--- Log opened Fri May 27 00:00:12 2016 | ||
-!- mf1008 [~mf1008@unaffiliated/mf1008] has quit [Quit: Cave quid dicis, quando, et cui] | 00:02 | |
-!- mf1008 [~mf1008@unaffiliated/mf1008] has joined ##hplusroadmap | 00:10 | |
fenn | all this fussing about the cost of producing large quantities 178m2Hf is pointless, there are other nuclear isomers, and 180Ta is naturally occurring | 00:27 |
---|---|---|
fenn | the main source of controversy/disinformation is over whether stimulated gamma emission works at all | 00:28 |
fenn | hafnium was chosen because it's predicted to be the easiest to trigger | 00:28 |
fenn | that's it | 00:28 |
fenn | it's like saying "helium-3 is so expensive, we'll never achieve fusion power" | 00:29 |
JayDugger | A hand-grenade sized nuke has non-military applications. | 00:35 |
fenn | also it could be a rocket or a drone | 00:38 |
fenn | i think they were more interested in the fact that you could dial the power down to whatever you wanted for a given situation | 00:39 |
fenn | so you could decide to blow up the entire city block, or just a building | 00:39 |
JayDugger | Depressing thoughts. | 00:40 |
fenn | really the last thing this military needs is more expensive bombs | 00:40 |
fenn | we're already sending $20m cruise missiles at goat herders | 00:40 |
JayDugger | I'll disagree. A credible nuclear deterrent "seems" to have prevented a major power war since 1945. | 00:41 |
fenn | sure, but what has the F-35 accomplished | 00:41 |
JayDugger | Um...wow, where to start? | 00:41 |
JayDugger | Kept LM running? | 00:41 |
fenn | lol | 00:42 |
JayDugger | Taught Boeing to switch to drones? | 00:42 |
JayDugger | Drove McDonnell-Douglas out of business? | 00:42 |
JayDugger | Provided an impetus for the next generation of airborne high-power microwave and directed energy weapons? | 00:42 |
fenn | i don't follow | 00:43 |
fenn | "Lockheed is studying integrating a fiber laser onto the aircraft that uses spectral beam combining to channel energy from a stack of individual laser modules into a single, high-power beam, which can be scaled up or down for various levels of effects. Adding a laser would give the F-35 the ability to essentially burn missiles and other aircraft out of the sky" | 00:43 |
JayDugger | When I worked in the defense side of aerospace, one of the rumors about why LM's design won was-- | 00:43 |
fenn | still i don't really see what that has to do with the f-35 in particular | 00:44 |
JayDugger | The F35 has a certain amount of prestige as the "next generation" fighter, which attracts other, less glamorous projects, to ride on its coattails. | 00:44 |
JayDugger | Among these, directed energy weapons and high-power microwaves, which haven't always enjoyed their current high point in the defense industry's versino of the hype cycle. | 00:45 |
fenn | well it seems like they're actually useful now | 00:46 |
JayDugger | I have also heard that the F35's winning design had abundant on-board electrical power, and that this counted in its favor during the competition. | 00:46 |
JayDugger | I agree. It seems so. No one talks about combat use, so that caps it. | 00:47 |
JayDugger | I am a little surprised that the US Navy hasn't done more, given they've war machines powered by multiple fission reactors that have an entire ocean to use a heat sink. | 00:48 |
fenn | busy fighting land wars in asia | 00:49 |
JayDugger | At a guess, they've got hands full with rail guns and the other new ship class, and haven't the attention to spare for reintroducing another class of capital ships. | 00:49 |
fenn | yeah they are focusing on "all electric ship" fleets | 00:49 |
JayDugger | That will make a nice advantage. | 00:50 |
fenn | i guess | 00:50 |
fenn | the main threat seems to be china's super mega death rockets | 00:50 |
fenn | anti-carrier missiles | 00:50 |
JayDugger | Not really, no. | 00:51 |
JayDugger | I think, and I don't know. It seems more for domestic consumption on both sides. | 00:51 |
fenn | hm ok | 00:51 |
fenn | i don't really understand naval warfare | 00:51 |
JayDugger | You aren't missing much. If you're really interested, there's educational games for the topic at least as good for learning it as Kerbal Space Program is for its topic. | 00:52 |
JayDugger | But back to the F35, that sure is one (way too) expensive plane. | 00:53 |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-upszxeneasoyrwiy] has joined ##hplusroadmap | 00:56 | |
fenn | PatrickRobotham: is that your real name? | 00:57 |
PatrickRobotham | yes | 00:57 |
fenn | i want to make a robotic hamlet some day | 00:57 |
fenn | .wik Hamlet_(place) | 00:58 |
