2014-12-05.log

--- Day changed Fri Dec 05 2014
-!- Boscop [me@unaffiliated/boscop] has quit [Quit: Boscop]00:13
-!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap00:20
-!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has joined ##hplusroadmap00:23
-!- Burn_ [~Burn@pool-71-191-174-26.washdc.fios.verizon.net] has quit [Read error: No route to host]00:27
-!- Burninate [~Burn@pool-71-191-174-26.washdc.fios.verizon.net] has joined ##hplusroadmap00:47
-!- lichen [~lichen@c-50-139-11-6.hsd1.or.comcast.net] has joined ##hplusroadmap01:13
-!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has quit [Remote host closed the connection]01:22
-!- Vutral [ss@mirbsd/special/Vutral] has quit [Ping timeout: 245 seconds]01:29
-!- Vutral [uNBH5PZDFs@mirbsd/special/Vutral] has joined ##hplusroadmap01:37
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Ping timeout: 245 seconds]01:55
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap02:03
-!- NilsHitze [~hitze@80.190.141.50] has joined ##hplusroadmap02:11
NilsHitzemornin02:11
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-fwziiiwkfoivycvd] has joined ##hplusroadmap02:30
fennhopefully i'm not the only one who sees the irony in this: http://blog.codinghorror.com/the-incredible-linktron-5000tm/02:53
-!- saurik_ is now known as saurik02:59
-!- gene_hacker [~chatzilla@c-50-137-46-240.hsd1.or.comcast.net] has quit [Ping timeout: 260 seconds]03:05
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap03:15
ebowdenpaperbot: http://www.nature.com/neuro/journal/vaop/ncurrent/full/nn.3887.html03:19
paperbothttp://libgen.org/scimag/get.php?doi=10.1038%2Fnn.388703:19
ebowden\:D/03:20
ebowdenpaperbot: http://www.nature.com/nature/journal/vaop/ncurrent/full/nature13982.html03:24
paperbothttp://libgen.org/scimag/get.php?doi=10.1038%2Fnature1398203:24
ebowden\:D/03:25
ebowdenpaperbot: http://www.cell.com/cell-metabolism/abstract/S1550-4131%2814%2900500-203:36
ebowden+tel duces http://www.liveleak.com/view?i=2cf_134176574203:40
-!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 272 seconds]03:42
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap03:50
-!- cpopell2 [~cpopell@c-76-26-144-132.hsd1.dc.comcast.net] has joined ##hplusroadmap03:54
-!- cpopell [~cpopell@c-76-26-144-132.hsd1.dc.comcast.net] has quit [Ping timeout: 240 seconds]03:55
ebowdenpaperbot: http://www.jneurosci.org/content/34/35/1184404:07
paperbothttp://libgen.org/scimag/get.php?doi=10.1523%2FJNEUROSCI.4642-12.201404:07
ebowden\:D/04:15
-!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap04:17
ebowdenpaperbot: http://onlinelibrary.wiley.com/doi/10.1002/jnr.490330407/abstract04:21
ebowdenpaperbot: http://www.ncbi.nlm.nih.gov/pubmed/2212755604:44
paperbothttp://libgen.org/scimag/get.php?doi=10.1007%2Fs00213-011-2594-804:44
-!- Vutral [uNBH5PZDFs@mirbsd/special/Vutral] has quit [Ping timeout: 245 seconds]04:48
-!- weles [~mariusz@wsip-174-78-132-9.ri.ri.cox.net] has joined ##hplusroadmap05:04
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Read error: Connection reset by peer]05:06
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap05:07
-!- Boscop [~me@46.246.12.60] has joined ##hplusroadmap05:08
-!- Boscop [~me@46.246.12.60] has quit [Changing host]05:08
-!- Boscop [~me@unaffiliated/boscop] has joined ##hplusroadmap05:08
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has quit [Quit: Lost terminal]05:11
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap05:11
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-fwziiiwkfoivycvd] has quit [Quit: Connection closed for inactivity]05:17
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has quit [Quit: Lost terminal]05:20
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap05:23
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has quit [Client Quit]05:26
-!- ebowden_ [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap05:41
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Read error: Connection reset by peer]05:41
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap05:47
-!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 244 seconds]05:53
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 255 seconds]06:10
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap06:12
-!- ebowden_ [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Remote host closed the connection]06:18
-!- chido [chidori@pasky.or.cz] has joined ##hplusroadmap06:25
-!- weles [~mariusz@wsip-174-78-132-9.ri.ri.cox.net] has quit [Ping timeout: 244 seconds]06:26
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Ex-Chat]06:29
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has quit [Quit: Lost terminal]07:09
kanzureblah07:17
RedMEdicblah blah blah blah07:18
kanzurebasically...07:23
RedMEdichttps://www.youtube.com/watch?v=SgBj7Mc_4sc07:24
archelshttps://www.kickstarter.com/projects/chaibio/open-qpcr-dna-diagnostics-for-everyone07:24
RedMEdicHow much will this cost?07:25
archelssome $1299, according to that page07:26
RedMEdicDamn07:26
kanzuretitle https://www.youtube.com/watch?v=SgBj7Mc_4sc07:28
kanzure.title https://www.youtube.com/watch?v=SgBj7Mc_4sc07:28
yoleauxThe Poo in You - Constipation and Encopresis Educational Video - YouTube07:28
-!- Vutral [~ss@mirbsd/special/Vutral] has joined ##hplusroadmap07:37
-!- NilsHitze [~hitze@80.190.141.50] has quit [Quit: NilsHitze]07:39
