2014-12-19.log

--- Day changed Fri Dec 19 2014
-!- thundara [~thundara@despair.OCF.Berkeley.EDU] has joined ##hplusroadmap00:00
-!- shubhamgoyal [~shubhamgo@119.56.119.40] has joined ##hplusroadmap00:13
-!- shubhamgoyal [~shubhamgo@119.56.119.40] has quit [Read error: Connection reset by peer]00:13
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has joined ##hplusroadmap00:47
-!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 265 seconds]01:13
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has quit [Remote host closed the connection]01:17
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Remote host closed the connection]01:27
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap01:27
-!- ebowden_ [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap01:30
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Ping timeout: 258 seconds]01:32
-!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap01:37
-!- poppingt` [~poppingto@105.231.27.142] has joined ##hplusroadmap01:41
-!- poppingt` is now known as poppingtonic01:42
-!- poppingtonic is now known as Guest2604501:42
-!- Guest26045 is now known as poppingtonic01:44
-!- poppingtonic is now known as Guest7464501:44
-!- Guest74645 is now known as poppingtonic`01:44
-!- poppingtonic` is now known as poppingtonic01:45
-!- poppingtonic [~poppingto@105.231.27.142] has quit [Changing host]01:45
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has joined ##hplusroadmap01:45
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has quit [Remote host closed the connection]02:21
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap02:22
-!- poppingt` [~poppingto@105.231.27.142] has joined ##hplusroadmap02:22
-!- poppingt` [~poppingto@105.231.27.142] has quit [Remote host closed the connection]02:25
-!- poppingt` [~poppingto@105.231.27.142] has joined ##hplusroadmap02:28
-!- poppingt` [~poppingto@105.231.27.142] has quit [Remote host closed the connection]02:31
-!- shubhamgoyal [~shubhamgo@spnp240085.spnp.nus.edu.sg] has joined ##hplusroadmap02:47
-!- poppingt` [~poppingto@105.231.27.142] has joined ##hplusroadmap02:48
-!- poppingt` is now known as poppingtonic02:49
-!- poppingtonic [~poppingto@105.231.27.142] has quit [Changing host]02:49
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has joined ##hplusroadmap02:49
-!- shubhamgoyal [~shubhamgo@spnp240085.spnp.nus.edu.sg] has quit [Remote host closed the connection]02:52
-!- shubhamgoyal [~shubhamgo@spnp240085.spnp.nus.edu.sg] has joined ##hplusroadmap02:55
-!- shubhamgoyal [~shubhamgo@spnp240085.spnp.nus.edu.sg] has quit [Remote host closed the connection]02:56
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has quit [Remote host closed the connection]02:59
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has joined ##hplusroadmap03:00
-!- shubhamgoyal [~shubhamgo@spnp240085.spnp.nus.edu.sg] has joined ##hplusroadmap03:06
-!- ThomasEgi [~thomas@2a02:810b:33f:dc18:fcaf:895c:2533:ba1e] has joined ##hplusroadmap03:10
-!- ThomasEgi [~thomas@2a02:810b:33f:dc18:fcaf:895c:2533:ba1e] has quit [Changing host]03:10
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap03:10
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap03:26
-!- ebowden_ [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Read error: Connection reset by peer]03:27
-!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has quit [Remote host closed the connection]03:28
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has quit [Remote host closed the connection]03:30
-!- shubhamgoyal [~shubhamgo@spnp240085.spnp.nus.edu.sg] has quit [Remote host closed the connection]03:42
-!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 240 seconds]03:51
-!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap04:12
-!- shubhamgoyal [~shubhamgo@84-37.comp.nus.edu.sg] has joined ##hplusroadmap04:18
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Remote host closed the connection]04:22
-!- ebowden_ [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap04:24
-!- shubhamg_ [~shubhamgo@84-37.comp.nus.edu.sg] has joined ##hplusroadmap04:25
-!- shubhamgoyal [~shubhamgo@84-37.comp.nus.edu.sg] has quit [Read error: Connection reset by peer]04:25
-!- eudoxia [~eudoxia@r190-133-112-194.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap04:32
archelswhy the devil is Taylor & Francis mailing me on my personal e-mail account that I now have access to Connection Science till Dec 31st?04:42
archelsjust... out of the blue04:43
-!- weles [~mariusz@wsip-174-78-132-9.ri.ri.cox.net] has joined ##hplusroadmap05:07
-!- eudoxia [~eudoxia@r190-133-112-194.dialup.adsl.anteldata.net.uy] has quit [Quit: Leaving]05:27
-!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 256 seconds]05:30
-!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap05:38
archelshttp://www.mauve.plus.com/opensourcehw.txt05:43
archels"Why open-source hardware is hard"05:43
kanzure"“Our proposed use of seasteading for medical research,” says Allison, “will be designed to supplement the FDA and shift some of the regulatory burden away from their overworked team and to a more agile regulatory option. If properly organized and with careful design, seasteading has the potential to alleviate some of the stress on regulatory agencies by taking on medical research …  Simply stated, I think we can mirror the ...05:46
kanzure... beneficial aspects of existing review structures and combine them with a lean, agile seastead venture.”"05:46
kanzure"The FDA, founded in 1906, has held a monopoly over the production of regulations in the US medical field for more than a century, and the citizens they are charged with serving need them to be challenged by offshore competitors to reinvigorate its mission.  "05:46
kanzure"Nothing can match the awe and power of our own bureaucracy."05:47
-!- eudoxia [~eudoxia@r190-133-112-194.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap05:59
-!- ebowden_ [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Ping timeout: 272 seconds]06:06
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap06:10
kanzure"Proof-of-concept: Deterministic SSH (ed25519) key generation from a master secret + key handle"" https://github.com/mithrandi/ssh-key-generator06:17
kanzure"If you use the same keypair to access someaccount@somehost and otheraccount@otherhost, someone who observes that both accounts have the same public key for login knows that the two accounts (most probably) belong to the same person. Sometimes this is undesireable for privacy reasons, but the only way to avoid it is to generate two different keypairs for use with the two different hosts. Generating a keypair is not hard, but managing ...06:17
kanzure... large numbers of keypairs like this (eg. one per host) individually is difficult and tedious to do; this proof-of-concept shows that you can have the multiple keypairs without needing to store / manage / back up anything more than a single "master key". Meanwhile, someone looking at the public keys can not tell that the keys are related, so privacy is retained."06:17
-!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 264 seconds]06:39
-!- eudoxia [~eudoxia@r190-133-112-194.dialup.adsl.anteldata.net.uy] has quit [Remote host closed the connection]06:46
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Remote host closed the connection]06:48
-!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 240 seconds]07:12
-!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has quit [Ping timeout: 245 seconds]07:12
-!- shubhamg_ [~shubhamgo@84-37.comp.nus.edu.sg] has quit [Remote host closed the connection]07:16
-!- shubhamgoyal [~shubhamgo@84-37.comp.nus.edu.sg] has joined ##hplusroadmap07:17
-!- shubhamgoyal [~shubhamgo@84-37.comp.nus.edu.sg] has quit [Ping timeout: 258 seconds]07:21
-!- Viper168_ [~Viper@unaffiliated/viper168] has joined ##hplusroadmap07:37
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 250 seconds]07:39
-!- thundara [~thundara@despair.OCF.Berkeley.EDU] has quit [Quit: Lost terminal]07:47
-!- thundara [~thundara@despair.OCF.Berkeley.EDU] has joined ##hplusroadmap07:47
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has joined ##hplusroadmap07:52
