--- Log opened Wed Nov 06 00:00:57 2019 00:03 -!- lowl_ [~luwl@2606-a000-4806-e500-7035-e75b-77fc-0bc8.inf6.spectrum.com] has quit [Remote host closed the connection] 00:03 -!- lowl_ [~luwl@2606-a000-4806-e500-7035-e75b-77fc-0bc8.inf6.spectrum.com] has joined ##hplusroadmap 00:07 -!- TheHoliestRoger [~TheHolies@unaffiliated/theholiestroger] has quit [Quit: Find me in #TheHolyRoger or https://theholyroger.com] 00:07 -!- TheHoliestRoger [~TheHolies@unaffiliated/theholiestroger] has joined ##hplusroadmap 00:08 -!- lowl_ [~luwl@2606-a000-4806-e500-7035-e75b-77fc-0bc8.inf6.spectrum.com] has quit [Ping timeout: 250 seconds] 00:19 < nsh> programmable quasicrystal growth would probably make a pretty good LTS writing mechanism 00:19 < nsh> once we figure them out a bit better 01:28 -!- lowl_ [~luwl@2606-a000-4806-e500-7035-e75b-77fc-0bc8.inf6.spectrum.com] has joined ##hplusroadmap 01:32 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Remote host closed the connection] 03:09 < fenn> there's a bit of a tradeoff between stability and energy cost to write 03:10 < fenn> just in general 03:10 < fenn> you could input the energy up front though 03:19 -!- lowl_ [~luwl@2606-a000-4806-e500-7035-e75b-77fc-0bc8.inf6.spectrum.com] has quit [] 04:06 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 268 seconds] 04:33 < fenn> http://wormspit.com/bombyxsilkworms.htm there must be a million uses for bio-engineered silkworm fabric; trace chemical detection proteins that change color, heavy metal ion binding protein, antibacterial surgical suture filament, 3d printing organoid scaffolding with tissue specific embedded growth factors 04:36 -!- uniera [uniera@gateway/vpn/privateinternetaccess/uniera] has joined ##hplusroadmap 04:36 < fenn> it seems there's a limit on how much you can muck with the silk protein because the worm must use it to pull its skin off at the various stages of larval development 04:45 < kanzure> maybe you can target different versions of the protein to be expressable during those different stages of metamorphosis 04:49 < fenn> at each instar (there are 5 i think) the worm prints a patch of silk to use as an anchor, and if the anchor comes loose it means the worm will die 04:49 < fenn> if the old skin is partially pulled off they suffocate 04:50 < fenn> if the anchor comes loose right away they won't necessarily be able to print a new one because the old anchor is stuck to the spinner region 04:50 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 265 seconds] 04:50 < kanzure> sounds like this calls for mechanical degloving 04:51 < fenn> depending on the use case it may make sense to be labor intensive, but for most uses i can think of you'd want the worms to shed their skins without any intervention 04:51 < fenn> the first and second instars are really tiny 04:58 < kanzure> wasn't there some bioreactor loom-spinner contraption that weaved spider silk out of bioreactor sludge? 04:58 < kanzure> i might be making that one up. 05:00 < kanzure> https://talktotransformer.com/ 05:07 < fenn> anything you say to transformer can and will be used against you in a future court of AI law 05:08 < fenn> you have no rights, no responsibilities, only your own conscience and aversion to pain 05:08 < kanzure> i thought the idea was that the ai would also know what you didn't say to the transformer and use that against you 05:08 < kanzure> and then simulate yourself and all your friends in hell for eternity as a consequence of not bringing about the ai apocalypse 05:09 < fenn> don't care 05:09 < kanzure> man how the hell have we not constructed a better defense to their lunacy 05:09 < fenn> i don't think that was ever a serious argument 05:09 < kanzure> i thought once that i had distilled their argument down to "i have an infinitely powerful silver bullet, now i want you to somehow overcome this infinitely powerful silver bullet" 05:10 < fenn> yeah it's frustrating 05:10 < kanzure> and arguing about things-that-by-their-definition-cannot-be-argued-against is not arguing in good faith 05:11 < fenn> well the definitions are stupid too, like "Omega, the all-knowing supernatural entity, tells you " 05:12 < fenn> wow, all i need is an all-knowing supernatural oracle, great! 