-!- JayDugger [~jwdugger@pool-173-57-55-138.dllstx.fios.verizon.net] has quit [Quit: Leaving.]07:45
-!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has quit [Remote host closed the connection]07:52
-!- RedMEdic [~RedMedic@CPE-69-23-98-42.new.res.rr.com] has quit [Ping timeout: 240 seconds]08:01
-!- RedMEdic [~RedMedic@CPE-69-23-98-42.new.res.rr.com] has joined ##hplusroadmap08:02
-!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has joined ##hplusroadmap08:05
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap08:17
-!- weles [~mariusz@wsip-174-78-132-9.ri.ri.cox.net] has joined ##hplusroadmap08:21
kanzuresteve coles http://www.latimes.com/local/obituaries/la-me-stephen-coles-20141205-story.html08:25
kanzurehehe "a 103-year-old woman to practice as a pediatrician and a 106-year-old woman to drive."08:26
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap08:36
kanzuremore openworm stuff in the news http://www.newscientist.com/article/mg22429972.300-first-digital-animal-will-be-perfect-copy-of-real-worm.html08:40
-!- _Sol_ [~SolGr@c-69-141-24-242.hsd1.nj.comcast.net] has quit [Ping timeout: 245 seconds]08:41
-!- _Sol_ [~SolGr@c-69-141-24-242.hsd1.nj.comcast.net] has joined ##hplusroadmap08:42
-!- eudoxia [~eudoxia@r179-25-166-141.dialup.adsl.anteldata.net.uy] has quit [Quit: Lost terminal]08:50
-!- nmz787_i [nmccorkx@nat/intel/x-iauzjrapqyprnpmz] has joined ##hplusroadmap08:59
-!- chido [chidori@pasky.or.cz] has quit [Quit: leaving]09:00
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Ping timeout: 252 seconds]09:20
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap09:25
kanzure"The uneasy relationship between mathematics and cryptography" http://www.ams.org/notices/200708/tx070800972p.pdf09:46
kragenis that Koblitz of the Koblitz elliptic curves?09:54
kanzureheh #bitcoin-wizards just asked that same question09:56
kragenyes09:57
kanzure.title https://plus.google.com/+KajSotala/posts/ZGtGMRbHumM10:13
yoleauxA sophisticated hacker group is targeting highly placed corporate personnel…10:13
kanzurehttps://securelist.com/blog/research/66779/the-darkhotel-apt/10:14
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-gndtyetlulumzjjm] has joined ##hplusroadmap10:17
kanzurehttp://einsteinpapers.press.princeton.edu/10:34
kanzurehttp://www.einstein.caltech.edu/ with links to various highlights10:34
kanzure"To Max Planck, on receiving credible death threats — Einstein writes that he cannot attend the Scientist’s Convention in Berlin because he is “supposedly among the group of persons being targeted by nationalist assassins.”"10:35
kanzurehah10:35
-!- Zinglon [~Zinglon@ip565f6f48.direct-adsl.nl] has joined ##hplusroadmap10:38
kanzurewhat? his second year instructor just-so-happened to be minkowski? that seems a little unlikely... http://einsteinpapers.press.princeton.edu/vol1-trans/4910:38
kanzureer i meant http://einsteinpapers.press.princeton.edu/vol1-trans/48 but same thing10:38
-!- TMA [tma@twin.jikos.cz] has quit [Ping timeout: 256 seconds]11:01
-!- TMA [tma@twin.jikos.cz] has joined ##hplusroadmap11:02
-!- ybit [~ybit@unaffiliated/ybit] has quit [Quit: leaving]11:48
-!- heath [~heath@unaffiliated/ybit] has joined ##hplusroadmap11:48
nmz787_ihttp://www.antha-lang.org/12:00
nmz787_i.title12:01
yoleauxWelcome to Antha - Antha12:01
nmz787_i'coding biology'12:01
nmz787_i"Antha is a high-level language for biology, making it easy to rapidly compose reproducible work flows using individually testable and reusable Antha Elements."12:01
nmz787_i"© 2014 Antha Authors. Code licensed under the GPL 2.0 License. Documentation licensed under CC BY 3.0. Proudly sponsored by Synthace"12:01
nmz787_ihttps://github.com/antha-lang12:01
nmz787_ikanzure: plz give your summary when ready12:01
RedMEdicAre we discussing Antha again12:02
RedMEdicAntha is bullshit12:02
nmz787_ihttp://www.antha-lang.org/docs/intro.html12:03
-!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Read error: Connection reset by peer]12:04
nmz787_iwhen was it previously discussed?12:04
nmz787_i(I can't grep the logs right now)12:04
nmz787_ithis doesn't look terrible, though I'd prefer something non-java (or javascrpit I guess) https://github.com/antha-lang/antha/blob/master/examples/bradford/bradford.an12:05
nmz787_ior is this go?12:06
-!- superkuh [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap12:06
nmz787_ihmm, I cannot find an API function list12:08
-!- drewbot [~cinch@ec2-54-166-133-142.compute-1.amazonaws.com] has quit [Read error: Connection reset by peer]12:10
kanzurehttp://in-theory.blogspot.com/2007/08/swift-boating-of-modern-cryptography.html12:10
-!- drewbot [~cinch@ec2-54-145-76-87.compute-1.amazonaws.com] has joined ##hplusroadmap12:10
kanzurenmz787_i: use http://transcriptic.com/platform instead of antha12:10
kanzure"Greetings, At least one competing scientist has e-mailed me today to dispute the assertion, placed in the media, that Dr. Coles had "over 100" scientific journal articles published."12:13
kanzurewtf12:13
kragen"Without this technology, buying and selling things online would be extremely inconvenient and companies like amazon and ebay would probably not exist.12:14
kragenthis seems unlikely to me12:14
kragenyou'd just have to use something like kerberos and get a TGT to talk to Amazon instead of verifying that their cert isn't on the CRL12:15
kanzureworst case scenario you would have lots and lots of escrow12:15
kragenescrow doesn't make things less convenient12:16