-!- shubhamg_ [~shubhamgo@118.189.209.93] has joined ##hplusroadmap07:54
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has quit [Ping timeout: 255 seconds]07:57
-!- Vutral [ss@mirbsd/special/Vutral] has quit [Ping timeout: 245 seconds]08:06
-!- shubhamg_ [~shubhamgo@118.189.209.93] has quit [Read error: Connection reset by peer]08:08
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has joined ##hplusroadmap08:08
-!- Vutral [ss@mirbsd/special/Vutral] has joined ##hplusroadmap08:22
-!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has joined ##hplusroadmap08:28
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has quit [Remote host closed the connection]08:34
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has joined ##hplusroadmap08:34
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has quit [Ping timeout: 255 seconds]08:39
-!- nmz787_i [nmccorkx@nat/intel/x-zsdqhhzkpqeshzgj] has joined ##hplusroadmap08:57
-!- nmz787_i1 [~nmccorkx@134.134.139.76] has joined ##hplusroadmap08:58
-!- nmz787_i [nmccorkx@nat/intel/x-zsdqhhzkpqeshzgj] has quit [Ping timeout: 250 seconds]09:02
-!- cuba [~qba@static.217.217.251.148.clients.your-server.de] has quit [Read error: Connection reset by peer]09:24
-!- cuba [~qba@static.217.217.251.148.clients.your-server.de] has joined ##hplusroadmap09:25
-!- cuba [~qba@static.217.217.251.148.clients.your-server.de] has quit [Read error: Connection reset by peer]09:27
-!- cuba [~qba@static.217.217.251.148.clients.your-server.de] has joined ##hplusroadmap09:30
-!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap09:33
-!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has joined ##hplusroadmap09:47
-!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap09:49
-!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 244 seconds]10:43
-!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap10:55
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has joined ##hplusroadmap11:03
-!- cuba_ [~qba@static.217.217.251.148.clients.your-server.de] has joined ##hplusroadmap11:07
-!- saurik_ [saurik@carrier.saurik.com] has joined ##hplusroadmap11:08
-!- saurik [saurik@74.208.44.105] has quit [Ping timeout: 244 seconds]11:08
-!- Vutral [ss@mirbsd/special/Vutral] has quit [Ping timeout: 244 seconds]11:08
-!- JayDugger [~jwdugger@pool-173-57-55-138.dllstx.fios.verizon.net] has quit [Ping timeout: 244 seconds]11:08
-!- soylentbomb [~k@unaffiliated/soylentbomb] has quit [Ping timeout: 244 seconds]11:08
-!- cuba [~qba@static.217.217.251.148.clients.your-server.de] has quit [Ping timeout: 244 seconds]11:08
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Ping timeout: 244 seconds]11:08
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Ping timeout: 244 seconds]11:08
-!- JayDugger [~jwdugger@pool-173-57-55-138.dllstx.fios.verizon.net] has joined ##hplusroadmap11:08
-!- soylentbomb [~k@d53-64-122-232.nap.wideopenwest.com] has joined ##hplusroadmap11:08
-!- nmz787_i [~nmccorkx@192.55.55.39] has joined ##hplusroadmap11:09
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap11:10
-!- nmz787_i1 [~nmccorkx@134.134.139.76] has quit [Ping timeout: 265 seconds]11:11
-!- Vutral [ODnrGzs1Cm@mirbsd/special/Vutral] has joined ##hplusroadmap11:11
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap11:12
-!- nmz787_i1 [~nmccorkx@134.134.139.76] has joined ##hplusroadmap11:32
-!- nmz787_i [~nmccorkx@192.55.55.39] has quit [Ping timeout: 240 seconds]11:34
-!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 255 seconds]11:42
-!- Twey [~twey@unaffiliated/twey] has quit [Ping timeout: 265 seconds]11:48
kanzurehttps://medium.com/@maradydd/when-nerds-collide-31895b01e68c12:02
kanzurealthough i would rather see her writing articles about the brain parts she keeps12:03
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-qhschfvoelcvyprv] has joined ##hplusroadmap12:03
kanzuretmux ctrl-b z (zoom) is the best thing since sliced bread12:12
-!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 265 seconds]12:45
-!- weles [~mariusz@wsip-174-78-132-9.ri.ri.cox.net] has quit [Ping timeout: 272 seconds]12:58
-!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has quit [Remote host closed the connection]13:03
fenndo you usually have multiple panes up in a single terminal?13:10
-!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has joined ##hplusroadmap13:10
fenn"With readline, ^A is beginning-of-line, ^B is backward-char13:12
fennWith the default settings, to go to the beginning-of-line you have to – ^A a inside a screen session – while ^A would be enough inside a tmux session. Now, to go one char backward, you have to – ^B within screen – ^B b within tmux13:12
fennWhich of these two actions do you do the most? I bet this is the first!13:12
fennNow, as a few others, I’v longed solved this problem, by using M-w as the prefix (and M-a and M-q for nested and nested-nested screen-or-tmux session)13:12
fennAs a bonus, (this is actually why I originally started looking for a new prefix), you may also find these prefixes are easier&quicker to use than either ^A or ^B!"13:12
fennthe windows key is just sitting there staring at me mockingly13:13
kanzurefundrawtransaction https://github.com/bitcoin/bitcoin/pull/550313:16
kanzurethis is a useful thing13:16
fennfun draw the funnest drawing program around13:17
fennhow is this different from just creating an unfunded transaction?13:18
kanzurethis selects unspent outputs to use as the input to the transaction you are creating13:18
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has quit [Ping timeout: 245 seconds]13:23
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has joined ##hplusroadmap13:31
delinquentmeso IO limited processes...13:35
delinquentmethrow redis at it right?13:35
archelsNew Myo Releases - Including Raw Data!13:42
archels"The biggest addition to the Windows and Mac SDKs is the ability to capture raw EMG data using the onEmgData() event."13:42
archelswell, that's about the best response to a consumer outcry ever13:42
archelsI know they already did have something in the pipeline when that happened, but they sure spun it around nicely13:43
fennthat should be the first thing they published13:46
fennall the fancy algorithms come next13:46
fenndon't congratulate someone for dragging their feet until people complained about basic functionality13:47
fenn"All an attacker needs in order to exploit Misfortune Cookie is to send a single packet to your public IP address. No hacking tools required, just a simple modern browser."13:52
fenn(assuming you have a vulnerable router)13:52
fenn"esearchers have distinctly detected approximately 12 million readily exploitable unique devices connected to the Internet"13:54
archelsnumber of ISPs that are going to write to their customers: 013:54
fennmaybe someone could do a remote exploit that shows a warning page13:55
nmz787_i1delinquentme: or use threads13:55
kanzuresee bitcoin-dev convo13:56
fenni'm going to have to see a lot of transaction subgraphs before i really understand bitcoin privacy implications14:01
fennthe coinalytics stuff didn't make a lot of sense unfortunately14:01
fennmisfortune cookie affects router "models by D-Link, Edimax, Huawei, TP-Link, ZTE, and ZyXEL," and 200 others14:02
fennalso a device vendor could remotely patch their device by using the exploit?14:03
delinquentmelololol sooo im about to take a hard up negotiations call14:04
delinquentmeto calm ones self : start at the physiological level.14:04
fennput on your suit14:04
delinquentmedeep breaths, heating pad on feet, warm tea14:04
fenn.wik hachimaki14:05
yoleaux"A hachimaki (鉢巻, "helmet-scarf") is a stylized headband (bandana) in Japanese culture, usually made of red or white cloth, worn as a symbol of perseverance, effort, and/or courage by the wearer." — http://en.wikipedia.org/wiki/Hachimaki14:05
fennmagnesium citrate or beta blockers14:05
delinquentmehow about household items?14:06
delinquentmeis that in something I can drink?14:06
fenna banana has some magnesium14:07
fennepson salt bath lasts a few hours14:07
fennalso it's best to stand up14:08
delinquentmewhen taking calls?14:08
fennyeah14:08
fennunless you're invalid or something14:08