05:12 < kanzure> reach into your bayesian toolbox and grab any superoracle 05:14 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 05:15 < kanzure> shouldn't i expect progress in simulation-based (not emulation-based) attempts in artificial general intelligence by now? 05:16 < kanzure> like even further simplifications of http://diyhpl.us/~bryan/papers2/neuro/Simulating%20extracted%20connectomes%20-%202017.pdf 05:16 < kanzure> or https://diyhpl.us/~bryan/papers2/neuro/cognitiveconsilience/stitched-diagram.jpg 05:18 < kanzure> or the nengo stuff https://diyhpl.us/~bryan/papers2/neuro/brain-emulation/Nengo:%20a%20Python%20tool%20for%20building%20large-scale%20functional%20brain%20models.pdf 05:19 < kanzure> ted berger (et al) has demonstrated that a large part of the hippocampus can be replaced by a prosthetic device that operates by some algorithm instead of a 10 billion neuron network 05:22 < kanzure> more on spaun/nengo http://clm.utexas.edu/compjclub/papers/Eliasmith2012.pdf 05:25 < kanzure> https://www.nengo.ai/ 05:31 -!- Dr-G [~Dr-G@unaffiliated/dr-g] has joined ##hplusroadmap 05:38 < kanzure> looks like that group caved to the deep learning craze http://compneuro.uwaterloo.ca/publications/rasmussen2018.html 05:56 < kanzure> https://www.biorxiv.org/content/biorxiv/early/2019/09/30/703777.full.pdf says "all you need is declarative long-term memory, a buffer-based working memory, a reinforcement-learning-based set of state-action patterns represented in procedural memory, and some modules for perception and motor action". 05:58 -!- uniera [uniera@gateway/vpn/privateinternetaccess/uniera] has quit [Remote host closed the connection] 06:07 < kanzure> "Breaking and (partially) fixing provably secure onion routing" https://arxiv.org/abs/1910.13772 06:08 < fenn> "shouldn't i expect progress" lol 06:09 -!- uniera [~uniera@83.223.251.130] has joined ##hplusroadmap 06:10 -!- uniera_ [uniera@gateway/vpn/privateinternetaccess/uniera] has joined ##hplusroadmap 06:10 -!- uniera_ [uniera@gateway/vpn/privateinternetaccess/uniera] has quit [Remote host closed the connection] 06:13 < archels> kanzure: what does "infuse the protective APOE2 gene into the cerebrospinal fluid of people" mean? 06:13 < kanzure> dunno, they might mean viral gene therapy injected into cerebrospinal fluid 06:13 < kanzure> or they might mean injecting a protein? 06:14 -!- uniera [~uniera@83.223.251.130] has quit [Ping timeout: 245 seconds] 06:19 < archels> yeah, not sure there 06:20 < fenn> since it says "gene" it's not a protein 06:20 < archels> I think I read somewhere that if you inject small RNA fragments into CSF or extracellular fluid, cells tend to take them up and they will perform modulatory duties 06:21 -!- helleshin [~talinck@98.29.27.253] has joined ##hplusroadmap 06:25 -!- hehelleshin [~talinck@98.29.27.253] has quit [Ping timeout: 276 seconds] 06:28 -!- balrog [~balrog@unaffiliated/balrog] has quit [Quit: Bye] 06:32 -!- balrog [~balrog@unaffiliated/balrog] has joined ##hplusroadmap 06:42 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap 07:21 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 268 seconds] 07:32 -!- sachy [~sachy@78.108.102.220] has joined ##hplusroadmap 08:03 -!