kragenjust riskier12:16
kanzureoh right, the claim was inconvenience and not impossibility12:16
kragenaye12:17
kanzurewhy would anyone dispute some dead guy's obituary saying x number of publications12:18
kragenlies anger some people12:19
nmz787_ijrayhawk:  ah hahahahah I just saw a spam email in my spam folder with the title "Overeating Sugar... CURES Diabetes?"12:21
nmz787_ikanzure: I know of transcriptic... does your indication mean you've previously reviewed antha?12:22
heathnmz787_i: last night12:31
heath..is when antha was mentioned12:31
heathpostgresql 9.4.0 to be released next week for anyone who cares12:33
heathhttp://www.postgresql.org/message-id/26295.1417708564@sss.pgh.pa.us?utm_source=dbweekly&utm_medium=email12:33
-!- _0bitcount [~big-byte@81.61.34.185.dyn.user.ono.com] has joined ##hplusroadmap12:33
_0bitcount"Wanderers" http://vimeo.com/108650530  . How I miss Carl Sagan.12:41
heath_0bitcount: linked a few days ago :)12:48
heathpretty neat vid though12:48
_0bitcountSorry for reposting. :-)12:48
_0bitcountI wonder if we'll ever make it to those places. It will probably easier to send AI-powered spaceships with ultra-HD sensors to report back to Earth.12:51
archelsthis one is worth reposting12:51
-!- weles [~mariusz@wsip-174-78-132-9.ri.ri.cox.net] has quit [Ping timeout: 258 seconds]12:55
-!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap13:03
heathi guess that came off a bit annoying, apologies13:04
kragen_0bitcount: it's easier today to weave cloth with automatic electric looms, but that doesn't stop enthusiasts from hand-weaving their own cloth13:09
kragenit's mostly a question of will, which is to say, budget13:10
_0bitcountkragen, true, but I think we humans will be very different from our present fragile being by then. Our bodies will have plenty of improvements to withstand the harsh environment of space.13:12
nmz787_i_0bitcount: very cool vid13:13
_0bitcountIt really is.13:17
archels_0bitcount: the space faring thing is just around in the corner--in fact we're already doing it now. In comparison to that, biological modifcations are much much more inchoate.13:17
-!- RedMEdic [~RedMedic@CPE-69-23-98-42.new.res.rr.com] has quit [Ping timeout: 260 seconds]13:18
_0bitcountarchels, I truly hope that is about to happen. However, I read about how many in governments and big, old industries don't want the space frontier to be open...13:19
_0bitcount... except for their own, limited and restricted purporses.13:19
archelsokay, I don't want to get into a whole free market rhetoric here13:21
archelsshall we say... whatever works? :)13:21
_0bitcountMaybe I am being a bit negative. Maybe, for today's technology, establishing permanent bases in the Moon and Mars is not that different from crossing the Bering Strait or the Atlantic Ocean centuries ago.13:22
_0bitcountarchels, I am confident that the pressures of the market will do their part on opening space for a lot more.  ;-)13:23
kragen_0bitcount: yeah, it's totally plausible that we'll modify our bodies for spacefaring13:24
kragenbut we're not actually that fragile13:24
_0bitcountkragen, but won't be easier/cheaper in, say, 30 years to send a highly autonomous fleet of ships to Jupiter, Europa, Saturn, etc than to make a habitable pod?13:27
kanzureit'll also be cheaper to send emulated brains than fleshmatter, but what's your point?13:29
_0bitcountMy point is that we might miss visiting those worlds if it ends up being cheaper and easier to send our artificial servants.13:31
_0bitcountAnd maybe experience those environments through virtual reality.13:31
kanzurethere is nothing glamorous about being stuck in a metal tin can for 40 years13:32
archelskragen: eventually yeah, but these are wildly different timescales13:32
kanzurecalling an emulated brain a servant is a little strange but if you want to force non-cooperative people into doing stuff hey whatever13:33
_0bitcountkanzure, unless it's like the Enterprise, holodeck and all.  :-)13:33
_0bitcountkanzure, it was just a poetic way of thinking about it. They will probably feel more like a part of ourselves.13:35
archelshttps://asciinema.org/13:36
archels_0bitcount: Mars gets dull pretty quickly if you're just looking around13:38
archelsand the finite speed of light will make telepresence useless, anyway13:38
archelsso then we're back to sending WBEs, but I don't think that's what you meant13:39
kanzureprobably wont need to send *whole* brains either13:39
archelswhat manner of evil is this now13:39
_0bitcountarchels, whether WBE or pure AI, that was my idea, that sending "someone" not made of flesh would be easier and more manageable.13:40
archelsoh okay, yeah I'm 100% with you on that one13:41
_0bitcountSo space tourism will end up being something we do around Earth's orbit, maybe on the Moon.13:41
archelsthat's very different from sending "artificial servants" though and experiencing things vicariously through VR13:42
_0bitcountSo are we not considering AIs like servants anymore?  :-)13:43
archelswell, either *you* go, or you stay at home and look at a video feed with six and half minute delay13:44
_0bitcountNot that I have anything against conceding them full citizenry. I am sure I will love my daily AI helpers more than many among the people I have to deal with.13:45
_0bitcountarchels, "video" sounds very XX century. How about ultra-HD 3D systems? I we can buffer enough of the surroundings, an immersive experience is going to be realistic enough.13:49