fennthe point here is to be alert and confident, but not in panic mode14:09
fenncaffeine helps with alertness, standing and power poses helps with the confidence, magnesium reduces adrenaline effects14:10
fennalso did you read the kalzumeus article on negotiating?14:11
-!- juri_ [~juri@50.242.254.37] has joined ##hplusroadmap14:11
fenn.g kalzumeus salary negotiation14:11
yoleauxhttp://www.kalzumeus.com/2012/01/23/salary-negotiation/14:11
fennprobably too long to read now14:12
fenn"You know what people find persuasive?  Their own words.  People love their own words.  When you talk to them, you should use their own words."14:13
fenn"do not start negotiating until you already have a Yes-If.  (Yes-If we agree on terms.)  Do not start negotiating from No-But.  (No-But we might hire you anyway if you’re really, really effing cheap.)"14:13
kanzurei think the coinalytics stuff may make more sense if they gave you access to actual graphs14:16
kanzurerather than just blog posts14:16
kanzure(i have only seen thier blog posts)14:16
kanzuredelinquentme: if you want you can phone me up and we can talk salary 512-203-050714:17
fennhttp://www.zdnet.com/article/coin-sized-nuclear-batteries-to-revolutionise-electronics/14:17
delinquentmekanzure, actually that would be huge -- you can chat now?14:17
-!- juri_ [~juri@50.242.254.37] has quit [Ping timeout: 265 seconds]14:17
kanzureyep14:18
-!- skyraider [uid41097@gateway/web/irccloud.com/x-jvaqrrlaonhbmhkv] has joined ##hplusroadmap14:22
-!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap14:24
-!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 265 seconds]14:29
-!- juri_ [~juri@50.242.254.37] has joined ##hplusroadmap14:29
nmz787_i1fenn: oO this article seems pretty good14:32
nmz787_i1great timing as I may be able to negotiate some in the next week or two14:33
-!- juri_ [~juri@50.242.254.37] has quit [Read error: Connection reset by peer]14:33
-!- juri_ [~juri@50.242.254.37] has joined ##hplusroadmap14:34
fenni just filter stuff and regurgitate it; i am a nauseated sponge14:34
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has quit [Read error: Connection reset by peer]14:41
kanzureas for transaction graphs i usually just click around on blockchain.info or webbtc.com14:43
kanzureif you would like a full blockchain dataset i could make one available for you to play with14:44
kanzurehttp://www.scientificamerican.com/article/monkey-see-monkey-speak-video/14:48
nmz787_i1someone recommended reading this re: software engineering... I haven't read it yet http://www.ucalgary.ca/icic/files/icic/49-IJSSCI-1201-CogFoundSE.pdf14:48
nmz787_i1that krak.wav didn't have any sound in it that sounded like 'crack' to me14:51
-!- juri_ [~juri@50.242.254.37] has quit [Ping timeout: 245 seconds]14:52
-!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap14:54
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has joined ##hplusroadmap15:00
kanzure"It's basically a specialization of the base rate fallacy. If, say, idiots can still be right by chance 10% of the time, and experts are right 90% of the time, but only 10% of the population are experts, then lucky idiots will make up fully half of the population of people who are right."15:00
-!- nmz787_i1 [~nmccorkx@134.134.139.76] has quit [Ping timeout: 265 seconds]15:00
kanzurefrom https://news.ycombinator.com/item?id=877314915:01
kanzure"I'm representing one of the competitor payloads, lettuceonmars. We originally wanted to send algae too, but had to rule them out because of payload specifications. More info about our porjct ca be found here: <http://www.lettuceonmars.com/>"15:08
kanzurevia Ravasz <ravaszmeister@gmail.com>15:09
-!- nmz787_i [~nmccorkx@192.55.54.42] has joined ##hplusroadmap15:24
-!- Viper168_ [~Viper@unaffiliated/viper168] has quit [Ping timeout: 250 seconds]15:40
-!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has quit [Remote host closed the connection]15:42
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap15:48
fenn“This is the first time that a professional linguist has tackled the data on call combinations in wild monkeys.”16:11
fennfor shame16:11
fennshameful if it's true, shameful if he was wrong and said it anyway16:12
kanzurein conclusion, we're screwed16:14
fennwas there actually a video?16:15
kanzuredunno, sorry16:15
fennlettuce is the most useless food you could pick16:18
-!- DumpsterD1ver [~loki@vpn166.sdf.org] has joined ##hplusroadmap16:24
kanzurehacker news says "already been done for dogs" http://languagelog.ldc.upenn.edu/myl/llog/FarsideDogTranslator.jpg16:49
fennHeyyyyyyyy...16:52
kanzure"the great list of things that you thought people had already taken care of, but turns out nobody has done, updated for the poor souls who may come across this in 20 or 40 years"16:57
fenni'm afraid to look16:57
-!- poppingtonic [~poppingto@unaffiliated/poppingtonic] has quit [Remote host closed the connection]16:58
kanzurenah, we'd probably have to write it16:58
fennomg what an epic fail on the part of blockchain.info http://www.reddit.com/r/Bitcoin/comments/2onm5r/blockchaininfo_security_disclosure/cmosmbn17:04
kanzurethey are not known for their skillz https://people.xiph.org/~greg/21mbtc.png17:05
fenniirc this is the exact same crypto fail that let people sign PS4 games/install linux again17:10
fennnot using a random nonce for ECDSA17:11
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has joined ##hplusroadmap17:14
kanzurei have a really bad idea and i hate myself for having this idea17:17
kanzurei don't know what this salary-tithing trend is really called, but people seem to do that17:18
fennwhich salary-tithing trend17:18
kanzureand singinst used to spend a ridiculous amount of effort shaming everyone into donating 50% of their salaries towards anti-ai research17:18
kanzure"because this is the most important importantness you will ever have influence on, in the history of forever"17:18
fennwell, potentially17:18
kanzure(i don't even know if charitynavigator covers these schemes)17:18
kanzurei was writing an email to reason@fightaging.org about this, but then decided against sending this straight to him, because he keeps secrets17:19
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Ex-Chat]17:19
kanzureand uh, the reason why i was typing to him was because of a stupid grg thread17:19
kanzurei think grg is filtering my emails, there should be at least one email from me in that thread, but whatever17:19
kanzureanyway, the awful idea goes like this:17:19
kanzuresomeone should put kickstarter-style campaigns together for salary-tithing-style campaigns17:19
kanzuree.g. get a group of people to send 40% of their salaries only if $400M is reached or whatever17:20
kanzureer, $400M/year17:20
fenni guess i don't see the point of making it a recurring cost17:20
kanzurehopefully this would be implemented not in the name of charity but in the name of projects that need to get done17:20
delinquentmeyou're convincing people ti give up 40% of their income?17:20
kanzureno17:21
kanzurei am just reporting about that17:21
fennrecurring costs leads to waste instead of fixing stuff17:21
fennsee any federal agency for example17:21
kanzuremy point though is that, as much as "send 40% of your salary to singinst" exists (and it's not singinst's idea by far), group-style "only if everyone else does it" hasn't been used on that idea yet17:22
fennso this is just kickstarter with a recurring donation17:22
kanzurei hate donations17:23
fennis it not a donation?17:23
kanzurethis is also why i hate myself for even talking about this trend17:23
kanzureno i just hate donations17:23
fennsome kickstarters are basically product launch with early reservations17:23
kanzurewell you're the one that said donation17:23
fennyou could have a similar service where you sign up for fair trade beard razors or whatever17:24
kanzureoh shit i said donating17:24
kanzurebut that was about singinst17:24
kanzureso you can't call me on that17:24
fenni dont see anything wrong with the idea of donations in general17:25
fenni wish people would do at least the same amount of research they do for buying a consumer product17:25
fennthere are a lot of corrupt "charities" that just embezzle 99% of the money, but you'd never know without doing research on them17:26