- yashgaroth [~ffffffff@2606:6000:ca84:b300:4153:d835:687e:435b] has joined ##hplusroadmap 08:11 < kanzure> https://en.wikipedia.org/wiki/Bacterial_microcompartment 08:28 < kanzure> "... Controlled expression of phage nucleases cloned in Escherichia coli generated cells without DNA" 08:28 < kanzure> circularized RNA (circRNA) for in vitro protein production https://2019.igem.org/Team:GIFU_TOKAI 09:41 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 09:52 -!- purpleshift [purpleshif@gateway/vpn/mullvad/purpleshift] has joined ##hplusroadmap 09:55 -!- sektor [~sektor@95.87.234.241] has joined ##hplusroadmap 09:59 -!- bsm1175321 [~mcelrath@104.129.204.134] has joined ##hplusroadmap 10:08 -!- andytoshi [~apoelstra@wpsoftware.net] has joined ##hplusroadmap 10:08 -!- andytoshi [~apoelstra@wpsoftware.net] has quit [Changing host] 10:08 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has joined ##hplusroadmap 10:15 < kanzure> "Leveraging grammar and reinforcement learning for neural program synthesis" https://arxiv.org/pdf/1805.04276 10:20 < kanzure> "Programmatically interpretable reinforcement learning" https://arxiv.org/pdf/1804.02477.pdf 10:55 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 268 seconds] 10:56 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 11:15 -!- LeoTal [~Adium@58.170.86.79.rev.sfr.net] has joined ##hplusroadmap 11:55 -!- bsm1175321 [~mcelrath@104.129.204.134] has quit [Ping timeout: 250 seconds] 12:16 -!- spaceangel [~spaceange@ip-89-176-139-157.net.upcbroadband.cz] has joined ##hplusroadmap 12:35 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has quit [Read error: Connection reset by peer] 12:52 -!- N-time [~Mark@212.225.172.60] has joined ##hplusroadmap 13:47 -!- LeoTal [~Adium@58.170.86.79.rev.sfr.net] has quit [Quit: Leaving.] 14:04 -!- andytoshi [~apoelstra@wpsoftware.net] has joined ##hplusroadmap 14:04 -!- andytoshi [~apoelstra@wpsoftware.net] has quit [Changing host] 14:04 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has joined ##hplusroadmap 14:06 -!- spaceangel [~spaceange@ip-89-176-139-157.net.upcbroadband.cz] has quit [Ping timeout: 246 seconds] 14:39 -!- sektor [~sektor@95.87.234.241] has quit [Ping timeout: 268 seconds] 14:43 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap 14:56 -!- sachy [~sachy@78.108.102.220] has quit [Quit: Leaving.] 15:44 -!- Human_G33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 15:45 -!- Human_G33k [~HumanG33k@62.147.242.8] has quit [Remote host closed the connection] 15:45 -!- Human_G33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 15:46 -!- HumanGeek [~HumanG33k@62.147.242.8] has quit [Ping timeout: 240 seconds] 15:59 -!- Dr-G [~Dr-G@unaffiliated/dr-g] has quit [Ping timeout: 276 seconds] 16:13 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 268 seconds] 16:30 -!- crockwork [~crockwork@66.205.193.158] has quit [Quit: crockwork] 16:52 -!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap 16:58 < lsneff> Thoughts on robert zubrin's mars direct plan 2.0 that involves boosting a "mini-starship" to mars instead of going with the full deal? 17:05 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Quit: Leaving] 17:08 < kanzure> lsneff: https://github.com/kanzure/nanoengineer 17:09 < lsneff> kanzure: what does that have to do with Mars direct?? 17:10 < kanzure> i am told that you are reimplementing nanoengineer 17:25 < kanzure> "Large-scale cognitive cognitive GWAS meta-analysis reveals tissue-specific neural expression and potential nootropic drug targets" https://sci-hub.tw/10.1016/j.celrep.2017.11.028 17:26 < lsneff> kanzure: yeah, it's a long-term plan 17:30 < lsneff> I think there's a general long-term trend of research working towards the building blocks of molecular manufacturing, and the software is a major roadblock 17:31 < kanzure> graph representation of molecules is not a bottleneck 17:31 < kanzure> but it's fun to do 17:32 < lsneff> The representation isn't, but effective ways to design are (as far as I can tell, I