archelslike Google street view? sure, but like you said, then we're just looking at a cached, Earthbound data stream13:51
kanzurei can't follow the conversation any more13:53
kanzuresend many cheap things this is good13:53
archelssending humans, too13:54
archelsbe they in digital or analog form13:54
nmz787_iwow, the comments here are vulgar and violent (I don't think I13:57
nmz787_i'd heard of this before today) https://www.kickstarter.com/projects/secretlabs/agent-the-worlds-smartest-watch/comments13:57
nmz787_iwow this is really cool https://www.youtube.com/watch?v=J9b0J29OzAU13:59
nmz787_i.title13:59
yoleauxWorld's Simplest Electric Train 【世界一簡単な構造の電車】 - YouTube13:59
nmz787_ichris_99: there's a UK BOM for that here http://hackaday.com/2014/12/04/amazing-sciences-simple-electric-train/#comment-221084214:01
chris_99heh cool, cheers14:02
chris_99oh i didn't think the bettery was doing anything, but apparently it's creating an electromagnet using the copper coil?14:05
nmz787_iyea14:09
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 265 seconds]14:15
-!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has joined ##hplusroadmap14:15
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap14:23
kragen_0bitcount: it might be easier and cheaper, but people might still do both14:30
-!- delinquentme [~dingo@74.61.157.78] has quit [Quit: Leaving]14:31
kragenI don't think anybody has a very clear idea of the timescales we're talking about here14:31
_0bitcountkragen, at what point in time are you expecting those capabilities to be common?14:34
kragensometime this century or next14:35
kragenmaybe next month14:35
_0bitcount?14:36
_0bitcountkragen, that's not very reassuring. Some could say we should just forget and focus on more urgent matters down here, on Earth.14:47
kanzurewhy do you care about what they say?14:47
_0bitcountkanzure, what do you mean?14:54
kanzurewhether or not it takes a month for capabilities to be common should be an independent analysis from whether or not there exist people who say "forget and focus on other problems"14:55
_0bitcountkanzure, since large scale problems require a lot of money, public relations is a big imperative in selling politicians and the public the need for something.14:57
_0bitcountHence having to be more or less accurate with the dates.14:57
_0bitcountImagine Elon Musk telling investors and potential customers that "a competitive electric vehicle would be ready by next month, maybe 2050."14:59
kanzureelon musk will still do whatever the fuck he wants independent of the existence of people who want to "forget and focus on other problems"14:59
kanzureso your argument is still garbage14:59
kragencalm down, kanzure15:01
_0bitcountkanzure, it might be garbage, but I don't like people like Tesla dying in destitution for lack of support.15:01
kanzurebad arguments are bad and should be rejected15:01
kanzurei odn't understand what destitution has to do with this15:01
kragenit'll be okay15:01
kanzureno it wont if you let bad arguments stick around your SNR drops15:01
kragen_0bitcount: designing a competitive electric vehicle doesn't require any basic science; even when Tesla started, it was just a matter of engineering and cost control.15:02
_0bitcountI was talking about Nikola Tesla in that last example.15:02
_0bitcountSorry for mixing things up.15:03
kanzureno it is elon who should apologize for violating namespace integrity15:03
kanzureactually didn't elon get into tesla after tesla was named tesla?15:03
kanzureso anyway, at most you can blame him for continuing to allow it to be named tesla15:03
kragenWhole-brain emulation, by contrast, requires new basic science, as does adapting our bodies to outer space.  Sending humans to Jupiter might be possible without new basic science, but we could only do it without new basic science on such a long timeframe that new relevant basic science will probably emerge.15:04
kanzureyou could send all the cadavers you like, does that help?15:04
kragenSo all of these things are deeply unpredictable.  That alone wouldn't justify "maybe next month".15:04
kanzuretesla didn't die in destitution because people who argued "forget and focus on other problems"15:05
kanzure*because of15:05
kragenBut maybe next month an AGI developed at Google's Switzerland office will FOOM and perhaps decide that sending AGIs to Jupiter is an eminently reasonable idea and should happen as soon as possible.  Say, within a week.15:05
kragenI mean these are just very difficult things to predict.15:06
_0bitcountkanzure, I was under the impression that he was kind of a humanist, so bankers and financiers ended up not wanting anything to do with him.15:08
kragenHe was insane, which made it hard to work with him.15:08
kanzureyou are both off on some weirdo political tangent now that i refuse to follow15:09
kanzurei still hold that whether or not it takes a month for capabilities to be common should be an independent analysis from whether or not there exist people who say "forget and focus on other problems"15:09
_0bitcountI'm not claiming any accurate, historical knowledge, only what I have read.15:09
kanzureand that you should not be telling me that the existence of people who are not interested in your proposals are the primary reason for your inability to execute15:09