kanzurethere are many reasons why donations are a bad idea (perhaps the most recent one that i've stumbled into is "fuck your free shirts, we're trying to start a t-shirt economy and you're killing our businesses by injecting free crap into our economy.")17:26
fennoh noes free stuff, call the libertarian police17:27
kanzureonly if you pay for the phone call17:27
fenni hope you realize how dumb that sounds, reading this on your linux-powered computer17:27
kanzuret-shirts are not software17:28
kanzurenot yet, anyway17:28
fennthere's a difference between a (tax-funded) subsidy and a charity17:29
fennalso technological innovation can reduce costs to zero17:29
fennthese things can all reduce revenue, or they can increase it17:29
* fenn shrugs17:30
kanzureshrug, another reason i am17:30
kanzurehey don't shrug at my shrug17:30
fennhey man i invented shrugging, you're infringing my social communication patent17:30
fennmethod of conveying apathy in ascii text stream17:31
kanzureif you are going to flood people with free stuff, flooding them with useful things like free software, cnc machines and the science ark would be much better than just consumer goods17:31
fennyou're missing the point entirely17:31
fennthe t-shirts are advertising for the charity, cleverly hacked into cognitive biases that say "a shirt is useful!" and get people to advertise for them unwittingly17:32
fennsomeone had to pay for the shirt at some point17:32
fennthat shirt cost comes out of their advertising budget17:32
kanzurethe t-shirts are advertising for the charity? when they send the t-shirts to $crapcountry?17:32
fennoh, that's a different situation. in that case the shirts are literally a donation in kind17:33
kanzurewhat situation did you think i was talking about17:33
fennor recycling, if you're cynical17:33
fennthe situation where they send you swag17:33
kanzureoh, that's just annoying17:33
fennok i see where you're coming from now17:34
fennunintended consequences of "the white man's burden" (for lack of a better phrase)17:35
kanzurethere's also some rant to be had about supply/demand/pricing and donations not being linked to market efficiency17:35
kanzurealthough donations are often linked to "market(ing) efficiency" hehe aren't i clever, i'll go sit down now17:35
fennhow do you type standing up? :)17:36
kanzuredesk happens to be correct height17:36
kanzureit's pretty nice for run-by typing and ranting17:36
-!- yashgaroth [~ffffff@2606:6000:cb85:6a00:3da2:23d0:76b4:3e6] has joined ##hplusroadmap17:37
fennwrt the t-shirt economy, that's a problem with income inequality and trade imbalance17:37
fennif t-shirts were worth something in $donating_country they probably wouldn't be giving them away wholesale17:37
kanzurethey should be trying to bootstrap a stable t-shirt printing industry (it's not like screen printing is black magic these days)17:38
fennbut t-shirts would still not be worth anything in $donating_country and the trade imbalance would persist17:38
kanzurewait why did i mention printing17:38
fenneven though $receiving_country has a t-shirt production business17:38
kanzureprinting is after you have t-shirt17:38
fennwhatever17:39
fennmost t-shirts are made in bangladesh17:39
fennthey at least have an export economy there17:40
fennanyway i agree that giving people capital goods is a better use of your charity dollars17:41
fennassuming they want it and can use it17:41
fennsomething about a replicator being used as a goat milking stool17:41
fennout of cartridges the milkmaid said17:42
fennan e-beam metal powder sintering machine on the ISS would have to be spinning to create centrifugal force to settle the dust17:43
-!- Quellvisk [~Quellvisk@gateway/tor-sasl/quellvisk] has joined ##hplusroadmap17:44
fennhello Quellvisk17:44
-!- Quellvisk [~Quellvisk@gateway/tor-sasl/quellvisk] has left ##hplusroadmap []17:44
fennwas it something i said17:44
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap17:44
fennQuellvisks (quell='kill, crush completely' visk='move quickly') are - or rather, were, since they're all extinct by now (see below) - huge and crushingly stupid herbivorous ruminants17:45
fennhttp://motherfuckingwebsite.com/  yesss17:51
-!- nmz787_i [~nmccorkx@192.55.54.42] has quit [Quit: Leaving.]17:51
jrayhawkhaha "<!-- yes, I know...wanna fight about it? --><script>[Google analytics shit]</script>"17:56
fennGood riddance, Goxchain!17:57
jrayhawkhttp://txti.es/ i like his captcha17:58
fennat least it tells you to leave it blank17:58
fenni've been classified as a robot because i put shit in fields that weren't labeled17:58
fennhow does the "show more options" button work?17:59
jrayhawksame way as every other button?18:00
fennhm ok nevermind18:00
fennsame way as every other button before 200018:00
jrayhawki assume you've seen how the CGI standard works18:01
fennyes18:01
fenni was confused because the url didn't change18:01
fennand it loaded/displayed really fast so i thought it was the same page18:02
fennonly accepts lowercase letters (a-z), numbers (0-9), and dashes (-). If you put anything else in there, txti will change it.18:07
fennthat's an odd decision, as punctuation is generally considered helpful18:07
fennoh that's the url18:08
kanzure17:17 < dgenr8> mempool art, 500+ tx all invalidated by a single respend http://i.imgur.com/ZTKAuqz.png18:11
kanzure17:47 < dgenr8> ah here it is... https://github.com/weilu/bitcoin-tx-graph-visualizer18:11
fennwtf kind of image url is this: https://camo.githubusercontent.com/6ea7433b519f66576c487ae4d038b9209eb51c86/68747470733a2f2f7261776769746875622e636f6d2f7765696c752f626974636f696e2d74782d67726170682d76697375616c697a65722f6d61737465722f6578616d706c652f7472616e73616374696f6e732e73766718:13
kanzurethey can't host content on github.com anymore because Daeken18:13
fennobviously i missed something18:16
kanzurealso this is an svg file and i think the filename might be related to svg things18:16
fennit's a ridiculously long hash18:16
kanzureDaeken did some proof-of-concepts for embedding html and js inside images that gets executed by most modern browsers18:16
kanzureso various cross-domain cookie-hijacking session-hijacking html5-js-api implications18:16
fenn"let's parse images for js just in case"18:17
fennwhat the bloody fucky18:17
kanzuresee https://news.ycombinator.com/item?id=420905218:17
kanzure"A PNG that's interpreted as HTML and loads itself as compressed JavaScript"18:17
fennis it just messing with the mime-type in the header?18:18
jrayhawkthe w3 doesn't care about security and it's impossible for anyone to simultaneously comprehend every w3 standard, so we get undefined emergent properties of standards that vendors can only write ad-hoc policy fixes to as they're discovered18:19
fenni mean i understand that images can contain other files, but why the fuck would a browser load html data from an image18:19
fennin an <img> tag no less18:19
kanzurebecause the browser renders almost any html it sees18:19
kanzureno the <img> tag you are seeing is inside the image18:19
fenni'm not seeing any <img> tag (though i believe this works on other browsers)18:20
kanzure-Vœi¼jawh<img onload=with(document.createElement('canvas'))p=width=4968,(c=getContext('2d')).drawImage(this,e='',0);while(p)e+=String.fromCharCode(c.getImageData(0,0,p,1).data[p-=4]);(t=top).eval(e) src=#>IDATxœ]TÛnã6Ú18:20
fennnope18:20
kanzureanother technique is http://demoseen.com/windowpane/nufl0wer.png.html18:21
kanzureabout http://daeken.com/superpacking-js-demos18:21
fenn‰PNG  IHDRIDATxœ]TÛ...18:21
fennhmm this didnt paste right but whatever18:21
kanzure"Rule #1 of shrinking demos: removing a byte becomes more difficult every time you remove a byte. It starts off trivial and runs very quickly into a brick wall."18:22
jrayhawkDaeken: can you add ;charset=UTF-8 to your Content-Type header?18:24
jrayhawk'Nicolás Alvarez' seems like a funny name18:25
kanzure18:18 < kanzure> so what sort of regular rate of transaction-invalidation do you see in your mempool?18:26
kanzure18:19 < dgenr8> only 10-20 a day http://respends.thinlink.com18:26
kanzure18:21 < dgenr8> a few weeks ago there was another pattern.  only 50 or so txes, but arranged in a dazzling self-referential mess18:26