yoleaux | "A hamlet is a type of settlement. The definition of hamlet varies by country. It usually refers to a small settlement, with a small population that is usually under 100, in a rural area, or a component of a larger settlement or municipality. Hamlets are typically unincorporated communities." — https://en.wikipedia.org/wiki/Hamlet_(place) | 00:58 |
-!- Gurkenglas [Gurkenglas@dslb-178-000-182-226.178.000.pools.vodafone-ip.de] has joined ##hplusroadmap | 01:13 | |
-!- Gurkenglas [Gurkenglas@dslb-178-000-182-226.178.000.pools.vodafone-ip.de] has quit [Ping timeout: 246 seconds] | 01:39 | |
nmz787_i | I just think of Dr. Robotnic, from Sonic the Hedgehog | 01:39 |
nmz787_i | sorry: Robotnik | 01:40 |
-!- nmz787_i [~ntmccork@134.134.139.78] has quit [Quit: Leaving.] | 01:47 | |
-!- jcorgan [~jcorgan@unaffiliated/jcorgan] has quit [Quit: Lost terminal] | 02:20 | |
fenn | time to break out the phage http://www.cnn.com/2016/05/26/health/first-superbug-cre-case-in-us/index.html | 03:11 |
-!- sandeepkr [~sandeep@111.235.64.4] has joined ##hplusroadmap | 03:40 | |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-upszxeneasoyrwiy] has quit [Quit: Connection closed for inactivity] | 03:50 | |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has quit [Ping timeout: 260 seconds] | 04:10 | |
-!- jtimon [~quassel@4.28.134.37.dynamic.jazztel.es] has joined ##hplusroadmap | 04:39 | |
-!- ArturSha1 [~ArturShai@212.97.22.183] has quit [Ping timeout: 258 seconds] | 04:40 | |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-afpmaxqekkcndfuf] has joined ##hplusroadmap | 04:56 | |
-!- ArturSha1 [~ArturShai@31.29.0.57] has joined ##hplusroadmap | 05:00 | |
-!- Orpheon [~Orpheon@7.152.62.81.dynamic.wline.res.cust.swisscom.ch] has joined ##hplusroadmap | 05:12 | |
kanzure | hmph | 05:46 |
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 240 seconds] | 05:55 | |
-!- Aurelius_Work2 [~cpopell@209.48.69.2] has quit [Ping timeout: 276 seconds] | 05:59 | |
-!- Aurelius_Work [~cpopell@209.48.69.2] has joined ##hplusroadmap | 06:06 | |
FourFire | kanzure, do you know of research into microbial resonance frequencies? | 06:11 |
kanzure | no | 06:12 |
archels_ | http://www.cyborgnest.net/ | 06:12 |
kanzure | no | 06:14 |
archels_ | why not | 06:15 |
-!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 260 seconds] | 06:40 | |
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap | 06:51 | |
-!- Aurelius_Work2 [~cpopell@209.48.69.2] has joined ##hplusroadmap | 07:06 | |
-!- Aurelius_Work [~cpopell@209.48.69.2] has quit [Ping timeout: 260 seconds] | 07:09 | |
-!- jcorgan [~jcorgan@unaffiliated/jcorgan] has joined ##hplusroadmap | 07:23 | |
-!- strangewarp [~strangewa@c-76-25-206-3.hsd1.co.comcast.net] has quit [Read error: Connection reset by peer] | 07:36 | |
-!- strangewarp [~strangewa@c-76-25-206-3.hsd1.co.comcast.net] has joined ##hplusroadmap | 07:37 | |
catern | seems inferior to a northpaw | 07:48 |
-!- augur [~augur@2602:304:cdac:e260:9402:8177:386d:700d] has joined ##hplusroadmap | 07:59 | |
maaku | Very disappointed that it only vibrates | 08:56 |
maaku | Has there been any Research into a surface with neural affinity? | 08:56 |
-!- Gurkenglas [Gurkenglas@dslb-178-000-182-226.178.000.pools.vodafone-ip.de] has joined ##hplusroadmap | 08:57 | |
-!- nmz787_i [~ntmccork@134.134.139.78] has joined ##hplusroadmap | 09:03 | |
-!- Orpheon [~Orpheon@7.152.62.81.dynamic.wline.res.cust.swisscom.ch] has quit [Remote host closed the connection] | 09:03 | |
maaku | a wonderful description of biology techniques : http://www.cell.com/cancer-cell/fulltext/S1535-6108%2802%2900133-2 | 09:22 |
-!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap | 09:23 | |
-!- augur [~augur@2602:304:cdac:e260:9402:8177:386d:700d] has quit [Remote host closed the connection] | 09:28 | |
-!- nmz787_i1 [~ntmccork@134.134.139.78] has joined ##hplusroadmap | 09:29 | |
-!- nmz787_i [~ntmccork@134.134.139.78] has quit [Remote host closed the connection] | 09:29 | |
-!- sandeepkr_ [~sandeep@111.235.64.4] has joined ##hplusroadmap | 09:33 | |
-!- sandeepkr__ [~sandeep@111.235.64.4] has joined ##hplusroadmap | 09:34 | |
-!- sandeepkr__ [~sandeep@111.235.64.4] has quit [Read error: Connection reset by peer] | 09:34 | |
nmz787_i1 | maaku: I am glad you found that... it was a good read when I found it ~4 years ago (in a history of radio humanitites-class) | 09:37 |
maaku | yeah it's actually quite educational about how various biological techniques work | 09:37 |
-!- sandeepkr [~sandeep@111.235.64.4] has quit [Ping timeout: 244 seconds] | 09:37 | |
-!- sandeepkr__ [~sandeep@111.235.64.134] has joined ##hplusroadmap | 09:38 | |
-!- sandeepkr_ [~sandeep@111.235.64.4] has quit [Ping timeout: 246 seconds] | 09:38 | |