kragenHopefully you're talking to _0bitcount here.  I'm well aware that my own inability to execute is entirely self-generated!15:10
kragenAlso, I'm not trying to execute any of those things.15:10
kanzureright, i mean _0bitcount15:10
kanzureand i'm glad to hear that you're "well aware that my own inability to execute is entirely self-generated" hehe15:11
kanzurearguably you should be trying to execute at least some of those things, though15:11
kanzureuh, in this context i'm not sure which subset15:12
kragenI mean maybe a different environment would be more congenial to it and compensate better for my weaknesses.  But fundamentally they're my weaknesses, not the world's.15:12
_0bitcountkanzure, I like that attitude. :-)  I frequently remind myself that I should be doing what I want, not what is, apparently, expected of me.15:12
kragen_0bitcount: what you want is not necessarily what you should be doing15:12
kragenI mean that's sort of a pure hedonist viewpoint, exacerbated by a failure to recognize your own irrationality15:13
kanzurewanting the impossible is a recipe for a bad dinner15:13
_0bitcountI wasn't thinking about space, just more mundane things.15:13
kanzurewell, i mean, wanting to cheat around certain amounts of effort investments is a recipe for bad tastes in your mouth15:13
kragenyeah15:14
kanzurealthough for certain things it is attainable to cheat15:14
kanzuremostly these are already existing things15:14
kanzureif i had more working memory then i would elaborate on this concept but i can't picture it correctly15:15
kanzurei don't think it's precisely irrationality, just something like a mismatch of effort expectation versus outcome estimation15:15
kragenyou need the right amount of "cheating" to be successful.  Too much "cheating" and you defeat yourself, like Coyote nailing his eyelids shut15:15
kragenToo little "cheating" and most of your hard work is wasted.15:15
kanzure"cheating" is stuff like "naaah i'm sure the government will do this for me"15:16
kragenyeah, that's a good strategy in many cases15:16
kanzureso far the government has not put people on mars, so arguably no?15:17
-!- Zinglon [~Zinglon@ip565f6f48.direct-adsl.nl] has quit [Ping timeout: 245 seconds]15:17
-!- eudoxia [~eudoxia@179.26.181.44] has joined ##hplusroadmap15:17
kragenno, but it did pay for the development of most of the technology needed to do so.15:17
kragenif that's your goal, you should focus your efforts on things where they will make more difference, for example, things the government is less likely to do for you.15:18
kragenAnd maybe figuring out how to avoid coming into direct conflict with the government.15:19
kragene.g. do your stem-cell research in South Korea, not the US.15:19
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Ex-Chat]15:41
kanzurehmm one of the openworm people is doing bitcoin things https://github.com/bitcoin/bitcoin/pull/523115:42
kanzurestalkermatrix wins again15:42
paskyluke-jr is working on openworm too?15:50
paskyhe runs eligius, my favorite mining pool (when i was still mining bitcoins)15:50
kanzurenope not luke-jr15:51
kanzure"ABISprotocol"15:51
kanzurepasky: hi15:53
kanzurei have been meaning to bug you about ai things15:54
paskyhi :)15:55
paskybug on15:55
kanzurepage 4 figure 2 http://diyhpl.us/~bryan/papers2/neuro/randall-oreilly/Biologically%20based%20computational%20models%20of%20high-level%20cognition.pdf15:56
kanzurepage 3 figure 3 http://diyhpl.us/~bryan/papers2/neuro/randall-oreilly/Towards%20an%20executive%20without%20a%20homunculus:%20computational%20models%20of%20the%20prefrontal%20cortex%20basal%20ganglia%20system.pdf15:56
kanzurepage 6 figure 4 of http://arxiv.org/vc/arxiv/papers/1008/1008.5161v1.pdf15:56
kanzurewhat do you think?15:56
kanzuremy primary concern is that this is already highly conserved between animals and therefore seems to be insufficient to make general intelligence 15:57
pasky*scratches head*16:01
paskyi'm not sure i can think anything about it before reading these papers in detail16:01
paskyobviously the arxiv paper seems most comprehensible to me16:02
kanzurei'm really on the fence as to whether or not to like that paper16:02
paskybut the title of that figure is rather amusing16:02
kanzureyes :)16:02
kanzureif you want some really long reading then this paper (miscategorized, so bite me) is good http://diyhpl.us/~bryan/papers2/neuro/randall-oreilly/Towards%20a%20universal%20cortical%20algorithm:%20Examining%20hierarchical%20temporal%20memory%20in%20light%20of%20frontal%20cortical%20function.pdf16:03
kanzure(this was also found on arxiv, which is a nice trend)16:03
paskyah, HTMs16:04
kanzuresorta16:05
paskyi think i need to pick something to focus on here - so basically what you are pondering is how the brain decides which things to remember?16:06
kanzuremore like, given what we have previously discussed and your insistence on symbolic computation, do you think this is an interesting direction even if it is explicitly non-symbolic16:07
kanzureuh, and questions like, methods of estimation of the plausibility of an implementation option (such as those diagrams) doing anything interesting without buying 100 billion supercomputers for testing16:08
-!- yashgaroth [~ffffff@2606:6000:cb85:6a00:a4c1:2b23:d54c:e64a] has joined ##hplusroadmap16:08
-!- nmz787_i [nmccorkx@nat/intel/x-iauzjrapqyprnpmz] has quit [Quit: Leaving.]16:08