fennmaybe it was a trading site that got a stale block18:30
fennthat superpacking js in images thing could be big trouble for malware-infected-website scanners18:34
-!- Qfwfq [~WashIrvin@unaffiliated/washirving] has quit [Ping timeout: 272 seconds]18:34
fennthe js can encrypt the malware, or at least scramble it enough that simple pattern matchers don't see anything bad18:34
kanzurei dunno if the js has access to content outside the image actually, i haven't tried18:35
fennit doesn't have to; it can just contain an entire webpage18:35
fenninside the "image"18:35
kanzurehttp://blog.sucuri.net/2014/02/new-iframe-injections-leverage-png-image-metadata.html18:36
kanzurehrmm not quite the same thing18:36
kanzure"This last technique is called PNG Bootstraping and is pretty standard practice in the demoscene and creative coding scene focusing on size optimization." hehe18:37
kanzure"In retrospect, the dynamic favicon trick seen in Defender of the Favicon ( also from 2008 ) was abused much quicker, in only 2 years, to phish people using tabnapping."18:38
kanzure"store your data in the number pi" https://github.com/philipl/pifs18:39
kanzurei thought this was a joke until i noticed there seems to be people actually using this18:39
-!- Qfwfq [~WashIrvin@unaffiliated/washirving] has joined ##hplusroadmap18:39
fennhey i came up with that independently18:40
fennArithmetic Coding!18:44
fennwasnt there an altcoin like this18:45
fenn"why use a hash function when we can just use pi, it's random right?"18:46
kanzureat this point there's basically a law that says there already exists an altcoin for every single bad idea you can possibly imagine18:46
kanzuresee http://mapofcoins.com/18:46
fenni need more classes of quotes, sarcastic quotes, scare quotes, verbatim quotes, paraphrase quotes...18:46
kanzure"Power and timing side channels for Strong Physical Unclonable Functions and their efficient exploitation" http://eprint.iacr.org/2013/851.pdf18:47
fennOne coin to rule them all, One coin to find them, One coin to bring them all and in the darkness bind them18:48
kanzureoh how clever -_-18:49
fennit should start way before 200818:50
fenneven checks are a sort of cryptocurrency18:50
kanzureer what18:51
fennbank notes18:51
kanzurehuh?18:51
fennthe security features are based on gravure/guilloche or some kind of engine turning unique set of gear ratios18:52
kanzurego on18:52
fennwell that's basically a private key18:52
kanzurehehe http://retractionwatch.com/2014/12/19/elsevier-retracting-16-papers-faked-peer-review/18:52
kanzure"Sixteen papers are being retracted across three Elsevier journals after the publisher discovered that one of the authors, Khalid Zaman, orchestrated fake peer reviews by submitting false contact information for his suggested reviewers."18:52
kanzureaww "editors have learned to be more skeptical of reviewer details provided by authors, especially contact details not connected to institutions"18:53
kanzure"If someone recommends a reviewer, we suggest [editors] verify the email address against SCOPU"18:54
fennhttps://en.wikipedia.org/wiki/Geometric_lathe18:54
fennThe lathe was able to generate intersecting and interlacing patterns of fine lines in various shapes, which were almost impossible to forge by hand-engraving. They were used by many national mints.18:54
kanzure"A “small minority” to scientists could be translated as 1% or 5%. 1% or 5% of 12.5 million is 125000 and 625000 papers, respectively."18:55
kanzure"In my opinion, based on quite a few stories related to Elsevier, a deep distrust of Elsevier and its editorial integrity, has now set in. And, to cover all 12.5 million papers, Elsevier has hired ONE “full-time staff member with a PhD in physics and history as a managing editor to do the grunt work on cases like this”? Something is really, really wrong with this picture."18:55
fennyou can submit your own reviewers? who thought that was a good idea18:55
kanzurehaha18:55
kanzure"hindsight is 20/20"18:56
kanzurecould it be possible that scientists legitimately don't have any better ideas than conventional papers for communicating ideas18:58
kanzurei really can't think of an alternative explanation that doesn't sound completely crazy18:58
kanzureoh right, not incentivized to fix things(?)18:58
fennthat's more like it18:58
kanzureno alternative is particularly obvious though18:58
fennum, how about the www18:58
kanzuremore of the same18:58
fennmore what?18:59
kanzureokay, more accessible i guess, but you are still writing papers18:59
fennmy point is the www was invented by scientists for scientists to communicate data and ideas18:59
fenn"we" are still stuck in the paper mindset even with ridiculously flexible tools at our disposal19:00
kanzurei suppose i should be glad that they aren't all converting their ideas into superlong videos19:00
kanzurei better not poke that bear19:01
fennor pdf files with embedded executable whatevers19:01
fenni like the idea of something like engelbart's zooming hierarchical interface19:02
fennwhere you have an abstract and table of contents, and the subheadings expand when you click on them19:02
fennwhen everything is expanded you have the raw data, lab notebooks, commentary, everything19:02
kanzuresounds like some terrible contraption invented by borges19:02
fennyes it is potentially infinite19:03
kanzure"you read the sentence, but then you realize it's a book and ted nelson is laughing at you"19:03
kanzuremeh19:03
fennbut it's also just a webpage19:03
fennthis channel is potentially infinite but it's still just an irc channel19:04
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-qhschfvoelcvyprv] has quit [Quit: Connection closed for inactivity]19:04
fennthe structure and language of a scientific paper has changed a lot since the 1600s19:05
kanzurenot many alternatives available19:05
kanzure"all concepts must be submitted as executable code that gets uploaded to the massive science ai, or else" wont work19:05
fenni could suggest some but you'd hat eme19:07
fennascii text has staying power19:07
kanzureat least within math, everything should have importable models and functions19:09
kanzurewell, nearly everything19:09
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has quit [Remote host closed the connection]19:09
fennbut then it would be code!~19:09
kanzurecan't have that19:09
fenndid i mention i hate hieroglyphics19:09
fenna big seduction of math is you can be lazy and abuse your reader by not explaining everything, like what the letters mean19:11
fennthen you get to say "behold, the succinctness of my equation"19:11
fenneven though it's not really reproducible19:11
fennand any compiler worth its salt would throw an error19:12
fennhas anyone done a bitcoin key encoding in tengwar yet19:15
-!- skyraider [uid41097@gateway/web/irccloud.com/x-jvaqrrlaonhbmhkv] has quit [Quit: Connection closed for inactivity]19:15
maakuwhy would you want to?19:16
fenni don't know19:16
fennchinese characters would probably be better19:17
kanzuremaaku: i was thinking about which constraints i would want to impose on a project if i was going to pretend to try to implement something ai-like, and these are the rules i figure are important http://diyhpl.us/wiki/projects/human-like-cognitive-abilities/19:17
kanzuremaaku: can you think of any additional constraints that should be imposed19:17
fenncrypto stuff all kinda looks like this to me: https://en.wikipedia.org/wiki/File:One_Ring_inscription.svg19:18
maakukanzure: strike AI, that sounds like generally good advice for any project ;)19:19
kanzureoh19:19
kanzurewell, i intended it to be more ai-specific19:20
kanzuresurely there are additional constraints that ai projects should have that other projects should not19:20
fenndeep strike AI, for long-range assault in hostile environments19:20
fenn.wik deep strike19:21
yoleaux"An attack aircraft (also called a strike aircraft or attack bomber) is a tactical military aircraft that has a primary role of attacking targets on the ground or sea, with greater precision than bombers, and which is prepared to encounter stronger low-level air defenses." — http://en.wikipedia.org/wiki/Attack_aircraft19:21
