--- Day changed Fri Nov 28 2014
00:14 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
00:18 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 244 seconds]
00:49 -!- JayDugger [~jwdugger@pool-173-57-55-138.dllstx.fios.verizon.net] has quit [Quit: Leaving.]
01:00 -!- lichen [~lichen@c-50-139-11-6.hsd1.or.comcast.net] has joined ##hplusroadmap
01:14 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
01:18 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 258 seconds]
01:19 < archels> augur: thought of a similar scheme to downconvert the bat/rodent auditory range to human frequencies
01:20 < archels> this may or may not be related http://hackaday.com/2014/01/10/a-vibrating-timepiece/
02:02 -!- ParahSailin [~parahsail@unaffiliated/parahsailin] has quit [Ping timeout: 258 seconds]
02:03 -!- ParahSailin [~parahsail@unaffiliated/parahsailin] has joined ##hplusroadmap
02:15 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
02:18 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 255 seconds]
02:54 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap
03:15 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
03:17 -!- altersid is now known as hypersid
03:18 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 255 seconds]
03:28 -!- hypersid [~sid@altersid.net] has quit [Ping timeout: 250 seconds]
03:29 -!- hypersid [~sid@altersid.net] has joined ##hplusroadmap
03:56 < maaku> augur: minsky's 'frames' are magic boxes
03:57 < maaku> how frames are learnt, how they are integrated into context-aware connections, and how that connecdtion array changes over time is completely unspecified, and is also the real heart of the matter
04:15 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
04:19 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 255 seconds]
04:40 -!- eudoxia [~eudoxia@r190-134-51-44.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap
04:41 < chris_99> kanzure, you know you have a trinocular amscope, do you have any piccys from it per chance?
04:43 -!- Qfwfq [~WashIrvin@unaffiliated/washirving] has quit [Ping timeout: 255 seconds]
04:49 -!- Qfwfq [~WashIrvin@unaffiliated/washirving] has joined ##hplusroadmap
05:15 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
05:19 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 245 seconds]
05:25 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 244 seconds]
05:57 < kanzure> chris_99: http://diyhpl.us/~bryan/irc/microscope/IMG_20140620_193236.jpg
05:57 < kanzure> chris_99: http://diyhpl.us/~bryan/irc/microscope/IMG_20140621_135552.jpg
05:58 < chris_99> ooh wow cheers, so what adapter did you use for a camera, and what camera did you use?
05:58 < eudoxia> cool, is that brain tissue
06:03 -!- BobaMa [bobama@kapsi.fi] has joined ##hplusroadmap
06:04 < archels> probably not
06:05 < kanzure> no
06:05 < kanzure> also no adapter
06:05 < kanzure> camera was just some shitphone
06:05 < kanzure> i recommend an adapter
06:09 < chris_99> ah wow, did you take a photo through the twin lenses you normally look through then
06:10 < kanzure> i was using the trinocular port.
06:11 < chris_99> aha
06:21 -!- JayDugger [~jwdugger@pool-173-57-55-138.dllstx.fios.verizon.net] has joined ##hplusroadmap
06:59 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Ex-Chat]
07:06 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 240 seconds]
07:08 -!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has quit []
07:13 -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 255 seconds]
07:21 -!- Zinglon [~Zinglon@D549A77D.cm-10-1a.dynamic.ziggo.nl] has joined ##hplusroadmap
07:22 -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap
07:22 -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Excess Flood]
07:22 -!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has joined ##hplusroadmap
07:23 -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap
07:24 < kanzure> 21:28 < zooko> kanzure: my wife, ambimorph, wrote a tagging system this summer: https://github.com/ambimorph/protagonist
07:24 < kanzure> 05:44 < kanzure> "There is a subdirectory named ".protagonist/tags", and a subdirectory ".protagonist/tags/t" for every existing tag, t. Any file which is tagged with t is given a unique identifier and a hard link in the directory t."
07:24 < kanzure> 05:44 < kanzure> i dunno about this...
07:24 < kanzure> i don't really like the idea of using hardlinks like this
07:25 < kanzure> if you are going to be using filenames then you might as well be making a tagging file sytsem
07:25 < kanzure> er, tagging file system
07:26 -!- Merovoth [~Merovoth@gateway/tor-sasl/merovoth] has joined ##hplusroadmap
07:32 -!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 255 seconds]
07:34 < pasky_> kanzure: how would that fit with "A major design constraint of this project is to provide seamless compatibility with Tahoe-LAFS backup storage." ?
07:34 < sheena> ugh. i hate this thing. I posted about essential oils and now i hav eno way to know if anyone replied :(
07:35 < pasky_> though I think in general the better option is getting rid of anything that processes your directories and doesn't support symlinks
07:43 -!- yashgaroth [~ffffff@2606:6000:cb85:6a00:cdbf:9d7c:fcfd:b08d] has joined ##hplusroadmap
07:45 -!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap
07:48 < kanzure> sheena: nobody replied afaik
07:48 < kanzure> pasky_: so then you agree that syminks are the wrong thing to do there...?
07:57 < kanzure> http://f6fvy.free.fr/rtl_sdr/Some_Measurements_on_E4000_and_R820_Tuners.pdf
08:00 < kanzure> http://www.taylorkillian.com/2013/08/sdr-showdown-hackrf-vs-bladerf-vs-usrp.html
08:01 -!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
08:02 < kanzure> "50MHz –
08:02 < kanzure> er
08:02 < kanzure> "50 MHz - 6 GHz"
08:02 < kanzure> "It should be noted that the HackRF is not capable of full duplex communication unlike the other boards. This means that in order to switch from receive to transmit and vice versa, commands must be sent from the controller every time. This is only supposed to take microseconds when the decision to switch is made by the microcontroller, but in the case of complex signals being processed on the PC, it could take a lot longer to switch."