-!- sandeepkr__ [~sandeep@111.235.64.134] has quit [Remote host closed the connection] | 09:38 | |
-!- bsm1175321 [~mcelrath@38.121.165.30] has quit [Ping timeout: 240 seconds] | 09:41 | |
-!- ArturSha1 [~ArturShai@31.29.0.57] has quit [Ping timeout: 264 seconds] | 10:10 | |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has joined ##hplusroadmap | 10:13 | |
-!- bsm1175321 [~mcelrath@38.121.165.30] has joined ##hplusroadmap | 10:18 | |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has quit [Remote host closed the connection] | 10:21 | |
-!- bsm1175321 [~mcelrath@38.121.165.30] has quit [Read error: Connection reset by peer] | 10:28 | |
maaku | xentrac_: Moravec's paradox is interpreted too narrowly, even by Moravec himself I think | 10:30 |
maaku | the correct observation is that there is only tenuous corrolation between how difficult something seems, and how difficult it actually is, because things our complex brains are good at seem simple, literally intuitive to us | 10:31 |
maaku | so e.g. sensorymotor is actually moderately hard problem whereas forward logical inference (the math we take literally decades to teach our kids) can be done with simple machines at superhuman speeds | 10:32 |
maaku | machine learning is good, but I think the key innovation is/will be integration of knowledge. and people are just starting to look in that direction in the machine learning space (e.g. thought vectors out of Google X) | 10:33 |
maaku | jcorgan: this is much like we were talking about in zurich | 10:33 |
xentrac_ | maaku: I think that's a much narrower interpretation than the one I'm familiar with, although I couldn't swear my interpretation is Moravec's | 10:39 |
xentrac_ | a broader interpretation is that there's a fairly strong correlation between how difficult something seems and how difficult it actually is (in a computer) but it's negative | 10:40 |
xentrac_ | well, I guess that only applies to things that people actually do | 10:40 |
maaku | xentrac_: well i think we're saying the same thing so there's no need to get down to definitions | 10:40 |
xentrac_ | (there are things that computers do, but people don't, that seem difficult and are difficult) | 10:41 |
xentrac_ | so, you could be right that there are still things that are going to be difficult | 10:41 |
maaku | there is also now very few things, if any, that people do which computers cannot | 10:42 |
maaku | the problem is just doing one program that does all the things | 10:42 |
maaku | hence the important I think on integration | 10:42 |
xentrac_ | yeah, that might turn out to be difficult, given that we don't know how to do it | 10:42 |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has joined ##hplusroadmap | 10:48 | |
maaku | i don't think so. google "thought vectors" | 10:49 |
jcorgan | maaku: my intuition is that integration of these narrow(er) modules would still on result in something like the human unconscious, and there would remain the issue of how to generate goal directed behavior | 10:50 |
maaku | jcorgan: planning engines? | 10:50 |
maaku | er, there should have been no question mark. that's my answer. | 10:50 |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has joined ##hplusroadmap | 10:51 | |
maaku | I'm very confident that "goal directed behavior" reduces down a planning engine. | 10:51 |
jcorgan | that sounds like GOFAI | 10:51 |
-!- esmerelda [~andares@unaffiliated/jacco] has joined ##hplusroadmap | 10:55 | |
maaku | jcorgan: I think the distinction between GOFAI vs connectionist isn't helpful. apples to oranges. | 10:55 |
xentrac_ | I suspect it won't turn out to be difficult, but it could | 10:55 |
kanzure | it would be useful to have a million brains sliced and scanned for connectomics and antibody labeling reasos | 10:56 |
kanzure | *reasons | 10:56 |
xentrac_ | my point is that reaching superhuman performance on ImageNet with convnets means we've overcome the Moravec obstacles | 10:56 |
maaku | You can have a planning engine that runs on neural circuits. That's probably what our prefrontal cortex is. | 10:56 |
jcorgan | unfortunately our only actual example is a result of a brutal reinforcement learning regime resulting in life or death :) | 10:56 |
kanzure | why is that unfortunate? we can poke and prod brains as much as we like. | 10:56 |
maaku | sure, but start with c. elegans (and people are working on this) | 10:57 |
kanzure | yeah about that, what was the reason for starting with c. elegans anyway | 10:57 |
maaku | well understood | 10:58 |
kanzure | iirc the dynamics are still not figured out, but i'm not sure they need to be in the first place for those 301 neurons.. | 10:58 |