kanzureseems like there probably should be ai-specific project rules19:22
kanzurealthough now i am failing to come up with any reasons19:22
fennsounds like begging the question, cart before the horse thinking19:22
kanzurehuh?19:22
fennging why should there be rules?19:23
fenns/ging//19:23
kanzurewell, let's broaden this to be about also brain emulation software development19:23
fennpick one19:23
kanzureand there are clearly some things that are totally inappropriate/wrong to be doing when doing brain emulation development19:23
kanzurelike cyc19:23
kanzurecyc would be a wrong thing to be doing19:23
fenncyc is not brain emulation19:24
kanzurecorrect19:24
kanzurethat is a very good reason why cyc should not be a part of that development effort19:24
fennit's a semantic database of common knowledge19:24
maakusomething more AI specific but along those lines... i would say have an architectural grounding in some sort of analytical theory19:24
kanzureas opposed to a grounding in what?19:24
maakutoo many ai projects are of the sort 1) do something low level at scale, 2) ..., 3) strong ai will "emerge"19:25
fennproof of convergence19:25
maaku(machien learning and neural network people seem to fall pray to this especially)19:26
kanzurehmh19:26
kanzure-h19:27
fennin machine learning we had to show that the optimization function converged; this was done by showing that the kernel was convex by taking the double derivative...19:27
fennit dredged up a lot of calculus i had forgotten and was a pain in the ass with zero utility19:27
fennit seems arrogant to believe you can know the outcome of a system you don't understand19:29
maakufenn: i understand there is plenty of theoretical backing to machine learning. not the point.19:29
fennit's a very linear, low complexity field of math19:30
maakuthe fallacy people fall into is saying "this construct can learn any classification problem", "i theorize that the brain is just 50 billion of these constructs", "therefore, if I had 50 billion of these at scale, that's all i need!"19:30
fennit only gives interesting results because there's a lot of variables to tweak19:30
maakuyou also need to connect those 50 billion things in a specific way, and that's where the magic is...19:31
fenni'm not sure if "eureqa" counts as machine learning or not.19:31
kanzurenow rant about "deep learning"?19:31
kanzure(word2vec and such)19:31
fenndeep learning is just more of the same, piled higher and deeper19:31
fenn(literally)19:31
maakukanzure: no i don't intend to do that. deep learning is an important advance19:31
maakubut it's also not all there is to say as far as intelligence goes19:32
fennthe brain has a lot of other things besides neocortex19:32
fenndeep learning probably accurately models what happens in the neocortex19:32
fennbut it would never feel, or hunger for something, or focus, or remember what happened 3 seconds ago19:33
maakuright19:35
fennword2vec feels like cheating because it skips the whole visual processing part19:36
fennand the auditory processing part of reading19:37
kanzure"Morphometry of human cerebral cortex microcirculation: General characteristics and space-related profiles" http://ftp.tuebingen.mpg.de/pub/kyb/bweber/papers_db/2008/Lauwers/Neuroimage%202008%20Lauwers.pdf19:38
kanzurewhy hasn't todd uploaded 3d models and data dumps at low/high resolution to his serverz19:41
kanzureisn't this something he should be doing19:41
kanzurehe owes us for like, stuff19:41
kanzurehe owes us for our moral support :p19:42
fenngo todd, woo19:43
fennwe're out of peanuts over here19:43
kanzure"In a companion paper (Lorthois et al., Neuroimage, in press), we perform the first simulations of blood flow in an anatomically accurate large human intra-cortical vascular network (~ 10000 segments), using a 1D non-linear model taking into account the complex rheological properties of blood flow in microcirculation."19:45
kanzure"This model predicts blood pressure, blood flow and hematocrit distributions, volumes of functional vascular territories, regional flow at voxel and network scales, etc."19:46
fennjeez this guy in #bitcoin has 256GB of RAM19:46
kanzuremy laptop has 32 GB.. 19:46
fenni guess that's small potatoes compared to supercomputers but still, i'm jealous19:46
fenni have a bag of 1MB ram sticks that i haven't figured out what to do with yet19:51
-!- ebowden [~ebowden@CPE-121-223-157-226.lns2.bat.bigpond.net.au] has joined ##hplusroadmap20:05
kanzurehmm.20:12
kanzureoh i suppose i should add "no emergence" to the list20:14
kanzurebut that's sort of covered by hand waving20:14
kanzure"but what about the experience of actually holding a book in your hands? you can't get that over the net" - batman beyond filling your head with lies since 199820:21
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has joined ##hplusroadmap20:21
fenna book is probably the worst example they could have picked20:22
fenn.wik strange days20:22
yoleaux"Strange Days is the second album by American rock band The Doors, released in September 1967. It was a commercial success, initially earning a gold record and reaching #3 on the Billboard 200 albums chart. The album also yielded two top 30 hits and eventually a platinum certification." — http://en.wikipedia.org/wiki/Strange_days20:22
fennhuh20:22
fenn.wik strange days film20:23
yoleaux"Strange Days is a 1995 Americanscience fiction action thriller film directed by Kathryn Bigelow. Co-written and produced by her ex-husband James Cameron and co-written by Jay Cocks, it stars Ralph Fiennes, Angela Bassett, Juliette Lewis, Tom Sizemore, Vincent D'Onofrio, and Michael Wincott." — http://en.wikipedia.org/wiki/Strange_Days_(film)20:23
fennbah20:23
kanzurein this episode, batman beats up a hyperintelligent gorilla because nobody wants a gorilla walking around town20:27
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has quit [Quit: Leaving]20:27
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has joined ##hplusroadmap20:29
fennracist20:30
fenni'm running out of faux indignation20:31
-!- yottabit [uid36770@gateway/web/irccloud.com/x-pdxapmoviclejnpq] has joined ##hplusroadmap20:40
kanzure"must be hot in that suit" "you get used to it" and he tells jokes20:40
-!- shubhamgoyal [~shubhamgo@118.189.209.93] has quit [Remote host closed the connection]20:42
-!- DumpsterD1ver [~loki@vpn166.sdf.org] has quit [Ping timeout: 245 seconds]20:45
-!- DumpsterD1ver [~loki@vpn166.sdf.org] has joined ##hplusroadmap20:48
kanzurewut https://www.indiegogo.com/projects/the-braincard-powered-by-a-neural-network-chip20:48
kanzure"Neuromorphic Chip with 1024 artificial neurons"20:49
fennis that supposed to be a lot20:49
kanzure"'re looking to raise at least $200k to start manufacturing in volume, which will make the BrainCard as cheap as possible. "20:50
kanzurei don't think so, tim20:50
fennsomething like this shouldn't be on indiegogo20:50
kanzureright. definitely fucked up.20:50
fennhe can't deliver if doesn't make the goal, so where does the money go?20:50
kanzurehttp://neuromorthings.com/20:50
kanzureugh "In this mode the BrainCard acts as a ‘Right’ brain hemisphere for image/pattern recognition and the CPU device acts as a ‘Left’ brain hemisphere for procedural/logic & communications."20:51
fennoh i guess they have 1000 prototypes already which is just about $200k at $199 each20:51
kanzure1024 neurons isn't going to do much20:52
kanzure"we'd like to crowdsource our research and development through you!" because you can't figure out what to do with them20:54
kanzure"FPGA (Xilinx Spartan 6 – XA6SLX16-2CSG324C) with up to 14,579 logic elements"20:55
fenni feel like if it did anything useful they would have demos20:56
kanzureright.. and a non-hardware implementation for demonstration purposes maybe, "it does this, but now imagine that happening 1000x faster"20:57
fennbut 1000x faster gets you nothing because its inputs aren't changing20:58
fennif they could switch between 100 sets of 1024 neurons every second that might be useful20:58
fennthat's almost bee-level20:59
kanzurethat seems like something a gpu would be better at20:59
fennyes20:59
fennreally the hardware should be able to do more20:59
fenna gpu is more general purpose than a hardware neuron simulation21:00