08:02 < pasky_> kanzure: they'd be less wrong than hardlinks... i didn't read the README all that carefully, but of all the wrong options, symlinks might be some of the least wrong ones
08:03 < pasky_> i think reiserfs was solving some of these things, iirc, but in the end noone cared enough
08:03 < pasky_> (i don't even know what's the usecase, who would like to tag their files and how)
08:03 < kanzure> use case is my collection of papers
08:04 < kanzure> instead of storing things with file names it would be nice to have tags
08:05 < kanzure> i am fine with using an additional application layer on top of a file system if necessary
08:15 -!- ebowden_ [~ebowden@147.69.179.205] has quit [Remote host closed the connection]
08:25 -!- drewbot [~cinch@ec2-54-197-106-143.compute-1.amazonaws.com] has quit [Remote host closed the connection]
08:26 -!- drewbot [~cinch@ec2-54-196-206-59.compute-1.amazonaws.com] has joined ##hplusroadmap
08:34 < JayDugger> sheena, check the logs.
08:34 < eudoxia> there are no logs after wednesday
08:35 < sheena> JayDugger: i dont know how :(
08:36 < kanzure> wide-band websdr http://websdr.ewi.utwente.nl:8901/
08:52 < kanzure> what was the sideways parabola i just saw
08:59 -!- yottabit [uid36770@gateway/web/irccloud.com/x-lcrydekpayeqaqbw] has joined ##hplusroadmap
09:00 -!- yottabit [uid36770@gateway/web/irccloud.com/x-lcrydekpayeqaqbw] has quit [Changing host]
09:00 -!- yottabit [uid36770@unaffiliated/ybit] has joined ##hplusroadmap
09:00 -!- yottabit [uid36770@unaffiliated/ybit] has quit [Changing host]
09:00 -!- yottabit [uid36770@gateway/web/irccloud.com/x-lcrydekpayeqaqbw] has joined ##hplusroadmap
09:08 -!- eudoxia [~eudoxia@r190-134-51-44.dialup.adsl.anteldata.net.uy] has quit [Quit: Lost terminal]
09:17 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap
09:48 < kanzure> http://www.sexviahex.com/ "Software EXploitation Via Hardware EXploitation" or "SExViaHEx" (as we jokingly refer to it) teaches you how to reverse engineer and exploit software on embedded systems via hardware. It teaches all this against real-world Commercial Off The Shelf (COTS) products such as routers, game systems, and other appliances.  This course has an intense focus on results oriented vulnerability discovery (not just hardware ...
09:48 < kanzure> ... hacking and tinkering for fun)." ($4k/week)
09:50 < chris_99> eek 'spensive
09:51 < chris_99> i assume you saw the weaponre post on HN?
09:51 < kanzure> yep http://weaponre.com/blog.html
09:52 < kanzure> not as expensive as singularity university
09:52 < chris_99> for $4 you could buy a decent microscope and learn how to decap yourself though
09:52 < chris_99> *4k
09:54 < yottabit> http://www.sexviahex.com/uploads/2/4/4/8/24485815/software_hardware_exploitation_training.pdf
09:54 < chris_99> oh ta, lets have a look what it entails
09:54 < chris_99> hmm still expensive imo
09:55 < fenn> augur: a few orders of magnitude higher hearing and you could triangulate radio transmissions
09:58 < kanzure> .wik bus pirate
09:58 < yoleaux> "The Bus Pirate is a universal electronic open hardware tool to program and interface with communication buses and program various chips, such as AVRs from Atmel and PICs from Microchip Technology." — http://en.wikipedia.org/wiki/Bus_pirate
10:02 < fenn> sheena: you could compare the "chemical resistance" ratings of your particular plastic against a chemical that's similar to your essential oil
10:02 < fenn> but it's complicated because you don't know which part of the molecule is reacting with the plastic
10:04 < fenn> the major reason would be that common plastics (PP and LDPE) are rather permeable to oils and will retain the smell and probably evaporate over time
10:05 < fenn> metal would be better if you're just worried about breakage
10:09 < kanzure> "gdb tricks you should know or be ashamed of not knowing" https://blogs.oracle.com/ksplice/entry/8_gdb_tricks_you_should
10:11 -!- Vutral [ss@mirbsd/special/Vutral] has quit [Ping timeout: 244 seconds]
10:11 < kanzure> fun fact: yottabit is probably ybit
10:12 < yottabit> yeah
10:12 < yottabit> fun fact: gnusha is still down
10:13 < kanzure> hmm maybe i should fix this
10:17 < yottabit> andytoshi: thanks for mentioning speed-reading, that has been on the todo list, and it's time to fix that
10:17 < fenn> sheena: otoh metal can catalyze oxidation and crosslinking reactions
10:18 < kanzure> i think sheena is busy doing car engine rebuild
10:18 < kanzure> under a meter of snow
10:18 < fenn> fun fun fun
10:18 < fenn> at least she doesn't have to worry if it doesn't run, since there's nowhere to drive
10:19 < fenn> does she have a dog sled?
10:19 < kanzure> http://images.drivebc.ca/bchighwaycam/pub/cameras/156.jpg
10:19 < kragen> yottabit: http://www.slate.com/articles/briefing/articles/2000/02/the_1000word_dash.html suggests that speed-reading is mostly bunk
10:20 < kanzure> yes well sort of, she has this http://3.bp.blogspot.com/-mlGojBNigIA/UunzkeGHboI/AAAAAAAAUW8/Th3aC3meLBM/s1600/999951_10151845555016526_623689869_n.jpg
10:20 < kragen> that is, the methods that people have claimed to achieve it so far are not effective
10:20 < kanzure> kragen: at minimum it seems to be safe to say that people do in fact read at different rates
10:21 < kanzure> "At Carver's direction, the 16 brainiacs read passages from Reader's Digest condensed editions under controlled conditions: None of them could read faster than 600 words per minute and retain more than 75 percent of the information contained in the texts." text matters though...