kanzureoh this is curious, "The frontal lobes have long been among the least well understood regions of the brain. Early experiments using lesion and electrical stimulation failed to identify any clear functions for the majority of frontal cortex, resulting in their being considered the "silent lobes" for much of the twentieth century. It is only in the last several decades that new techniques have allowed researchers to begin to build an ...16:10
kanzure... understanding of frontal lobe function."16:10
paskyhehe, i would certainly not say about myself that i'm in favor of "symbolic computation" at all; but maybe definitions of symbolic computation vary16:10
kanzureregarding that last link, perhaps starting at page 31 would be better16:10
kanzureoh it is possible that i am misremembering your design proposal thingy16:11
paskybut so far I don't even really understand what are the diagrams *about* - i thought it's about memory updates, but now it seems more like about action selection; i'll just read the papers i guess16:11
paskymaybe it's both16:11
kanzurethey are mixing memory and action selection sort of16:11
kanzureexecutable memory of sorts16:11
-!- _0bitcount [~big-byte@81.61.34.185.dyn.user.ono.com] has quit [Quit: Leaving]16:15
pasky(in general, i'm not in favor of insisting to emulate the brain down to details to reproduce general intelligence, biologically plausible neurons etc.; but doing *architecturally* similar things makes a lot of sense, and i'm e.g. huge fan of distributed representations)16:32
kanzure"distributed representations"?16:33
paskyi'll need some time to digest this; i've read pages 31-47 of that last link and it's really interesting16:33
paskykanzure: ever heard of word2vec?16:33
kanzure"This tool provides an efficient implementation of the continuous bag-of-words and skip-gram architectures for computing vector representations of words. These representations can be subsequently used in many natural language processing applications and for further research."16:33
kanzure"The word2vec tool takes a text corpus as input and produces the word vectors as output. It first constructs a vocabulary from the training text data and then learns vector representation of words. The resulting word vector file can be used as features in many natural language processing and machine learning applications."16:33
paskybasically, the point is that each word is represented by a long vector (typically 100D), which originally was a hidden layer of an ANN16:34
paskyand the vector space emerging from this is very regular16:35
eudoxiai think i heard about this today at lunch16:35
kanzureso this is just giant multi-dimensional matrix math?16:35
eudoxiavector operations preserve semantics16:35
paskythe popular example is that doing aritmetics is meaningful; if you do king + woman - man, the word with smallest cosine distance to that vector is "queen"16:35
paskyit doesn't work 100% but it works surprisingly often16:36
paskyand such properties are completely implicit in the system; it really does come from an internal representation in a neuron network trained for some context-based word prediction task16:36
paskyso these are distributed representations in natural language processing16:36
paskynow in computer vision you also have them16:37
paskyyou probably heard about convolutional neural networks that can detect objects on images with extremely high precisions, do pretty impressive segmentation tasks etc.16:37
paskyand detect cats in videos :)16:37
paskyagain it's an ANN that has as output layer that indicates something like the type of object on the image16:38
paskybut if you tear that output layer away and look at the last hidden layer representation, you again get a 100D vector that is a distributed representation with quite rich description of the semantics of the image16:38
paskynow the funny thing is that you can learn mapping between image and word vectors to build text descriptions of images16:39
paskyyou start with some training set, but then on the testing set you are able to describe images using *unseen* words - so based on training set about cars and cats, you can label a picture of a plane as "airplane" even if you never saw one in the training set16:41
paskyit's where the "deep learning" fashion in AI is going now16:41
paskyand of course all of these tasks are about recognition stuff, there is more niche research on how to take actions (the famous deepmind paper on atari games), but noone seems to have *any* *fricking* *clue* how to get planning involved in this all16:43
pasky(of course everyone has many ideas, but nothing that works on any interesting scale; I don't think we even have good testcases for this)16:44
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap16:44
pasky(to clarify - word2vec actually is a trivial case of ANN, so it's often not even described as an ANN; but it's in the same abstraction space)16:46
kanzureso you are claiming that a large enough vector of characteristics/properties/attributes is equivalent to a hidden layer in an ANN?16:47
kanzureor what is the transformation function exactly between the two16:48
paskyit's some tensor, just matrix math16:48
paskyor maybe even just matrix multiplication, i'm not sure now16:48
kanzureso is the claim that these vectors are losslessly convertable into trained neural network topologies with weights n' stuff16:49
kanzureconvertible16:49
paskyoh, nope16:49