kanzurei strongly doubt these are biologically-accurate neurons too21:00
kanzureprobably just the minimum model21:00
fennfor a given area of silicon, you should be able to fit more simulated neurons on an asic than a gpu21:00
fennit's probably just a neural net21:01
fennok useless day completed21:04
kanzurethat's a lousy way of going about life21:05
fenni liked meredith's article on weird nerds21:06
kanzurei wrote a few mountains of code, wrote specs, tests, and pretended to be a leet haxor21:06
fenni ate coconut yogurt and baked brownies21:07
fenni guess that alt-m screen switching will probably save me from getting carpal tunnel21:08
kanzurei also picked a fight with an ayn rand immortalist so that was a mistake21:08
kanzure(grg crap)21:09
fennayn rand's head was frozen and stored on the secret nazi moon base21:09
kanzurenah his argument was literally "Objectivism is the only way that it is possible to see that government funding of life extension is non-optimal"21:10
fenni don't care who is funding it as long as someone is21:10
kanzureuh, s/extension/extension research/ of course21:10
kanzure(and that's about when i had my terrible idea)21:11
fennall that free science grant money is wrecking our science economy!21:12
kanzureyes21:12
kanzurewell, lots of other things are too21:13
fennif only the illuminati elite were in charge...21:13
kanzurelike intellectual property law21:13
kanzureand broken incentive models that don't work like you'd want21:13
fenn(i was kidding, if it weren't obvious)21:13
kanzureit21:13
kanzureis not21:13
kanzure(standing..)21:13
fennfor the record, i will piss on ayn rand's grave for naming her shitty philosophy "objectivism"21:13
kanzurego find her21:14
kanzure"Valhalla, NY"21:15
fennthere's even a video tour21:15
fenn.title youtu.be/OGgwyW6GNrI21:15
yoleauxAyn Rand Grave - YouTube21:15
fennfitting name for a cemetery town21:17
fennhigh school mascot is the fighting valkyries21:17
-!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap21:17
kanzure"Machine leanring: The high-interest credit card of technical debt" https://static.googleusercontent.com/media/research.google.com/en/us/pubs/archive/43146.pdf21:19
fennis this just saying "keep it simple, stupid"21:21
kanzurestill reading21:22
kanzure"In [7], dependency debt is noted as a key contributor to code complexity and technical debt in classical software engineering settings. We argue here that data dependencies in machine learning systems carry a similar capacity for building debt. Furthermore, while code dependencies can be relatively easy to identify via static analysis, linkage graphs, and the like, it is far less common that data dependencies have similar analysis ...21:23
kanzure... tools. Thus, it can be inappropriately easy to build large data-dependency chains that can be difficult to untangle."21:23
fennthe heroku people had some stuff to say about making sure old code continues working by defining interfaces strictly and static linking (or whatever equivalent is for scripting languages)21:24
kanzure"Or suppose that for efficiency a particular signal will no longer be computed; are all former consumers of the signal done with it? Even if there are no references to it in the current version of the codebase, are there still production instances with older binaries that use it? Making changes safely can be difficult without automatic tooling."21:26
kanzure"A remarkably useful automated feature management tool was described in [6], which enables data sources and features to be annotated. Automated checks can then be run to ensure that all dependencies have the appropriate annotations, and dependency trees can be fully resolved. Since its adoption, this approach has regularly allowed a team at Google to safely delete thousands of lines of feature-related code per quarter, and has made ...21:26
kanzure... verification of versions and other issues automatic. The system has on many occasions prevented accidental use of deprecated or broken features in new models."21:26
fenni'm not really following this article21:28
fennare they talking about how long to cache intermediates?21:29
kanzure"we have a bunch of experience and you should listen to us"21:29
fennwow great, now please explain your wisdom in english21:29
fenna few concrete examples would do wondersf21:30
kanzurethey probably can't talk about proprietary google machine learning thingies21:30
fennit's like they are from a different world that happens to overlap some of the same words we use21:35
kanzuretheir description of 1000 unused codepaths is basically a real thing that happens21:36
fenn"it was found possible to rip out tens of thousands of lines of unused experimental code paths."21:36
fennthat has nothing to do with machine learning, that's just bad software engineering practices21:37
kanzurethis article is mostly about software engineering ("technical debt")21:39
fenni guess "paying technical debt" is a bizarro-world term for "refactoring"21:40
fennjeez they even cite Fowler21:41
maakukanzure: maybe one more constraint: there is no single magic silver bullet which gives general artificial intelligence21:42
fennevolution!21:42
fennand fish!21:42
kanzurenow for your daily dose of "totally motivated reasoning" http://decentral.ca/what-the-data-says-this-is-going-to-escalate-quickly/21:43
maakumaybe this reflects my own biases, but while a special-purpose-everything AI doesn't work (as mentioned), neither does the  "one single universal rule" approach21:43
kanzuresponsored by ray kurzweil's wig21:43
fenn.wik salmon of wisdom21:43
yoleaux"The Salmon of Knowledge (Irish: bradán feasa) is a creature figuring in the Fenian Cycle of Irish mythology. (It is sometimes identified with Fintan mac Bóchra, who was known as "The Wise" and was once transformed into a salmon.)" — http://en.wikipedia.org/wiki/Salmon_of_wisdom21:43
kanzuremaaku: if i had a program that was intelligent, then is that a magic silver bullet?21:44
maakukanzure: talking about the insides of the program21:44
kanzurei think that i would argue that there seems to be a general cognitive architecture "conserved" across species that seems to indicate all known brains operate roughly the same way21:45
kanzurei am having trouble, sorry, i just am i guess (probably 'cause i'm tired)21:45
maakuany AI project with a chance at success needs to be somewhere in the middle ground -- it ivery likely from both AI history and neurological structure that general intelligence consists of multiple modes of thinking, each with strengths and weaknesses, integrated in some way21:45
kanzuresoftware is close to magic anyway or something21:45
maakuyeah i think we're talking past each other21:46
kanzureif you are arguing that "no single method will magically work", then how would anything have a chance of working?21:46
kanzurewe are probably talking past each other, hehe21:46
fennhe's identifying a trend in the unfounded pronouncements of past AI researchers21:46
kanzurei'm with you so far21:47
fenn"neural networks will solve everything" or "lambda calculus is god's gift to mankind"21:47
kanzurehow about something like, "only make claims backed up by evidence"21:47
fennyou mean an existence proof? (ala constructivism)21:47
maakui'm talking about a common failure mode, where a would-be AGI developer shows how neural net update rule X, or cellular automata Y, or technobable gizmo Z is _technically_ capable of learning or expressing any computation21:48
kanzureah i see.... yeah the computational reductionism has been annoying for sure21:48
* kanzure glares at wolfram21:48
fenn.wa the meaning of life, the universe, and everything21:48
yoleauxfenn: Sorry, no result!21:49
maakuright. wolfram, kurzweil, the NARS guy (wang? i forget his name and too lazy to google)21:49
JayDuggerNot even a Douglas Adams quote?21:49
fennsuch a let down21:49
kanzurewolfram probably thinks he's better than douglas adams21:49
kanzureokay, cheap shot21:49
JayDuggerGood point. He has cause.21:49
maakuso, therefore, (this is where they go wrong), all we need is X, Y, or Z running on enough computing power, etc.21:49
-!- shubhamgoyal [~shubhamgo@119.56.115.169] has joined ##hplusroadmap21:50