10:21 -!- Vutral [ss@mirbsd/special/Vutral] has joined ##hplusroadmap
10:21 < yottabit>  usually i am scanning and getting the gist of an article
10:21 < kragen> skimming can indeed be quite fast
10:22 < kragen> https://en.wikipedia.org/wiki/Speed_reading#Claims_of_speed_readers
10:22 < yottabit> i feel like this is faster than speed reading, but i do feel that these speed reading aids have their place if you need to consume a large amount of text quickly
10:22 < kanzure> i can read arbitrarily slow, so i would also posit that my other reading capabilities are faster.
10:23 < fenn> speed-reading would be great if someone also devised a method of speed-thinking
10:23 < kanzure> instantaneous thinking is not all that great. most of it gets forgotten.
10:24 < fenn> 90% of everything is crap
10:24 < kanzure> that's not what i mean
10:24 < kanzure> unless you mean "90% of everyone's ability to remember their instantaneous-style thoughts is crap"
10:25 < fenn> i mean, even if you forget 90% of it, it's okay as long as you remember the 10% that's worth remembering
10:25 -!- CheckDavid [uid14990@gateway/web/irccloud.com/x-nnucrjeoemeeawjo] has joined ##hplusroadmap
10:25 < kanzure> yeah i think it's way worse than that
10:25 < fenn> but realistically you'd forget 90% of the 10% that's important
10:25 < kanzure> on average i would guess that i am losing a lot of resolution per moment
10:26 < kanzure> way more than 90%
10:26 < fenn> because of noise?
10:26 < kanzure> no because working memory
10:26 < fenn> what do you mean resolution?
10:26 < kanzure> i have to dumb a lot of stuff down so that i can carry it forward to the next moment
10:35 < kragen> yeah, people in psychotic states seem to be thinking really, really fast
10:35 < kragen> in fact that's one of the distinguishing features of psychosis!
10:35 < yottabit>   curl $1 | unfluff | jq -r .text | speedread
10:35 < yottabit> done
10:35 < kragen> but they also have a really hard time holding onto a train of thought for long periods of time, such as 2 seconds
10:35 < yottabit> https://github.com/pasky/speedread
10:36 < kragen> like they can't remember what they were thinking and saying two seconds ago
10:36 < kanzure> they are not just trains of thoughts, they are terribly structured graphs
10:36 < fenn> so you're saying most people are psychotic
10:36 < kragen> fenn: fennetic.net is down? or am I smoking crack?
10:37 < fenn> yes it's down, i haven't figured out why yet
10:37 < kanzure> dns
10:37 < kragen> I saw an estimate once that the bandwidth of recording stuff in long-term memory is about half a bit per second
10:37 < kanzure> that's really unfortunate
10:37 < kanzure> damn
10:38 < kragen> that's why writing things down is useful
10:40 < fenn> plato ftw
10:40 < kanzure> why are they called trains of thought
10:40 < fenn> kragen is that "per second of sleep" or "per second of all day"
10:41 < yottabit> 1:37 PM <kragen> yeah, people in psychotic states seem to be thinking really, really fast
10:41 < yottabit> that's an interesting observation
10:41 < kanzure> "The term "train of thoughts" was introduced and elaborated as early as in 1651 by Thomas Hobbes in his Leviathan, though with a somewhat different meaning (similar to the meaning used by the British associationists)"
10:41 < yottabit> i know a crazy person, and this is true
10:41 < kanzure> "By Consequence, or train of thoughts, I understand that succession of one thought to another which is called, to distinguish it from discourse in words, mental discourse.
10:41 < kanzure> When a man thinketh on anything whatsoever, his next thought after is not altogether so casual as it seems to be. Not every thought to every thought succeeds indifferently."
10:41 < fenn> kanzure: because one thought leads to the next, like a train of cars
10:41 < kanzure> ugh
10:41 < kanzure> lost my space tether of thought
10:42 < kanzure> yeah something seems broken there
10:42 < yottabit> i feel that as as developer it can be beneficial to be thinking really, really fast
10:43 < yottabit> i.e. it's nice if you can multitask
10:43 < kanzure> if that is true then it would also be beneficial to have multiple developers at your command simultaneously
10:43 < fenn> multitasking is a lie
10:43 < kanzure> i think multitasking seems to work if you use other humans
10:43 < yottabit> fenn: in the sense that you have bitcoin compiling in one screen
10:43 < fenn> but then your task is just delegating other humans
10:43 < yottabit> you need to realize, oh i should fetch pip requirements in another screen
10:44 < fenn> by this definition "multitasking" is having toast in the toaster while you're cooking eggs
10:44 < fenn> instead of staring slack-jawed at the toaster watching toast brown
10:45 < kanzure> "Another factor contributing to the need for a minimal rehearsal as a routine precursor to action was the frequent indeterminacy or ambivalence of the circumstances in which one had to act. If the perceived circumstances seemed to call for doing X, but one couldn’t be absolutely sure, as was often the case, then it usually helped to do a minimal rehearsal for X, to attune oneself for doing X, for a while before actually doing it. By ...
10:45 < kanzure> ... readying one’s X-ing abilities for action, including the perceptual abilities required for X-ing, the minimal rehearsal energised and focussed one’s perceptual interrogation of the present situation. It made one alert to features of the situation relevant to X-ing – to features conducive to X-ing and features incompatible with X-ing. The minimal rehearsal sustained one’s X-relevant perceivings and one’s readiness to X while ...