maaku | a 'standard' c. elegans has a fixed number of cells, fully categorized and numbered, simple dynamics and behavior, etc. | 10:58 |
maaku | the dynamics are not fully figured out, but yes that's the point -- upload a c. elegans is a maneageable, <5yr project | 10:59 |
xentrac_ | the lobsters will take longer :) | 11:00 |
-!- nmz787_i1 [~ntmccork@134.134.139.78] has quit [Quit: Leaving.] | 11:23 | |
-!- nmz787_i [~ntmccork@134.134.139.70] has joined ##hplusroadmap | 11:36 | |
-!- nmz787_i [~ntmccork@134.134.139.70] has quit [Client Quit] | 11:37 | |
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-sbiliwzqmhthezbe] has quit [Quit: faggory daggory doo!] | 11:39 | |
-!- nmz787_i [ntmccork@nat/intel/x-vlbzdicgjqwstpig] has joined ##hplusroadmap | 11:40 | |
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-wrybrfmyvrfclgwb] has joined ##hplusroadmap | 11:45 | |
kanzure | markram's statistical minicolumn neural wiring thingy could probably be extended to fix bad scan data from connectomics scanner stuff... | 11:46 |
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-wrybrfmyvrfclgwb] has quit [Changing host] | 11:55 | |
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has joined ##hplusroadmap | 11:55 | |
-!- anoncicada [anoncicada@unaffiliated/anarcanon] has quit [Changing host] | 11:55 | |
-!- anoncicada [anoncicada@gateway/shell/elitebnc/x-wrybrfmyvrfclgwb] has joined ##hplusroadmap | 11:55 | |
kanzure | and evolutionary search should be applied to connectomics results as well. question is whether there's relevant information from scan data that can be had that wouldn't be possible to generate otherwise (because otherwise you would just skip the scan step in the first place). | 11:59 |
kanzure | (and once acquiring scan data, evolutionary search from that starting point can be somewhat more productive since "scan data does not include enough data about ion channels and other components to work immediately for an emulation but perhaps nearby in design space you can find related designs that probably work" since brains aren't tightly specified by genomes anyway) | 12:00 |
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has joined ##hplusroadmap | 12:08 | |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has quit [Ping timeout: 260 seconds] | 12:14 | |
-!- nmz787_i [ntmccork@nat/intel/x-vlbzdicgjqwstpig] has quit [Quit: Leaving.] | 12:18 | |
-!- nmz787_i [~ntmccork@134.134.137.71] has joined ##hplusroadmap | 12:30 | |
-!- nmz787_i [~ntmccork@134.134.137.71] has quit [Client Quit] | 12:32 | |
-!- nildicit [~nildicit@unaffiliated/nildicit] has quit [Ping timeout: 252 seconds] | 12:44 | |
-!- jtimon [~quassel@4.28.134.37.dynamic.jazztel.es] has quit [Ping timeout: 244 seconds] | 12:45 | |
-!- nmz787_i [ntmccork@nat/intel/x-uoqpalmshjsumtip] has joined ##hplusroadmap | 12:48 | |
-!- nmz787_i [ntmccork@nat/intel/x-uoqpalmshjsumtip] has quit [Client Quit] | 12:53 | |
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-sclpbpmqgqumpbab] has joined ##hplusroadmap | 13:04 | |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has quit [Remote host closed the connection] | 13:18 | |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has joined ##hplusroadmap | 13:19 | |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has joined ##hplusroadmap | 13:21 | |
maaku | http://smbc-comics.com/comics/1464275028-20160526.png | 13:23 |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has quit [Ping timeout: 250 seconds] | 13:23 | |
maaku | "Once you realize there is no hope, you can relax and just enjoy the progress in machine learning" | 13:23 |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has quit [Ping timeout: 260 seconds] | 13:26 | |
kanzure | yeah i think people overestimate the extent to which humans are not interested in kill all humans | 13:35 |
kanzure | and for some reason they get worried about an ai killing everyone, meanwhile we seem to have people that want to do that anyway. so blaming this problem on ai is pretty dumb. | 13:36 |
maaku | we've got a pretty good track record for killing all humans | 13:37 |
maaku | the 20th century was great for that | 13:37 |
c0rw1n | not nearly all humans, obviously | 13:37 |
kanzure | i suppose, in general, you could argue that most of the humans who have ever lived are now dead (~100 billion dead) so yes we do pretty well at this | 13:37 |
c0rw1n | that's under 95% of all humans ever, as we're likr 7Bn now ... so death is <95% sure so far, which is probably good news | 13:39 |
jcorgan | it's something like 6% of all humans who have ever lived are still alive today | 13:39 |
