not rel­a­tivized to a par­tic­u­lar point in time (no time-dis­count­ing), the con­clu­sion will hold.

  Th­ese con­di­tions can be re­laxed fur­ther. Even if the welfare func­tion is not perfectly ag­grega­tive (per­haps be­cause one com­po­nent of the good is di­ver­sity, the marginal rate of pro­duc­tion of which might de­cline with in­creas­ing pop­u­la­tion size), it can still yield a similar bot­tom line pro­vided only that at least some sig­nifi­cant 5 com­po­nent of the good is suffi­ciently ag­grega­tive. Similarly, some de­gree of time dis­count­ing fu­ture goods could be ac­com­mo­dated with­out chang­ing the con­clu­sion.

  Clearly, the ex­tent to which we can ac­tu­ally colonize star sys­tems be­yond our own af­fects how strong an ar­gu­ment there is from as­tro­nom­i­cal waste (or as I would rather call it, our as­tro­nom­i­cal po­ten­tial). If we can in fact be con­fi­dent that we can colonize the en­tire reach­able uni­verse, that might be 10^17 stars in­stead of the 10^13 in just the Virgo Su­per­cluster. An even stronger ar­gu­ment than Bostrom states. On the other hand, if we can’t even colonize be­yond our star sys­tem, we’re just at 10^0 stars. Then there’d be no as­tro­nom­i­cal ar­gu­ment at all.

• Ruby 9 May 2019 0:47 UTC

  4 points

  in reply to: KevinKohler's comment on: Space coloniza­tion: what can we definitely do and how do we know that?

  That’s in­ter­est­ing. I agree that given that con­sid­er­a­tion the term “coloniza­tion” is pos­si­bly mis­lead­ing. I have been us­ing it more in the sense of “you have hu­man civ­i­liza­tion over there” rather than “the colonies of the king­dom of Bri­tain.” I think I don’t mind if the differ­ent “colonies” are au­tonomous.

[Question] How many galax­ies could we reach trav­el­ing at 0.5c, 0.8c, and 0.99c?

[Question] How many hu­mans could po­ten­tially live on Earth over its en­tire fu­ture?

• Ruby 8 May 2019 23:12 UTC

  4 points

  on: What are the claims/​ar­gu­ments made in Eter­nity in Six Hours?

  Ex­tract­ing my re­sponse from this post.

  Claims and As­sump­tions (not ex­haus­tive)

  • Self-repli­cat­ing probes for coloniza­tions could be launched to a frac­tion of light­speed us­ing fixed launch sys­tems such as coil­guns or quench­guns as (op­posed to rock­ets).

  • Only six hours of the sun’s en­ergy (3.8x10^26W) are re­quired to com­mence the coloniza­tion of the en­tire uni­verse.

    • A fu­ture hu­man civ­i­liza­tion could eas­ily as­pire to this amount of en­ergy.

  • Since the pro­ce­dure is con­junc­tion of de­signs and yet each of the re­quire­ments have mul­ti­ple path­ways to im­ple­men­ta­tion, the whole con­struc­tion is ro­bust.

  • Hu­mans have gen­er­ally been quite suc­cess­ful at copy­ing or co-op­ing na­ture. We can as­sume that any­thing done in the nat­u­ral world can be done un­der hu­man con­trol, e.g. self-repli­ca­tors and AI.

  • Any task which can be performed can be au­to­mated.

  • It would be ru­inously costly to send over a large coloniza­tion fleet, and is much more effi­cient to send over a small pay­load which builds what is re­quired in situ, i.e. von Neu­mann probes.

  • Data stor­age will not be much an is­sue.

    • Ex­am­ple: can fit all the world’s data and up­load of ev­ery­one in Bri­tain in gram of crys­tal.

  • 500 tons is a rea­son­able up­per bound for the size of a self-repli­cat­ing probe.

  • A repli­ca­tor with mass of 30 grams would not be un­rea­son­able.

  • An­ti­mat­ter an­nihila­tion, nu­clear fu­sion, and nu­clear fis­sion are all pos­si­ble rocket types to be used for de­cel­er­a­tion.

    • Pro­cesses like mag­netic sail, grav­i­ta­tional as­sist, and “Bus­sard ram­jet” are con­ceiv­able and pos­si­ble, but to be con­ser­va­tive are not re­lied on.

  • Nu­clear fis­sion re­ac­tors could be made 90% effi­cient. Cur­rent re­ac­tor de­signs could reach effi­cien­cies of over 50% of the the­o­ret­i­cal max­i­mum.

    • Any fall-off in fis­sion effi­ciency re­sults in a dra­matic de­crease in de­cel­er­a­tion po­ten­tial.

    • They ig­nore de­cel­er­a­tion caused by the ex­pan­sion of the uni­verse.

  • As­sume probe is of sturdy enough con­struc­tion to sur­vive a grenade blad (800kJ)

  • Re­dun­dancy re­quired for a probe to make it to a galaxy is given by R = exp(dAρ ) where is d is dis­tance to be trav­el­led (in co­mov­ing co­or­di­nate), A is cross-sec­tion of the probe, and ρ is the den­sity of dan­ger­ous par­ti­cles.

    • Danger­ous par­ti­cle size given as a func­tion of speed of the probe by equa­tion in the pa­per.

    • From slower probes (80%c and 50%c) re­dun­dancy re­quired is low, two probes are enough to en­sure one sur­vives.

    • If you have a 500T repli­ca­tor, you have more cross-sec­tion but also bet­ter abil­ity to shield.

    • Den­sity of mat­ter in space is much higher in in­ter­stel­lar space com­pared to in­ter­galac­tic space. Might not be pos­si­ble to launch uni­verse-coloniza­tion di­rectly from our sun.

  • Dyson spheres are very doable. As­sumed to have ¹⁄₃ effi­cien­cies over sun’s out­put (3.8x10^26)

    • We could dis­assem­ble Mer­cury and turn it into a Dyson sphere.

  • Launch sys­tems could achieve en­ergy effi­ciency of 50%.

  • Apart from risks of col­li­sion, get­ting to the fur­ther galax­ies is as easy as get­ting to the clos­est, the only differ­ence is a longer wait be­tween the ac­cel­er­a­tion and de­cel­er­a­tion phases.

  • Trav­el­ling at 50c% there are 116 mil­lion galax­ies reach­able; at 80% there are 762 mil­lion galax­ies reach­able; at 99%c, you get 4.13 billion galax­ies.

    • For refer­ence, there are 100 to 400 billion stars in the Milky Way, and from a quick check it might be rea­son­able to as­sume 100 billion is the av­er­age galaxy.

      • The abil­ity to colonize the uni­verse as op­posed to just the Milky Way is the differ­ence be­tween ~10^8 stars and ~10^16 or ~10^17 starts. A fac­tor of 100 mil­lion.

  • On a cos­mic scale, the cost, time and en­ergy needed to com­mence a coloniza­tion of the en­tire reach­able uni­verse are en­tirely triv­ial for an ad­vanced hu­man-like civ­i­liza­tion.

  • En­ergy costs could be cut by a fac­tor of hun­dred or thou­sand by aiming for clusters or su­per­clusters [of galax­ies] and spread­ing out from there.

Claims & As­sump­tions made in Eter­nity in Six Hours

[Question] What speeds do you need to achieve to colonize the Milky Way?

[Question] Could a su­per­in­tel­li­gent AI colonize the galaxy/​uni­verse? If not, why not?