10:45 < kanzure> ... the situation resolved itself one way or the other. It kept one motivated, and enabled a quick and full response when or if a situation fully conducive to X-ing did arrive."
10:45 < kanzure> i seem to have fallen into the weird part of the internet again
10:45 < kanzure> yes in general it seems good to minimize any unnecessary slack-jawing
10:45 < yottabit> multitasking is being concurrent, not necessarily performing tasks in parallel
10:47 < kanzure> which is task switching
10:47 < yottabit> that reads like the most idiotic thing i've said
10:47 < kanzure> shrug, fenn claims it is a lie, i claim task switching exists
10:49 < pasky_> yottabit: do you do that regularly?
10:49 < kanzure> "The actional detail of our ‘minimal rehearsal’ is difficult to specify. There seemed to be a trick to it. Possibly, this involved the actual commencing of the action being rehearsed – so that overt movement was incipient – quickly followed by the aborting of it. The knack was to get the commencing and the aborting optimally close together, so as to prime the action and render it incipient, without committing to any actual ...
10:49 < kanzure> ... movement. It involved a kind of ‘doing and not-doing’, a mere ‘making as if to’ do something. And the precision of the readying was important too. Part of the skill in minimal rehearsing was to ready just those muscles that would be involved in the anticipated action – and in the right combinations and orders – and no others. At any rate, the ability to rehearse an action in this special ‘minimal’ way seemed to come ...
10:49 < kanzure> ... naturally to us, and children acquired it early."
10:49 < pasky_> i gave up pretty quickly, myself :)
10:49 < yottabit> pasky_: depends on what i'm working on
10:50 < pasky_> yottabit: (eh, sorry; that == | speedread)
10:50 -!- pasky_ is now known as pasky
10:51 < yottabit> pasky_: oh :) it made sense when it came out, but i never used it.. i might start using it more with that function pasted above
10:56 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
10:57 < kragen> fenn: per second of all day
10:58 < kanzure> "Telling others to do things was closely related to telling others how to do things. Essentially, telling-how was just a more thorough and time-consuming version of telling-to."
10:58 < kragen> interesting use of "casual" in "his next thought after is not altogether so casual" — that meaning "coincidental" is not current in modern English, but it is in Spanish
10:58 < augur> fenn: thats not how radio works
10:59 < pasky> it is not current? i'm not a native speaker but it didn't strike me as odd
11:02 -!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap
11:04 < kanzure> (previous quotes from <http://www.derekmelser.org/essays/essaycognition.html>)
11:09 < yottabit> watched interstellar last night with a former cia agent who was active in the cold war era :)
11:09 < delinquentme> quantum compute: what are the machanimss by which it solves the protein folding problem?
11:09 < delinquentme> yottabit, I watched big hero 6 :P
11:09 < delinquentme> non stop tears dude.
11:09 < delinquentme> projecting all over the fucking place
11:12 < yottabit> it was refreshing to see a movie which tried to seem feasible while focusing on space exploration beyond our galaxy
11:12 < yottabit> there were many things which didn't make sense, but i'm glad something like that was made
11:12 < kanzure> so they have cars that don't have dust in them
11:13 < kanzure> but they can't figure out how to get dust out of their homes
11:13 < kanzure> nah it's cool just turn all the plates upside down
11:13 < yottabit> :)
11:13 < kanzure> get our giant walking ipods to do it
11:17 < fenn> ok, i defeated godaddy and succeeded despite their terrible interface and not remembering my username from 2007
11:18 -!- Vutral [ss@mirbsd/special/Vutral] has quit [Excess Flood]
11:18 < kanzure> is that what progress feels like?
11:19 < kanzure> yottabit: why didn't they take the whole family
11:19 < kanzure> seems like nasa was not in a position to argue
11:20 < yottabit> or if they had money to send several people to different planets, maybe they should have sent couples
11:20 < yottabit> and if they can't communicate from the new galaxy back home, how did his daughter know the other lady was still alive
11:20 < kanzure> osnap
11:20 < kanzure> nah it was something like 1-bit bandwidth
11:21 -!- Vutral [Xtc3cr2f6R@mirbsd/special/Vutral] has joined ##hplusroadmap
11:21 -!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 245 seconds]
11:22 < yottabit> they were receiving signals from back home, but they weren't able to send anything back to them
11:22 < kanzure> they could send a single bit apparently
11:23 < kanzure> so they could have used delays between their bits
11:32 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 258 seconds]
11:35 -!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
11:35 -!- snuffeluffegus [~snuff@ps357888.dreamhost.com] has joined ##hplusroadmap
11:37 < yottabit> http://www.infoq.com/news/2014/08/the-future-of-docker
11:37 < yottabit> looks like coinbase is using http://flynn.io
11:37 < yottabit> https://github.com/flynn/flynn
11:38 < kanzure> i still prefer fig
11:38 < yottabit> https://github.com/progrium/registrator
11:38 < kanzure> especially since docker acquired orchard
11:38 < kanzure> and especially since docker people are merging fig into docker
11:38 < kanzure> i haven't figured out why registrator is necessary if you're already using consul
11:39 < yottabit> i really like the idea of coreos, i don't like that fleet is so coupled with etcd
11:49 < fenn> heh before: http://www2003.org/cdrom/papers/refereed/p583/p583-gupta_files/image022.jpg and after: http://www2003.org/cdrom/papers/refereed/p583/p583-gupta_files/image024.jpg
11:50 < fenn> "content extraction algorithms"
11:52 < yottabit> https://dl.acm.org/citation.cfm?id=775182
11:52 < yottabit> .title
11:52 < yoleaux> DOM-based content extraction of HTML documents
11:52 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Ex-Chat]
11:53 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap
11:54 < fenn> yottabit: it's what your "unfluff" is based on
11:54 < yottabit> nice
11:54 < fenn> unfluff doesn't seem to be anywhere near as aggressive tho
11:56 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
12:00 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 255 seconds]
12:00 < kanzure> is there a name for the idea of "thinking originated in evolutionary history from vocalization, speaking, speech into mumbling, and then into silent speaking"?