kanzure | i was really disappointed by the recent "How to design a malevolent AI" research paper. it was not a howto guide at all. | 13:40 |
jcorgan | is the group here familiar with the AI-Foom debate? | 13:43 |
c0rw1n | yes | 13:43 |
jcorgan | (i'm new) | 13:43 |
-!- nmz787_i [~ntmccork@134.134.139.78] has joined ##hplusroadmap | 13:51 | |
kanzure | jcorgan: here's what we do http://diyhpl.us/wiki/hplusroadmap | 13:59 |
maaku | jcorgan: yes, quite familiar | 14:06 |
maaku | jcorgan: in my own opinion the situation is complicated, because most people's kneejerk "that's impossible" reaction to AI xrisk is of course wrong | 14:07 |
maaku | but EY's instant-FOOM position is just as absurdly extreme in the other direction | 14:07 |
maaku | AI xriskers have reason to be worried, but not over the hollywood risk scenarios presented in their own literature | 14:08 |
maaku | jcorgan: If I come across someone who's view of FOOM is influenced by EY's "That Alien Message", I have so far always been able to talk them out of it by going over all the things that go on in a single cognitive cycle | 14:14 |
maaku | and then back-of-the-envelope checking how many FLOPS are actually available for an imaginable first-gen AGI supercomputer | 14:15 |
maaku | realizing that the proper timescale for such risk is weeks, months, or years, not microseconds usually tempers people's concerns... | 14:16 |
maaku | (this also works as an argument for pursuing strong ai ASAP before hardware gets better/cheaper/ubiquitous) | 14:16 |
kanzure | annnnd others are perfectly alright with replacing most things with computronium if the result is generally better in at least a handful of dimensions | 14:17 |
-!- strangewarp [~strangewa@c-76-25-206-3.hsd1.co.comcast.net] has quit [Read error: Connection reset by peer] | 14:17 | |
jcorgan | we're in the slow zone anyway, right? | 14:17 |
maaku | oh how i wish there was a beyond + transcend | 14:18 |
kanzure | yes, ever since hal finney's signal was dropped | 14:18 |
kanzure | (carrier error) | 14:18 |
maaku | sadest part of the universe is that there are so many distant regions we won't be able to fill with computronium before expansion of the universe gets us... | 14:21 |
andytoshi | welcome jcorgan | 14:22 |
kanzure | maaku: what about with giant laser writing by remote chemistry via atmospheric mixing etc | 14:22 |
kanzure | i guess we have better things to do with all that energy in the mean time | 14:23 |
maaku | huh. interesting | 14:24 |
maaku | well the vast majority of the visible universe is already outside of our future light cone | 14:24 |
andytoshi | it could slow down again, we don't have a good idea of why it's accelerating. fwiw | 14:24 |
maaku | we can see what happened there in the distant past, but they will never receive our light (and we can never visit) | 14:24 |
andytoshi | (the expansion) | 14:24 |
maaku | andytoshi: perfectly valid point | 14:24 |
maaku | and there is the unlikely possibility of worm holes, etc. | 14:25 |
maaku | correction: we don't have a fucking clue why it is accelerating ;) | 14:25 |
andytoshi | another problem though is that once the stars burn out it will be pretty hard to compute .. even if stuff is physically within our light cone maybe we won't have sufficient energy density by the time it's accessible to really use it | 14:25 |
andytoshi | maaku: hah, i was gonna say that but every time i say something like that here i get nailed by "nobody knows what nobody knows" | 14:26 |
kanzure | wait is there really no speculative technology for large-scale low-density computation? | 14:26 |
andytoshi | kanzure: i'm pretty sure it's possible even with existing tech. it'd just be slow | 14:27 |
andytoshi | in principle you can do computations with zero energy as long as it's reversible | 14:27 |
kanzure | slow is a matter of perspective in the 10^100 years after we run out of stars | 14:27 |
andytoshi | ah, this is very true | 14:27 |
maaku | there is a (currently unsolved?) question as to whether the absolute number of thoughts converges if you slow clock rates as available energy lowers | 14:28 |
maaku | it was one of Hal's interests and I remember him writing about it once | 14:28 |
maaku | if it does not converge then problem solved -- simulations get slower relative to absolute time, but we also have forever | 14:29 |
maaku | of course this is speculating over computronium as we imagine it now.. what if consciousness can exist as a fully reversible process? | 14:32 |
maaku | then the end state could be a steady-state reversible computronium with no energy loss at 0K | 14:32 |
kanzure | i thought we agreed that was one of the taboo words | 14:34 |