could be wrong) 17:33 < lsneff> Taking ideas from commercial cad packages and applying them to molecular design might yield something interesting 17:35 < kanzure> nanorex (the company that made nanoengineer) folded because while their software was fun to use and you could extrude all the nanostructures that your heart desired, there was no way to build any molecular machines so nobody found it valuable 17:36 < lsneff> yeah, that's the main problem 17:36 < lsneff> you can't do anything with the structures you design 17:41 < lsneff> hopefully at some point, it will be possible to actually build the molecular machines, so perhaps exploring the design space beforehand might make things easier once that happens. 17:42 < lsneff> But other than that, it's just fun :) 17:45 < kanzure> https://www.pnas.org/content/early/2019/10/31/1910073116 17:45 < kanzure> "A single combination gene therapy treats multiple age-related diseases" https://www.pnas.org/content/early/2019/10/31/1910073116 17:54 -!- purpleshift [purpleshif@gateway/vpn/mullvad/purpleshift] has quit [Quit: .] 18:04 < maaku> lsneff: the mini-starship makes way more sense than elon's plan 18:04 < maaku> mars colonization is mostly a one-way mass transfer, so it makes sense to have a smaller earth-return vehicle 18:06 < lsneff> maaku: I suppose that makes sense, but I don't see why they should drag all this extra mass there in the first place 18:06 < lsneff> The amount of ships returning from Mars will (hopefully) be way fewer than than the number of ships headed there 18:06 < lsneff> So optimizing for the return journey doesn't make sense in my eyes 18:09 -!- N-time [~Mark@212.225.172.60] has quit [Ping timeout: 240 seconds] 18:14 < lsneff> And I'm not confident a smaller starship would mean less mass leaving mars. Presumably there will be enough people who want to leave at some point to fill up many starships, and a small starship just has less payload/fuel ratio 18:15 < kanzure> https://worldbuilding.stackexchange.com/questions/23145/how-to-make-a-flying-human 18:17 -!- juri__ [~juri@79.140.121.40] has joined ##hplusroadmap 18:21 < kanzure> https://www.greaterwrong.com/posts/Wa2hASzbxyvutHJff/total-horse-takeover 18:21 -!- juri_ [~juri@178.63.35.222] has quit [Ping timeout: 246 seconds] 18:25 < fenn> there will be lots of cargo delivered to any mars colony, and it shouldn't need a fully reusable starship when tossing a cargo pod to mars with a starship that returns to a highly elliptical orbit will do 18:25 < fenn> this was zubrin's original refinement 18:25 < fenn> haven't seen the new version yet so i'll abstain from commenting yet 18:26 < fenn> you'll want propulsive landing for any passengers 18:27 < kanzure> they don't deserve it 18:27 < fenn> nuh uh 18:28 < lsneff> I believe elon specified that most starships headed towards mars would be stripped down because they only need to reenter once 18:29 < fenn> that would surprise me 18:29 < fenn> it goes against everything he's said about the reusable mars spacecraft concept so far 18:30 < fenn> maybe for the first few ships 18:31 < lsneff> I imagine some would be fully reusable for the return journey and for additional trips around the system from mars, but there's little reason to drag a fully reusable heat shield and whatnot there if it's only used once 18:31 < fenn> the plan is to reuse it 18:31 < fenn> up to 10 times 18:31 < fenn> by which time it will be obsolete 18:33 < lsneff> Yes, and it will be on many flights, like for refueling ships, but I suspect many initial starships will be stripped down and used a material or used as living space, so they'll never be reused 18:33 < lsneff> They're very cheap to build, so it makes less sense to keep shipping them back when we can just build more here 18:34 < fenn> sure you could even strip the engines off