12:04 < fenn> is this an idea that people hold to be true?
12:05 < fenn> danny hillis has some theories about mating calls and songs turning into language...
12:05 < kanzure> this person seems to http://www.derekmelser.org/essays/essaycognition.html
12:05 < kanzure> well, i am trying to skip the language part
12:05 < kanzure> monkey screams happen in the absence of a (at least human-discernable) language
12:06  * fenn looks at http://longnow.org/essays/intelligence-emergent-behavior-or-songs-eden/
12:06 < fenn> you seem to be quibbling over the definition of language
12:07 < kanzure> i didn't think this would have anything at all to do with language and i'm confused why anyone is bringing it up (even nsh)
12:07 < fenn> "It is taken as given that a genuine language is an abstract structure -- of representations, meanings, logical operators and combinatorial laws" wow that's a huge assumption
12:07 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Ping timeout: 240 seconds]
12:08 < kanzure> i emplore you to not get lost in the language bullshit
12:08 -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has quit [Ping timeout: 255 seconds]
12:11 < kanzure> i believe it is fair to assume that at one point whatever ancestors we had, if you go back far enough, did not have a thing approximating language
12:12 < fenn> i scrolled back a few pages and still don't know what you're actually after
12:13 < kanzure> oh, a handful of competing observations:
12:13 -!- Vutral [Xtc3cr2f6R@mirbsd/special/Vutral] has quit [Ping timeout: 272 seconds]
12:14 < kanzure> 1) rehersal is an interesting concept that was lacking from that mental inertia article (probably something about "a lack of rehersal"), but this observation is just coincidental i think
12:14 < kanzure> or rather, er, i don't mean to say rehersal specifically. the related notions from the page.
12:14 < fenn> "almost-rehearsal"
12:15 < kanzure> "mental action task caching and front-loading stuff, and loading approximations"
12:15 < kanzure> 2) i suspect there should be or is a gap in behavior from howling monkey clans and pre-language human tinkerers
12:15 < kanzure> which is relevant to ai things
12:16 < fenn> why 2
12:16 < kanzure> 3) something about multitasking or concurrent tasking and task switching made me think about the rehersal thing
12:17 < kanzure> well, 2 is interesting because presumably pre-language brain biology may have been simpler, i don't know
12:17 < yottabit> coreos is a minimal linux OS
12:17 < yottabit> did they build it from scratch or is it based on some other distros?...
12:18  * yottabit asks in #coreos
12:18 < kanzure> what's with the frequent focus of agi projects on human-level intelligence? why not dog-level...
12:18 < fenn> i state that there is no difference between howling monkey clans and pre-language humans
12:18 < fenn> if there even is such a thing as "pre-language"
12:18 < fenn> monkeys obviously have different calls for different threats/contexts
12:19 < fenn> and they use tools
12:19 < fenn> but even birds do that
12:19 < kanzure> i'll allow for simple grunts and sounds to be called language if you insist
12:19 < fenn> you can communicate important information with grunts
12:19 < kanzure> you can communicate anything with single bits
12:20 < fenn> a grunt is not a bit, but whatever
12:20 < kanzure> transport channel is really not interesting to me at the moment
12:20 < kanzure> or communication for that matter
12:20 < fenn> .wik pulse position modulation
12:20 < yoleaux> "Pulse-position modulation (PPM) is a form of signal modulation in which M message bits are encoded by transmitting a single pulse in one of  possible time-shifts. This is repeated every T seconds, such that the transmitted bit rate is  bits per second." — http://en.wikipedia.org/wiki/Pulse_position_modulation
12:21 < fenn> stupid math notation
12:21 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap
12:21 < fenn> one of 2^M possible time-shifts; bit-rate is M/T
12:23 < fenn> because humans evolved as social animals, communication is tightly integrated into our thought process
12:25 < fenn> minsky would say some bullshit like "humans are intelligent because the separate agents in their society of mind can communicate"
12:26 < kanzure> yawn
12:29 < kanzure> i don't know why that page started off with that assumption, because it seems unrelated to any of the other content
12:30 -!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap
12:39 < fenn> i don't understand this at all "The paper is in the form of a first-hand eye-witness account (by a survivor of the Lower Palaeolithic)"
12:40 -!- Vutral [ss@mirbsd/special/Vutral] has joined ##hplusroadmap
12:40 < kanzure> well it's certainly weird
12:40 < chris_99> How hard do you guys think it will be to create a woolly mammoth from this latest one they've found, out of interest
12:41 < kanzure> isn't this the one that requires using an elephant as a surrogate?
12:41 < kanzure> so, at minimum whatever resources are required to care for an elephant
12:41 < fenn> they can probably make a retarded broken mammoth that will die after a few months
12:41 < chris_99> i think that was the plan to use an elephant in someway
12:42 < chris_99> awh don't say that fenn, it'd be cool to have a herd of them marching through siberia
12:43 -!- snuffeluffegus [~snuff@ps357888.dreamhost.com] has quit [Remote host closed the connection]
12:44 < fenn> it's hard to find an elephant surrogate that's ovulating and you have permission to do science experiments on
12:44 < chris_99> i'm a bit confused what they do, so they've got the blood of the mammoth then what?