paskyit *is* weight vector of one hidden layer16:49
paskyinformation about other layers (if there are more) is not stored in it16:50
paskyin the word2vec context, you typically train the network then just keep that weight vector and throw away the network16:50
paskybecause the network just predicts the next word when you saw the last 10016:50
paskywe call it distributed representations because we imagine it like brain neuron activity; instead of having individual vector elements associated with fixed meaning, the semantics is stored in a complex way distributed over many elements16:54
-!- eudoxia [~eudoxia@179.26.181.44] has quit [Quit: leaving]17:15
kanzure"As a comparison, Uber raised money at a $18B valuation in June, and then closed a round at $40B a few days ago. While a smaller percentage increase, Instacart's $1.6B increase in value looks pretty trivial compared to Uber's $22B increase over the same time period."17:25
nmz787.title http://whatisgon.wordpress.com/2011/11/30/export-gimp-xcf-layers-as-jpegs/17:26
yoleauxExport Gimp XCF Layers as Jpegs | What is Going On?17:26
nmz787(using Python)17:26
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-gndtyetlulumzjjm] has quit [Quit: Connection closed for inactivity]17:27
nmz787and the API that roughly corresponds (the Python docs for GIMP seem pretty horrible to me, no API list) http://oldhome.schmorp.de/marc/pdb/file_jpeg_save.html17:27
-!- Qfwfq [~WashIrvin@unaffiliated/washirving] has quit [Ping timeout: 245 seconds]17:44
jrayhawknmz787: actually there were some very very very high-carb low-fat diet trials in, like, the fifties that reversed diabetes17:48
jrayhawkso low fat that compliance outside of a metabolic ward is nigh-impossible17:48
-!- Qfwfq [~WashIrvin@unaffiliated/washirving] has joined ##hplusroadmap17:50
kanzureer, the real diabetes?17:51
jrayhawkwhich one is real17:51
jrayhawkhttp://scholar.google.com/scholar?q=kempner+rice+diet17:53
kanzurepancreatic underproduction of insulin is the real one, right?17:53
nmz787paperbot: science17:55
nmz787paperbot: http://www.sciencedirect.com/science/article/pii/000293435090200217:55
jrayhawktype 1 diabetes is (at least) two diseases; one is autoimmunity targeting beta cells, two is various downstream effects of underproduction of insulin as a result of all the beta cells being dead.17:55
paperbotnmz787: get it yourself17:55
jrayhawkhahaha17:56
jrayhawktype 2 is cellular insulin resistance, generally caused by either oxidative stress or inflammation17:57
jrayhawkthere's also type 1.5 which is where the beta cells become so overburdened by compensatory production in type 2 that they all die off and you effectively get the second half of type 1 diabetes.17:57
jrayhawksome have argued for dementia being type 3, but that's unsettled17:58
kanzurei would definitely count 1.5 as real17:59
nmz787paperbot got rude17:59
jrayhawk1.5 is also pretty hard to get18:09
jrayhawkgenerally you need decades of pancreatic punishment to achieve it18:10
jrayhawki think pre-diabetes is bullshit, if you want something to be skeptical of18:11
kanzureis pre-diabetes supposed to be about type 1, type 2, type 1.5, or something else, or any of them?18:12
jrayhawkit's pre-type 218:12
nmz787https://www.youtube.com/watch?v=MKiP-4o3cFI18:12
nmz787.title18:12
yoleauxDirect Digital Synthesis (DDS) with Bil Herd - YouTube18:12
jrayhawkor, rather, in the spectrum of severity of type 2, it is low enough that doctors don't feel comfortable diagnosing an intervention or something18:13
jrayhawkthe entire concept of "standard lab ranges" in standard of care are also insane, so there's probably some reason for its existence derived therefrom18:15
nmz787fenn: https://www.youtube.com/watch?v=cVD5qKIY6ZM18:45
nmz787.title18:45
yoleaux3 Minutes On... The Intel 4004 Microprocessor - YouTube18:45
nmz787fenn: apparently they released tons of data on it a few years ago18:45
kanzurehttp://blog.uber.com/ride-ahead18:59
kanzure"This kind of continued growth requires investment. To that end, we have just raised a financing round of $1.2 billion, with additional capacity remaining for strategic investments. This financing will allow Uber to make substantial investments, particularly in the Asia Pacific region."18:59
nmz787http://firstmicroprocessor.com/documents/2013powerpoint.ppt19:12
nmz787.wik horners rule19:13
yoleaux"In mathematics, Horner's method (also known as Horner scheme in the UK or Horner's rule in the U.S.) is either of two things: (i) an algorithm for calculating polynomials, which consists of transforming the monomial form into a computationally efficient form; or (ii) a method for approximating the roots of a polynomial." — http://en.wikipedia.org/wiki/Horner's_method19:13
nmz787.wik booths multiply algorithm19:17
yoleaux"Booth's multiplication algorithm is a multiplication algorithm that multiplies two signed binary numbers in two's complement notation. The algorithm was invented by Andrew Donald Booth in 1950 while doing research on crystallography at Birkbeck College in Bloomsbury, London." — http://en.wikipedia.org/wiki/Booth's_multiplication_algorithm19:17
-!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 260 seconds]19:36
-!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap19:38
-!- Qfwfq [~WashIrvin@unaffiliated/washirving] has quit [Ping timeout: 252 seconds]19:44
-!- lichen [~lichen@c-50-139-11-6.hsd1.or.comcast.net] has quit [Quit: Lost terminal]19:48