kanzure | maybe it was another word we agreed to taboo though | 14:34 |
maaku | yeah well I'd need a word for whatever-it-is | 14:35 |
maaku | in this case do an uncontroversial s/consciousness/thinking process/ | 14:36 |
kanzure | i don't have a link to that hal finney thing, but i'm pretty sure it was in an email | 14:38 |
kanzure | "vixra (vixra.org), but they also (mostly) have the "funny" stuff." | 14:47 |
kanzure | (as an alternative to arxiv for uh i guess the accusation is crackpots) | 14:47 |
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap | 14:50 | |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-afpmaxqekkcndfuf] has quit [Quit: Connection closed for inactivity] | 14:50 | |
maaku | From the (un)common sense department: Somebody should start a charity for the purpose of cryopreserving fluent speakers and oral historians of dying/near-extinct languages | 15:01 |
maaku | Scope could be increased to include other people of global cultural interest, but extinct languages seems to maximally combine poor indigenous communities that can't afford Alcor with density of knowledge encoded in those brains | 15:05 |
maaku | I guess brain plastination would also work | 15:05 |
andytoshi | maaku: i wonder how most indigenous cultures would react to the idea of crypopreservation | 15:17 |
maaku | Not well I would imagine. But it would be worth a shot. | 15:18 |
maaku | I actually suspect the larger issue would be interment in Arizona vs whatever local burial ground & tradition. | 15:19 |
maaku | Plastination is a little better in that respect, although I don't know if there are any good/tested protocols for that yet | 15:19 |
andytoshi | yeah. if it was just about the person in question arguably we could just wake them up in the future with no warning | 15:20 |
andytoshi | but obvs the community will care what happens with the body | 15:20 |
maaku | Neuralpreservation would allow them to bury the body, as well as be cheaper for the charity. | 15:21 |
maaku | But I don't imagine that "some white guys want to come decapitate our elder" will go over very well. | 15:21 |
andytoshi | exactly | 15:21 |
andytoshi | i think it's perfectly reasonable to convince a couple people this is a good idea (in particular the sorts well-educated people we'd want to preserve). but not a whole community | 15:22 |
andytoshi | reasonable to expect to be able to* | 15:22 |
maaku | You'd probably need a half dozen or so first cases that aren't as culturally sensitive -- people with knowledge worth preserving but not a central role in the community, preferably with family that is integrated or at least well educated in the non-traditional culture | 15:23 |
maaku | wouldn't need the whole community. just that one grandma who remembers all the oral history stories her father told every friday at the community center | 15:23 |
andytoshi | in canadian first nations you'd find a few examples. there is certainly a lot of "fuck whitey" sentiment but also people are aware of modern ideas and culture, and there is a fairly urgent feeling (among people who care) that languages need to be preserved somehow | 15:24 |
andytoshi | and songs, stories, etc | 15:24 |
maaku | Right now there are a lot of grad students and post docs at e.g. Stanford linguistics that go around with microphones to remote villages and desperately try to get a few dozen hours of audio recording | 15:25 |
maaku | Just so that we have enough to classify these languages to an extent that we can write a one-paragraph about them rather than a single sentence in the encyclapedia. | 15:26 |
andytoshi | eek. | 15:26 |
maaku | But preserving just one brain.. you've got a fully fluent adult + memories, probably the equivalent of 10,000+ hours of audio | 15:26 |
andytoshi | http://www.cbc.ca/news/canada/once-vibrant-aboriginal-languages-struggle-for-survival-1.1173659 | 15:27 |
andytoshi | this is not nearly as bad as your examples, but "The situation is bleakest, however, in British Columbia, which is home to more than half of Canada's native languages. All are in danger of disappearing, with only about one in 20 aboriginals in the province still fluent — almost all of them elders" | 15:27 |
maaku | It makes just plain old economic sense if nothing else. A fixed ~$100k cost per preserved elder, vs years of effort put in by grad students & grant programs | 15:28 |
andytoshi | but also, this would be more politically feasible because these are not remote tribes, they're people who speak english or french (or know people who do) and communicate with the universities | 15:28 |