12:45 < kanzure> probably something like somatic cell nuclear transfer
12:45 < kanzure> but that does not guarantee anything
12:45 < kanzure> incubation might require other hormones and signals
12:45 < fenn> they extract a nucleus and implant it in a denucleated elephant egg cell, then poke it with electricity or something
12:46 < chris_99> just looking on the wiki page for that sounds interesting
12:46 < fenn> blood doesn't have much dna so they will probably use bone marrow or muscle tissue
12:46 < kanzure> you can read about it in my new book, "1001 odd transhumanist tricks"
12:46 < chris_99> haha
12:47 < chris_99> so this principle only works with animals that are vaguely related
12:47 < kanzure> hardly works at all
12:47 < chris_99> oh heh
12:47 < kanzure> to make an artificial womb that works you really really need to study existing wombs and pregnancies
12:48 < fenn> yeah it's hard to do even with the same species
12:48 < kanzure> and to convert an existing womb into something compatible with another dead species... ehhh
12:48 -!- maaku [~quassel@50-0-37-37.dsl.static.fusionbroadband.com] has quit [Read error: Connection reset by peer]
12:48 < fenn> the placenta is the least studied organ
12:48 < chris_99> is this the only method thats used to do this type of cloning?
12:48 < kanzure> "cloning" is easy man
12:50 < kanzure> well, you really need to get the dna into cells
12:50 < kanzure> nuclear transfer or nuclear injection seems sorta important for doing that
12:50 < chris_99> mmm true
12:50 < fenn> sheesh half the results for "placenta research" are about the health benefits of eating placenta
12:50 -!- Lemminkainen [uid2346@gateway/web/irccloud.com/x-bwpumoubtwkdzvwh] has joined ##hplusroadmap
12:50 < chris_99> eeek
12:50 < superkuh> I did some placenta tissue culture during my undergrad.
12:50 < kanzure> you might have more benefit taking an elephant genome and inserting mammoth-related junk instead
12:51 < kanzure> because you already know the elephant pregnancy thing works
12:51 < chris_99> "Dolly the sheep was born after 277 eggs were used for SCNT, which created 29 viable embryos"
12:51 < chris_99> ah yeah
12:51 < superkuh> Placental syncytium has a fascinating cytoskeleton.
12:52 < Lemminkainen> pix or gtfo superkuh I wanna see this cytoskeleton
12:52 < fenn> yeah it seems easier to sequence the mammoth genome and figure out what makes it different from an elephant
12:53 < fenn> it's probably the worst possible organism to try to do genetic engineering on
12:53 < chris_99> why's that?
12:54 < fenn> extremely expensive, long gestation time, long life cycle, ethics issues, animal rights issues, and it's a eukaryote...
12:54 < fenn> they should try to resurrect the extinct siberian tulip instead :P
12:55 < Lemminkainen> fenn orcas, blue whales, panda bears
12:55 < chris_99> haha, who wants a siberian tulip
12:55 < fenn> they should just try to breed a bigger lemur instead of all this panda junk
12:56 < Lemminkainen> welcome to the post-postmodern zoo
12:56 < fenn> https://en.wikipedia.org/wiki/File:Archaeoindris_fontoynonti.jpg needs optimizing for cuteness
12:57 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
12:57 < fenn> better than this guy tho https://en.wikipedia.org/wiki/File:Megaladapis.jpg
12:59 < fenn> Lemminkainen: nobody's trying to resurrect extinct orcas, pandas, or blue whales (yet)
13:00 < kanzure> i wonder if you could put a researcher into a whale placenta during gestation
13:00 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 240 seconds]
13:00 < chris_99> haha
13:01 < kanzure> not enough room for food storage
13:01 < Lemminkainen> fenn I was just naming worse organisms to try to do genetic engineering on
13:02 < fenn> a tasmanian tiger or a dodo would be a lot easier to start with
13:03 < fenn> passenger pigeon etc
13:03 < Lemminkainen> from the more general perspective, I don't understand the value of resurrecting a dead species without first stopping making more of them
13:04 < fenn> it's a PR stunt for biotechnology companies
13:04 < kanzure> you could probably do it for dna recovered from human skeletal remains
13:05 < fenn> that would be too easy to fake
13:05 < fenn> also laws against human cloning
13:05 < kanzure> "experience the unlimited joy of raising <some dead historical figure everyone hates"
13:05 < kanzure> +
13:05 < kanzure> +>
13:06 < kanzure> i guess there was a reason > didn't take the first time
13:06 < kanzure> as evidenced by the second time
13:10 -!- kenju254 [~kenju254@static-41-242-0-196.ips.angani.co] has quit [Ping timeout: 258 seconds]
13:12 < Lemminkainen> but that's PR of such limited value
13:12 < Lemminkainen> wouldn't it be better to, I don't know, reverse someone Alzheimer's?
13:13 < Lemminkainen> have we really decided that cloning mammoths is a more tractable problem than curing neurodegenerative disorders?
13:13 < fenn> it's easier to be optimistic about something nobody understands
13:13 < fenn> (nobody understands why cloning works)
13:14 -!- kenju254 [~kenju254@static-41-242-0-196.ips.angani.co] has joined ##hplusroadmap
13:14 < fenn> also people have been relatively succesful bumping around in the dark trying to clone stuff
13:14 < fenn> whereas alzheimers has lots of effort for not much success
13:16 < fenn> also there are a lot of cloning companies in korea that don't really know why they exist
13:16 < fenn> "clone your pet dog for $100,000"
13:21 < fenn> "It really helps that every time [Korean scientists] give a talk, they don't have to have an argument about whether an embryo is a person."
13:21 < Lemminkainen> the people who did the US attempt at commercial dog cloning live up in Mill Valley
13:27 < fenn> apparently Hwang Woo-suk is some kind of cloning rock-star "The national law-enforcement agency assigns officers to protect him. Korean Airlines flies him around the world for free. The minister of science and technology ranks at the top of the South Korean Cabinet—as high as the secretary of state or treasury in the United States."