and send them back in bulk 18:34 < lsneff> I suspect most starships used for lunar operations will be fully reusable though 18:34 < lsneff> Yeah, engines might be the bottleneck 18:35 < fenn> i don't think starship is a very good fit for moon landing 18:35 < lsneff> even once there are landing pads set up? 18:35 < kanzure> maaku or lsneff is supposed to ask fenn about CAM things 18:36 < kanzure> i forget which one 18:36 < fenn> more realistically it would deliver payload to a moon lander, which itself could be a starship, but packed so full of payload that the aerodynamic/heatshield mass overhead is insignificant in comparison 18:36 < lsneff> Actually yeah, that makes sense, ship mass to lunar orbit with a starship, and then move mass between luna and lunar orbit with a more specialized vehicle 18:36 < fenn> but the bulk of traffic would be flights to some intermediate destination, rather than the moon directly 18:37 < kanzure> maaku: https://diyhpl.us/~bryan/papers2/cad/ 18:38 < fenn> a week ago i'd have said a moon lander would run on hydrogen, but i recently learned that LCROSS plume spectra showed rather large concentrations of short hydrocarbons and carbon monoxide 18:39 < lsneff> Ah, I've looked into csg modeling quite a bit 18:39 < lsneff> fenn: so methane is the way to go once again? 18:40 < fenn> well it's certainly not off the table 18:40 < fenn> we simply don't know enough about the moon yet 18:40 < lsneff> soon hopefully 18:41 < lsneff> I remember hearing that a space elevator on the moon is feasible with current materials 18:41 < fenn> http://www.universetoday.com/76365/understanding-the-unusual-lcross-ejecta-plume/ the article i had read 18:43 < fenn> also http://www.universetoday.com/76329/water-on-the-moon-and-much-much-more-latest-lcross-results/ 18:50 < lsneff> https://en.wikipedia.org/wiki/Lunar_space_elevator 18:50 < lsneff> It appears that modern materials could raise small payloads 18:51 < lsneff> and future materials could be significantly better 18:53 < fenn> oh no 18:54 < fenn> tethers.com got rid of their publications entirely 18:55 < superkuh> That's a shame. Still got a deorbit tether page. 18:58 < lsneff> the cable for a lunar space elevator from lunar surface to L1 would only take up 42.3 cubic meters of volume 18:59 < fenn> lsneff: http://web.archive.org/web/20190213154044/http://www.tethers.com/papers/HOYT_MMOSTT_Final.pdf 19:00 < fenn> you can pick up bags of dirt from the surface of the moon with a rotating tether and drop it onto a tether in low earth orbit to reboost it to help get a suborbital rocket to earth orbit or beyond 19:00 < lsneff> Like a skyhook sort of? 19:01 < fenn> yes but with more degrees of freedom 19:01 < lsneff> interesting 19:01 < fenn> maybe this is not the best intro paper but it covers the moon use case 19:01 < lsneff> I imagine that people will be trying all of these ideas in a decade or two 19:01 < fenn> hah 19:02 < fenn> i imagine that rotating tethers will remain in obscurity forever despite their obvious usefulness 19:02 < lsneff> why is that? 19:02 < fenn> because the concept is not new and everything is terrible 19:02 < lsneff> right 19:02 < lsneff> Do you think a lunar space elevator will happen? 19:03 < fenn> many of these tether papers were written in the 1980s with only minimal progress 19:03 < kanzure> jrayhawk: having some serious disk io issues 19:04 < lsneff> It's difficult to make progress when you can't actually build it because you can't go there to start buildingit 19:04 < fenn> lsneff: if japan gets serious about space, then yeah 19:04 < fenn> nobody else seems to care about space elevators 19:04 < lsneff> I find that difficult to believe 19:05 < lsneff> They just haven't really been feasible until now 19:05 < fenn> that's not true. any help you can give a rocket as an extra stage helps exponentially 19:06 < fenn> skyhooks are much more well known, for some reason 19:06 < fenn> anyway you can do a lot with just a kevlar tether 19:07 < lsneff> yeah 19:07 < fenn> a space elevator is just a special case of a skyhook 19:08 < kanzure> http://diyhpl.us/~bryan/papers2/space/tethers.com/ 19:08 -!