TMA | maaku: that's still contingent on the ability to extract the data from there; can we be reasonably sure, that it will be possible from the specimens stored with current technology? | 15:29 |
maaku | TMA: yes. | 15:30 |
maaku | I mean there's really not much doubt. The full connectome is there. Things are literally frozen (ok, vitrified) in place. | 15:30 |
maaku | Brain plastination? I don't know. | 15:31 |
maaku | brain plastination preserves the connectome, but the connectome isn't everything. | 15:32 |
maaku | But vitrification (e.g. Alcor) would certainly work. | 15:32 |
maaku | Uncertainty in cryonics is over whether it would be medically possible to extract a viable human at the other end. | 15:34 |
maaku | But in terms of information preservation, it certainly does work. | 15:35 |
TMA | maaku: the neurons are said to be simple "fire when sum inputs is over threshold" [or so I have read] -- is the threshold setting preserved? in a separate meme/fact/infochunk I have been informed that the memory encoding seems to be related with cytosin methylation in selected neurons; other theories I have heard are even more exotic | 15:40 |
maaku | TMA: basically everything is preserved: the full cell structure, the proteome, relative positions, etc. | 15:42 |
maaku | In fact the reason why the blood if flushed with antifreeze is basically to make sure everything stays in the place it was at | 15:42 |
maaku | (vs. freezing with water, where crystal expansion moves everything around) | 15:42 |
maaku | Alcor's process actually makes the medical recovery significantly more difficult because they use highly toxic cryopreservation chemicals specifically because it reduces damage during vitrification and storage. | 15:44 |
maaku | (vs. Cryonics Institute which historically preferred simpler methods) | 15:44 |
TMA | maaku: I'll set my doubts on preservation viability aside for now. | 15:46 |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has joined ##hplusroadmap | 15:55 | |
kanzure | maaku: for long-term cryopreservation with specific dates in the future, it would be good to send oral historians sure but also both genders because yeah... | 15:56 |
jcorgan | i would imagine though that the process destroys any extracellular chemical gradients (neurotransmitters, hormones). but, perhaps these only matter on a very short-term time basis. | 15:56 |
kanzure | to answer the age-old "do the chemical gradients matter" question, i'm sure we could design some horrible experiments. | 15:58 |
maaku | kanzure: i assume tissue samples would get the information you care about | 15:58 |
kanzure | you can destroy chemical gradients inside synapses without using vitrification or cryopreservation of any kind | 15:59 |
maaku | but most cultures have different information transmitted down different gender lineages so you probably would want both | 15:59 |
kanzure | tissue samples plus reproduction tech/equipment, sure. yes. | 16:00 |
kanzure | (if it works) | 16:00 |
kanzure | why are there so few search results for "xangelix" | 16:01 |
kanzure | ah here we go, this is necessary | 16:01 |
kanzure | .title https://www.youtube.com/watch?v=WXgNo5Smino | 16:01 |
yoleaux | Live at the Necropolis: Lords of Synth | Adult Swim - YouTube | 16:01 |
maaku | wat did i just watch | 16:04 |
kanzure | i ask myself that all the time | 16:12 |
-!- Cory [~C@unaffiliated/cory] has quit [] | 16:17 | |
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-sclpbpmqgqumpbab] has quit [Quit: Connection closed for inactivity] | 16:22 | |
-!- justanotheruser [~Justan@unaffiliated/justanotheruser] has quit [Ping timeout: 276 seconds] | 16:29 | |
-!- Cory [~C@unaffiliated/cory] has joined ##hplusroadmap | 16:31 | |
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving] | 16:58 | |
-!- fleshtheworld [~fleshthew@2602:306:cf0f:4c20:105:839f:6fa9:fe35] has quit [Read error: Connection reset by peer] | 17:13 | |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has joined ##hplusroadmap | 17:20 | |
-!- augur [~augur@76-218-206-38.lightspeed.sntcca.sbcglobal.net] has quit [Ping timeout: 260 seconds] | 17:25 | |
-!- nildicit [~nildicit@unaffiliated/nildicit] has joined ##hplusroadmap | 17:38 | |
-!- nmz787_i [~ntmccork@134.134.139.78] has quit [Remote host closed the connection] | 18:06 | |
-!- nmz787_i [~ntmccork@134.134.139.78] has joined ##hplusroadmap | 18:06 | |
-!- Quashie [~boingredd@CPEbc4dfb5bf733-CMbc4dfb5bf730.cpe.net.cable.rogers.com] has joined ##hplusroadmap | 18:09 | |
-!- Gurkenglas [Gurkenglas@dslb-178-000-182-226.178.000.pools.vodafone-ip.de] has quit [Ping timeout: 246 seconds] | 18:13 | |
-!- dustinm [~dustinm@68.ip-149-56-14.net] has quit [Ping timeout: 260 seconds] | 18:14 | |
-!- dustinm [~dustinm@68.ip-149-56-14.net] has joined ##hplusroadmap | 18:14 | |