13:27 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 245 seconds]
13:27 < fenn> er, that last bit was because "Science is trendy in Korea"
13:28 < Lemminkainen> ah
13:28 < Lemminkainen> let's get Shinya Yamanaka similar accolades
13:28 < Lemminkainen> I want to put his tesselated face on a hoodie
13:30 < fenn> "Yamanaka received ... the Millennium Technology Prize in 2012 together with Linus Torvalds." uh, okay
13:36 < kanzure> yamanaka should be a name you already know
13:36 < kanzure> although receiving an award next to linus torvalds is amusing :)
13:37 < fenn> i'm bad with japanese names actually
13:37 < kanzure> satoshi hashimoto was a total phony
13:37 < fenn> yeah fuck that guy
13:38 -!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
13:40 < fenn> he's nowhere as cool as norio wakamoto
13:40 < kanzure> or tajiri nihei
13:45 < kanzure> google translate says "wéidài" and not "wei dai"
13:46 < fenn> .title https://www.youtube.com/watch?v=UucITQR0SrQ
13:46 < yoleaux> Azumanga Funny Scene - YouTube
13:46 < fenn> norio wakamoto is the fake cat thing
13:47 < kanzure> yamanaka dominates this folder basically http://diyhpl.us/~bryan/papers2/stem-cells/
13:48 < kanzure> eg http://diyhpl.us/~bryan/papers2/stem-cells/Induction%20of%20pluripotent%20stem%20cells%20from%20adult%20human%20fibroblasts%20by%20defined%20factors.pdf
13:48 < kanzure> "kazutoshi" hah
13:56 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
13:57 < fenn> oh he was Cell in DBZ
13:58 < kanzure> who?
13:59 < fenn> norio wakamoto
13:59 < kanzure> "Your search - filetype:pdf - did not match any articles published in Frontiers in Cellular Neuroscience."
14:00 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 240 seconds]
14:01 < kanzure> "Journal of Really Fancy Neuron Drawings"
14:01 < kanzure> .title http://journal.frontiersin.org/Journal/10.3389/fncel.2014.00145/full
14:01 < yoleaux> Frontiers | Patch-clamp recordings of rat neurons from acute brain slices of the somatosensory cortex during magnetic stimulation | Frontiers in Cellular Neuroscience
14:04 < kanzure> "Surprisingly, while TMS has been commercially available for decades, the actions of single pulse magnetic stimulation at the cellular level have not been directly studied. Some studies have suggested that that TMS activates cortical neurons antidromically, primarily at axonal bends, bifurcations, or terminations (Amassian et al., 1992; Maccabee et al., 1993, 1998; Kamitani, 2001; Hallett, 2007). Other investigations have claimed, mostly ...
14:04 < kanzure> ... by recording spinal volleys, that the action potential is generated more proximal to the soma (Edgley et al., 1990; Baker et al., 1995; Nielsen et al., 1995; Di Lazzaro et al., 2002; Terao and Ugawa, 2002; Pasley et al., 2009). Distal axonal activation evokes indistinguishable forward and backward information flow in the cortical network, suggesting that TMS provides a nonspecific reset signal (Walsh and Pascual-Leone, 2003)."
14:05 < kanzure> "In contrast, action potential initiation at the axon's initial segment elicits the normal, forward information flow in the cortical network. We recently investigated the effects of magnetic stimulation on single neurons using compartmental modeling (Pashut et al., 2011). Contrary to published models (Roth and Basser, 1990; Basser and Roth, 1991; Basser et al., 1992; Nagarajan et al., 1993; Abdeen and Stuchly, 1994; Roth, 1994; Ravazzani ...
14:05 < kanzure> ... et al., 1996; Ruohonen et al., 1996a; Davey and Epstein, 2000; Hsu and Durand, 2000; Kamitani, 2001; Hsu et al., 2003; Rotem and Moses, 2006; Silva et al., 2008; Salvador et al., 2011) our simulations predicted that TMS affects neurons in the central nervous system by somatic depolarization leading to initiation of actions potentials in the axon's initial segment (Pashut et al., 2011)."
14:05 < kanzure> "Driven by our theoretical predictions, we combined, for the first time, a patch-clamp setup designed for brain slice recordings with a custom-made magnetic coil. Using this novel setup magnetic stimulation was applied to acute brain slices and the response of cortical neurons recorded. Our recordings supported our theoretical prediction that the action potential was generated at the initial segment of the axon following somatic ...
14:05 < kanzure> ... depolarization during magnetic stimulation. Interneurons and pyramidal neurons responded differently to magnetic stimulation. We show, both experimentally and computationally, that the magnetic threshold of central nervous system neurons is correlated with the size of the soma, the current threshold of the neuron, and the orientation of the magnetic coil. In combination with our previous compartmental model, the current study ...
14:05 < kanzure> ... suggests a cellular mechanism for TMS."
14:06 < kanzure> welp.. okay then.
14:10 < kanzure> oh, correlated
14:10 < kanzure> "According to our computational prediction, magnetic stimulation induces the largest depolarization in the soma followed by action potential initiation at the axon's initial segment (Pashut et al., 2011). Proving this prediction requires simultaneous recording from the axon's initial segment and the soma. This experiment cannot be performed due to the limitation of our recording setup."
14:10 < kanzure> really? i thought everyone had multi-patch-clamp setups these days.
14:18 < kanzure> .wik patch clamp
14:18 < yoleaux> "The patch clamp technique is a laboratory technique in electrophysiology that allows the study of single or multiple ion channels in cells. The technique can be applied to a wide variety of cells, but is especially useful in the study of excitable cells such as neurons, cardiomyocytes, muscle fibers and pancreatic beta cells." — http://en.wikipedia.org/wiki/Patch_clamp
14:19 < kanzure> http://upload.wikimedia.org/wikipedia/commons/4/41/WholeCellPatchClamp-03.jpg
14:19 < kanzure> how is that a neuron?