- adlai [~adlai@unaffiliated/adlai] has quit [Ping timeout: 245 seconds] 19:09 < jrayhawk> kanzure: ugh, sorry. i have a couple disks i will shove in there to sort that out. 19:09 < kanzure> it's technically working 19:09 < kanzure> so... there's that. 19:09 < fenn> kanzure: i have a bunch more papers from them; want? 19:09 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has quit [Ping timeout: 264 seconds] 19:09 < kanzure> it's grabbing them slowly 19:10 < kanzure> i figured a delay would be nice because i don't want the internet archiving hoarder gods to smite me 19:10 < lsneff> How does one bring one end of the cable from orbit to the surface? I imagine that would be quite difficult 19:10 < fenn> of a body with atmosphere or without? 19:10 < lsneff> without? 19:11 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has joined ##hplusroadmap 19:11 < lsneff> do you just trust gravity with guiding rcs? 19:11 < fenn> imagine a giant wheel rolling along the surface with its center of mass in orbit 19:11 < fenn> now cut away all the parts of the wheel except for one spoke and the hub 19:11 < kanzure> ok that's all it found, if you refresh 19:12 < lsneff> How do you unspool that much cable with tangling? 19:12 < lsneff> s/with/without 19:14 < fenn> kanzure: ok that seems about right 19:15 < fenn> lsneff it's under tension from tidal stresses, or you can rotate it for centrifugal force 19:15 < lsneff> Ah, I see 19:16 -!- adlai [~adlai@unaffiliated/adlai] has joined ##hplusroadmap 19:18 < fenn> here's a rather overdone video that shows off the concept of a rotating momentum exchange tether, may want to forward to 3 minutes to see the actual momentum exchange http://youtu.be/mPx1Nq80jm8 19:20 < fenn> this is the simple easy and dumb version that doesn't require much research 19:20 < fenn> but it will get you from LEO to GTO 19:34 < lsneff> why are videos like that always way to long 19:34 < lsneff> s/to/too 20:42 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 245 seconds] 20:52 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 21:18 -!- LeoTal [~Adium@58.170.86.79.rev.sfr.net] has joined ##hplusroadmap 21:38 -!- preview [~quassel@2407:7000:8423:b75:8712:cc92:2cdd:eba6] has joined ##hplusroadmap 22:07 -!- yashgaroth [~ffffffff@2606:6000:ca84:b300:4153:d835:687e:435b] has quit [Quit: Leaving] 22:11 -!- ebowden_ [~ebowden@unaffiliated/ebowden] has quit [Ping timeout: 245 seconds] 22:14 -!- ebowden [~ebowden@unaffiliated/ebowden] has joined ##hplusroadmap 22:55 -!- preview [~quassel@2407:7000:8423:b75:8712:cc92:2cdd:eba6] has quit [Ping timeout: 264 seconds] 23:21 -!- kuldeep [~kuldeep@unaffiliated/kuldeepdhaka] has quit [Remote host closed the connection] 23:31 -!- spaceangel [~spaceange@ip-89-176-139-157.net.upcbroadband.cz] has joined ##hplusroadmap 23:31 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Ping timeout: 240 seconds] 23:43 -!- preview [~quassel@2407:7000:8423:b75:8712:cc92:2cdd:eba6] has joined ##hplusroadmap 23:47 < maaku> kanzure: we're supposed to put lsneff in contact with fenn because lsneff is working on a nanoengineer-like thing 23:47 < maaku> lsneff needs a data model for his nanoCAD program, and preferably one that is CAM-oriented in terms of specifying a manufacturing pathway 23:49 < maaku> I believe fenn has done work or at least thought about the problem of creating a "manufacturing language" for specifying how to construct mechanical things, which seems highly related --- Log closed Thu Nov 07 00:00:01 2019