-!- nmz787_i [~ntmccork@134.134.139.78] has quit [Quit: Leaving.] | 18:54 | |
-!- augur [~augur@2602:304:cdac:e260:add3:743c:db20:6793] has joined ##hplusroadmap | 19:15 | |
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 260 seconds] | 19:17 | |
-!- CuriousCat [~CuriousCa@unaffiliated/wye-naught/x-8734122] has joined ##hplusroadmap | 19:26 | |
-!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap | 19:36 | |
-!- yashgaroth [~yashgarot@2602:306:35fa:d500:f5e0:f867:a11d:8d52] has joined ##hplusroadmap | 19:40 | |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has quit [Ping timeout: 258 seconds] | 19:42 | |
-!- sandeepkr [~sandeep@111.235.64.4] has joined ##hplusroadmap | 19:53 | |
-!- comma8 [~comma8@2601:1c2:d00:c980:2263:b019:c515:9556] has joined ##hplusroadmap | 20:11 | |
-!- CuriousCat [~CuriousCa@unaffiliated/wye-naught/x-8734122] has quit [Quit: * * * * *] | 20:35 | |
-!- CuriousCat [~CuriousCa@unaffiliated/wye-naught/x-8734122] has joined ##hplusroadmap | 20:37 | |
-!- CuriousCat [~CuriousCa@unaffiliated/wye-naught/x-8734122] has quit [Quit: * * * * *] | 20:51 | |
-!- justanotheruser [~Justan@unaffiliated/justanotheruser] has joined ##hplusroadmap | 21:03 | |
nmz787 | abetusk: well BRLCAD can definitely export a number of 3D formats... the challenge is knowing which to use to for slicing. the BRLCAD developer who I talked to in their IRC room said rastering over the slice with their raytracing interface would definitely be a way forward... I guess I could try to export SVG (not sure if that would work, or be better for g-code production) | 21:25 |
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-zloidozicjmdafvy] has joined ##hplusroadmap | 21:31 | |
-!- ArturSha1 [~ArturShai@31.29.0.57] has joined ##hplusroadmap | 21:48 | |
-!- ArturSha1 [~ArturShai@31.29.0.57] has quit [Ping timeout: 244 seconds] | 21:54 | |
-!- irseeyou [~irseeyou@c-67-168-101-20.hsd1.wa.comcast.net] has left ##hplusroadmap [] | 22:20 | |
-!- CuriousCat [~CuriousCa@unaffiliated/wye-naught/x-8734122] has joined ##hplusroadmap | 22:25 | |
-!- ArturSha1 [~ArturShai@212.97.30.60] has joined ##hplusroadmap | 22:26 | |
-!- drewbot_ [~cinch@ec2-54-242-253-140.compute-1.amazonaws.com] has quit [Remote host closed the connection] | 22:30 | |
-!- drewbot [~cinch@ec2-54-204-190-255.compute-1.amazonaws.com] has joined ##hplusroadmap | 22:31 | |
-!- ArturSha1 [~ArturShai@212.97.30.60] has quit [Ping timeout: 240 seconds] | 22:31 | |
nmz787 | kanzure: http://cap.ee.ic.ac.uk/~pdm97/powermems/2005/pdfs/069_Piechna.pdf | 22:34 |
nmz787 | .wik Wave_disk_engine | 22:35 |
yoleaux | "A wave disk engine or wave disk generator is a type of pistonless rotary engine being developed at Michigan State University and Warsaw Institute of Technology. The engine has a spinning disk with curved blades." — https://en.wikipedia.org/wiki/Wave_disk_engine | 22:35 |
-!- ArturSha1 [~ArturShai@31.29.13.152] has joined ##hplusroadmap | 22:45 | |
-!- justanotheruser [~Justan@unaffiliated/justanotheruser] has quit [Ping timeout: 240 seconds] | 22:52 | |
-!- justanotheruser [~Justan@unaffiliated/justanotheruser] has joined ##hplusroadmap | 22:53 | |
-!- ArturSha1 [~ArturShai@31.29.13.152] has quit [Ping timeout: 240 seconds] | 22:56 | |
-!- jaboja [~jaboja@elh211.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap | 22:56 | |
-!- jaboja [~jaboja@elh211.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 250 seconds] | 23:04 | |
-!- ArturSha1 [~ArturShai@212.241.9.248] has joined ##hplusroadmap | 23:09 | |
-!- strangewarp [~strangewa@c-76-25-206-3.hsd1.co.comcast.net] has joined ##hplusroadmap | 23:13 | |
-!- ArturSha1 [~ArturShai@212.241.9.248] has quit [Ping timeout: 272 seconds] | 23:13 | |
-!- jaboja [~jaboja@vps.jaboja.pl] has joined ##hplusroadmap | 23:17 | |
-!- yashgaroth [~yashgarot@2602:306:35fa:d500:f5e0:f867:a11d:8d52] has quit [Quit: Leaving] | 23:17 | |
-!- augur [~augur@2602:304:cdac:e260:add3:743c:db20:6793] has quit [Remote host closed the connection] | 23:22 | |
-!- CuriousCat [~CuriousCa@unaffiliated/wye-naught/x-8734122] has quit [Quit: * * * * *] | 23:24 | |
-!- ArturSha1 [~ArturShai@31.29.14.198] has joined ##hplusroadmap | 23:25 | |
-!- jaboja [~jaboja@vps.jaboja.pl] has quit [Remote host closed the connection] | 23:27 | |
-!- nildicit [~nildicit@unaffiliated/nildicit] has quit [Ping timeout: 252 seconds] | 23:32 | |
-!- ArturSha1 [~ArturShai@31.29.14.198] has quit [Ping timeout: 246 seconds] | 23:38 | |
-!- ArturSha1 [~ArturShai@212.97.0.145] has joined ##hplusroadmap | 23:52 | |
--- Log closed Sat May 28 00:00:13 2016 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!