kanzureyeah i am highly skeptical of "if only we had more computing capacity"21:50
maaku-or- X, Y, Z attached to a proper perceiption system, or whatever supporting component they think is holding us back21:51
maakuor a big enough Cyc database, etc.21:51
maakulike i said, it might expose my biases, but I think the reasoning component itself might have quite a few different components to it which are not reduceable to one single principle21:51
maakuand due to past failures I tend to write off anybody who thinks they've discovered The One True Base Computation21:52
maakuat least not in a constructivist way21:53
maakuit's kinda like saying "I just discovered all computation can be reduced to a network of NAND gates. All we need to do is build trillions of NAND gates!"21:54
kanzurethis is a weird thing: http://diyhpl.us/~bryan/papers2/ai/Evolution%20of%20probabilistic%20consensus%20in%20digital%20organisms.pdf21:54
kanzurei am curious if you have a pre-baked rant about things like http://diyhpl.us/~bryan/papers2/ai/Understanding%20evolutionary%20potential%20in%20virtual%20CPU%20instruction%20set%20architectures.pdf21:55
kanzurethis seemed like an interesting direction but some of his implementation decisions were bizarre http://en.wikipedia.org/wiki/Tierra_(computer_simulation)21:59
fennmaaku isn't that true though21:59
kanzure"Tierra is a computer simulation developed by ecologist Thomas S. Ray in the early 1990s in which computer programs compete for central processing unit (CPU) time and access to main memory. In this context, the computer programs in Tierra are considered to be evolvable and can mutate, self-replicate and recombine. Tierra's virtual machine is written in C.[1] It operates on a custom instruction set designed to facilitate code changes and ...21:59
kanzure... reordering, including features such as jump to template[2] (as opposed to the relative or absolute jumps common to most instruction sets)."21:59
maakukanzure: rant? it looks like interesting work21:59
kanzure"The basic Tierra model has been used to experimentally explore in silico the basic processes of evolutionary and ecological dynamics. Processes such as the dynamics of punctuated equilibrium, host-parasite co-evolution and density-dependent natural selection are amenable to investigation within the Tierra framework. A notable difference between Tierra and more conventional models of evolutionary computation, such as genetic algorithms, ...22:00
kanzure... is that there is no explicit, or exogenous fitness function built into the model. Often in such models there is the notion of a function being "optimized"; in the case of Tierra, the fitness function is endogenous: there is simply survival and death."22:00
kanzuremaaku: sure, sorry, i classify many things as rants22:00
fennthats a terrible screenshot of tierra22:01
fennbetter screenshots http://web.stanford.edu/class/sts129/Alife/html/Tierra.htm22:02
-!- shubhamgoyal [~shubhamgo@119.56.115.169] has quit [Remote host closed the connection]22:02
fenna lot of the dynamics centered around jumping ahead in the memory map and building structures that benefitted you after the jump22:02
maakukanzure: so my planned architecture involves sort-of gp learning computational frames using MOSES : http://wiki.opencog.org/w/Meta-Optimizing_Semantic_Evolutionary_Search22:03
kanzurenone of the implementations of evolutionary computation look good to me22:04
maakuhaving an instruction set suitable to gp learning is an important detail. it'll be interesting to see what this group learned (I  saved the PDF for later review)22:04
kanzurenone of them seem to allow for the sorts of accidental shit that happens in the real world22:04
maakukanzure: yup. that's why evolutionary search is not the One Technique To Rule Them All22:04
kanzurewell, i mean, the evolutionary search implementers got lazy22:04
fennno it just means GA simulations are too low rez and their fitness functions suck22:05
kanzureehhhh22:05
kanzureprove it22:05
fennevolution happened22:05
fennwe are here now because of it22:05
fennhow is that not a silver bullet22:06
maakuMOSES is pretty good in that regard. The key takeaway from MOSES was that it needed to have a program description language that was specially crafted to have certain properties in order for the generated programs to be succinct ana analyzable22:06
maakufenn: so?22:06
maakuit's not even clear that evolution leads to intelligence22:07
fennok not necessarily "one technique to rule them all" but it does at least provide an existence proof of a silver bullet algorithm (however inefficient)22:07
maakuthere's a lot of anomolies around how homo sapiens evolved22:07
fennmaaku: are you doubting that you are in fact intelligent? or some sort of divine intervention? or what?22:07
maakuok scratch that. your misundstanding is simpler22:07
maakui'm not talking about the process for creating intelligence, i'm talking about how the intelligence works22:08
fennis there a difference?22:08
maaku"evolution" describes how the human brain came to be, not how it works22:08
maakuif i want to make a brain, "evolution!" is not an answer22:08
maakunot any more than "randomly assign transistor gates, check if it is intelligent, repeat."22:09
fennif your MOSES built an AI would you say it was MOSES22:09
fennevolution is not simply random though22:10
fennbah22:10
fennbruteforcing passwords is quicker with a markov chain than simply enumerating in binary22:11
fennlikewise, evolution finds quicker ways to adapt22:11
-!- HEx2 [~HEx@hexwab.plus.com] has joined ##hplusroadmap22:12
fennone of these adaptations is "intelligence"22:12
-!- HEx1 [~HEx@hexwab.plus.com] has quit [Read error: Connection reset by peer]22:12
fennintelligence adapted "artificial intelligence"22:12
fenni was going to say something about transposons and gene cassettes22:13
fennbut i think it's time for bed22:13
JayDuggerGood night.22:14
maakufenn: i'm not talking about using MOSES to build me an AI22:14
maakui'm talking about using MOSES as a component inside of an AI I design, for the purpose of learning new things in specific types of context22:14
maakugood night22:15
-!- yashgaroth [~ffffff@2606:6000:cb85:6a00:3da2:23d0:76b4:3e6] has quit [Quit: Leaving]22:18
-!- sheena [~home@173.181.61.137] has joined ##hplusroadmap22:34
-!- shubhamgoyal [~shubhamgo@119.56.117.248] has joined ##hplusroadmap22:46
-!- yottabit [uid36770@gateway/web/irccloud.com/x-pdxapmoviclejnpq] has quit [Quit: Connection closed for inactivity]22:48
-!- strages [sid11297@gateway/web/irccloud.com/x-yhufsnicqzxcvwvy] has quit [Ping timeout: 258 seconds]23:11
-!- strages [sid11297@gateway/web/irccloud.com/x-ocropcrzpiojrnvu] has joined ##hplusroadmap23:14
-!- shubhamg_ [~shubhamgo@119.56.117.243] has joined ##hplusroadmap23:14
-!- EnabrinTain [sid11525@gateway/web/irccloud.com/x-aepznyeobvxzvlhy] has quit [Ping timeout: 258 seconds]23:14
-!- EnabrinTain [sid11525@gateway/web/irccloud.com/x-erwzfmyqnaxvcgah] has joined ##hplusroadmap23:16
-!- shubhamgoyal [~shubhamgo@119.56.117.248] has quit [Ping timeout: 255 seconds]23:17
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has quit [Ping timeout: 258 seconds]23:18
-!- night [~Adifex@unaffiliated/adifex] has quit [Ping timeout: 258 seconds]23:20
-!- night [~Adifex@unaffiliated/adifex] has joined ##hplusroadmap23:22
-!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Excess Flood]23:24
-!- superkuh [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap23:25
-!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has joined ##hplusroadmap23:31
-!- shubhamg_ [~shubhamgo@119.56.117.243] has quit [Read error: Connection reset by peer]23:33
-!- akataphasia [~akataphas@108-203-53-229.lightspeed.sntcca.sbcglobal.net] has joined ##hplusroadmap23:37
-!- akataphasia [~akataphas@108-203-53-229.lightspeed.sntcca.sbcglobal.net] has left ##hplusroadmap []23:38
-!- HEx2 [~HEx@hexwab.plus.com] has quit [Ping timeout: 240 seconds]23:44
-!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap23:48
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 255 seconds]23:51
-!- Viper168_ [~Viper@unaffiliated/viper168] has joined ##hplusroadmap23:52

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