-!- Qfwfq [~WashIrvin@unaffiliated/washirving] has joined ##hplusroadmap19:50
kanzure"When it’s 3 A.M., and you’ve been debugging for 12 hours, and you encounter a virtual static friend protected volatile templated function pointer, you want to go into hibernation and awake as a werewolf and then find the people who wrote the C++ standard and bring ruin to the things that they love."19:56
kanzure"To a first approximation, we can say that accidents are almost always the result of incorrect estimates of the likelihood of one or more things."20:00
kanzure"A typical designer (20 years into a 40 year career) • Has less than 5,000 hours of real hands-on system experience … and, almost none of this is in the system’s real environment (When was the last time you saw a designer riding in an electronics bay of an aircraft?)"20:00
kanzurethis is from http://www.rvs.uni-bielefeld.de/publications/DriscollMurphyv19.pdf20:00
kanzure"We cannot rely on our experience-based intuition to determine20:01
kanzurewhether a failure can happen within required probability limits"20:01
kanzureer, i meant "We cannot rely on our experience-based intuition to determine whether a failure can happen within required probability limits."20:01
kanzure" An effect crossing a sufficient number of layers of abstraction [in some system's design] is indistinguishable from magic*"20:03
kanzure"How many Failure Modes and Effects Analysis (FMEA) procedures ask what would happen if one electrical part (a diode) changed into another (a capacitor)? At the electrical part level, this appears to be magic. And, yet, the simplest of failures (a crack) caused this transmogrification. The literature includes other examples of capacitors becoming resistors, transistors becoming silicon controlled rectifiers (SCRs), amplifiers becoming ...20:05
kanzure... oscillators, …"20:05
kanzure"I watched Roy Walford miss out on grants because he did Biosphere II, and wrote popular books. [...] And even Jared Diamond outgrew UCLA, though he had his ducks in a row, and he did everything right as an metabolism specialist."20:23
kanzuremetabolism? i wouldn't have guessed20:23
kanzure"Next year companies will make final investment decisions (FIDs) on a total of 800 oil and gas projects worth $500 billion and totalling nearly 60 billion barrels of oil equivalent, according to data from Norwegian consultancy Rystad Energy."20:33
kanzure"The brains of these early hominins were about the same size as that of a chimpanzee, although it has been suggested that this was the time in which the human SRGAP2 gene doubled, producing a more rapid wiring of the frontal cortex"20:44
kanzure"This protein in humans has been duplicated three times in the human genome in the past 3.4 million years, one duplication 3.4 million years ago (mya) called SRGAP2B, a second duplication 2.4 mya (called SRGAP2C) and one final duplication ~1 mya (SRGAP2D). The ancestral gene SRGAP2 is found in all mammals and the human copy has been renamed SRGAP2A. The 2.4 million year-old duplication (SRGAP2C) expresses a shortened version that 100% of ...20:45
kanzure... humans possess.[6] This shortened version SRGAP2C inhibits the function of the ancestral copy SRGAP2A and (1) allows faster migration of neurons by interfering with filopodia production and (2) slows the rate of synaptic maturation and increases the density of synapses in the cerebral cortex.[4]"20:45
kanzure(from <http://en.wikipedia.org/wiki/SRGAP2>)20:45
kanzure"This increase in human brain size is equivalent to every generation having an additional 125,000 neurons more than their parents."20:46
-!- yashgaroth [~ffffff@2606:6000:cb85:6a00:a4c1:2b23:d54c:e64a] has quit [Quit: Leaving]21:24
-!- Burnin8 [~Burn@pool-71-191-174-26.washdc.fios.verizon.net] has joined ##hplusroadmap21:26
-!- Burninate [~Burn@pool-71-191-174-26.washdc.fios.verizon.net] has quit [Ping timeout: 258 seconds]21:30
-!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has quit [Remote host closed the connection]21:36
kanzure"mindreading (also known as ‘theory of mind’)"21:51
kanzurehttps://events.ccc.de/congress/2014/Fahrplan/events/5956.html22:11
kanzure"In this presentation we will be discussing the path we took to successfully develop our own private server for Metal Gear Online on the Sony PlayStation 2 and PlayStation 3 video game consoles. Interestingly enough this was a private server that was developed after the original was already taken offline, so we did not have a live active server to help with the reverse engineering. Due to this we ran into some issues but ultimately ...22:11
kanzure... succeeded. We believe that the details of the techniques that we used will prove useful for anyone attempting similar actions in the future. The topics that we will discuss in this talk will cover a wide range of high and low level issues related to network protocol and binary reversing."22:11
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 240 seconds]22:25
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Ping timeout: 245 seconds]22:32
-!- cpopell2 [~cpopell@c-76-26-144-132.hsd1.dc.comcast.net] has quit [Max SendQ exceeded]22:34
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap22:34
-!- Guest84438 is now known as maaku23:08
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap23:13

Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!