14:20 < kanzure> ooh, "automated patch clamp"
14:21 < kanzure> "Schematic of a patch clamp system using a droplet suspension culture and gravity to position the cells above the pipette. Suction inside the pipette draws the cells to the tip of the pipette which then forms the gigaseal." (from <http://en.wikipedia.org/wiki/Automated_patch_clamp>)
14:24 < kanzure> huh they have used patch-clamp on spermatozoa to figure out sperm ion channels http://www.researchgate.net/publication/51192906_Rediscovering_sperm_ion_channels_with_the_patch-clamp_technique/file/e0b4952533505a6350.pdf
14:26 -!- pete4242 [~smuxi@boole.london.hackspace.org.uk] has quit [Remote host closed the connection]
14:35 < kanzure> "Vertical nanowire electrode arrays as a scalable platform for intracellular interfacing to neuronal circuits" http://www.people.fas.harvard.edu/~parklab/publications/nnano.2011.249.pdf
14:35 < kanzure> "we report a scalable intracellular electrode platform based on vertical nanowires that allows parallel electrical interfacing to multiple mammalian neurons. Specifically, we show that our vertical nanowire electrode arrays can intracellularly record and stimulate neuronal activity in dissociated cultures of rat cortical neurons and can also be used to map multiple individual synaptic connections."
14:42 < fenn> why do they use cultures instead of brain slices?
14:42 < kanzure> "Electromagnetic limits to radiofrequency (RF) neuronal telemetry" https://www.scienceopen.com/document_file/75a51973-8317-40dc-a50a-d97bdd4244a4/PubMedCentral/75a51973-8317-40dc-a50a-d97bdd4244a4.pdf
14:42 < kanzure> i don't know why they are assuming a receiver 1 meter away from the head. that's silly.
14:42 < kanzure> i don't know if you can keep brain slices alive long enough
14:50 < kanzure> .tw https://twitter.com/zooko/status/538392148127657984
14:50 < yoleaux> @petertoddbtc Challenge: key generation alg with entirely analog, homebuilt components: marbles, blocks, paper, rubber bands, etc. (@zooko, in reply to tw:538391798763118592)
14:50 < kanzure> .tw 538391798763118592
14:50 < yoleaux> The term "paper wallet" is incredibly misleading: your key was still generated digitally, and may very well have been compromised at birth. (@petertoddbtc)
14:55 < fenn> "Barbie's got the right idea: in cyberwarfare the only thing you can really trust with a secret is hardware you've audited yourself, like pen and paper."
14:56 < fenn> but what if they've got a conduction microphone attached to the floor that the table you're writing on is sitting on
14:57 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
14:58 < chris_99> or CCTV
14:58 < fenn> i wonder how much paper generating a keypair would require
14:58 < fenn> assuming you could do it without messing up
14:59 < kanzure> "Free-standing mechanical and photonic nanostructures in single-crystal diamond" http://nano-optics.seas.harvard.edu/publications/Mike_freestanding.pdf
14:59 < kanzure> "anisotropic plasma etching at an oblique angle" 
15:00 < kanzure> oh i can't tell if this is just electron beam lithography
15:01 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 272 seconds]
15:02 < fenn> it's plasma etching with mask, not electron beam
15:02 < kanzure> doh
15:05 < kanzure> heh here is a thing someone bothered to do: "Cell responses to metallic nanostructure arrays with complex geometries" http://www.sciencedirect.com/science/article/pii/S014296121400831X
15:05 < kanzure> "anopillar arrays of various shape, size, and spacing and different nanopillar-substrate interfacial strengths were fabricated and interfaced with fibroblasts and several unique cell-nanopillar interactions were observed using high resolution scanning electron microscopy. Nanopillar penetration, engulfment, tilting, lift off and membrane thinning, were observed by manipulating nanopillar material, size, shape and spacing."
15:05 < kanzure> *nanopillar
15:05 < nsh> paperbot?!
15:05 < nsh> .title http://link.springer.com/chapter/10.1007%2F978-3-642-21969-6_11
15:05 < yoleaux> Memory-Constrained Implementations of Elliptic Curve Cryptography in Co-Z Coordinate Representation - Springer
15:06 < kanzure> no access, try libgen
15:08 < kanzure> "Whole-cell patch-clamp recordings in freely moving animals" huh, that is neat
15:09 -!- HEx2 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
15:10 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Read error: Connection reset by peer]
15:11 < fenn> yeah the nanostructure cell scaffolding and stiffness matching looks promising for brain tissue culture and long-term implant viability
15:13 < kanzure> .title https://www.youtube.com/watch?v=nJpFvQ0YZLk
15:13 < yoleaux> Ultra high Resolution 3D Human Brain Model (BigBrain) in Atelier3D Viewer - YouTube
15:15 < kanzure> ftp://bigbrain.loris.ca/faq.txt
15:16 < kanzure> "The bigbrain is the brain of a 65yo woman with no neurological or psychiatric diseases in clinical records at time of death. The brain was embedded in parafin and sectionned in 7404 coronal histological sections (20 microns), stained for cell bodies. The bigbrain is the digitized reconstruction of the hi-res histological sections (20 microns isotropic)."
15:16 < kanzure> finally some real science
15:17 < kanzure> related tissue slice viewer (but doesn't have that same dataset?) http://www.tissuestack.org/
15:18 < kanzure> "The bigbrain videos were created using Atelier3D, a licensed software which is currently not distributed. The volume read in Atelier3D is at 20-micron isotropic, which is too big for file transfers. This is why reduced volumes at 100, 200, 300, 400 microns have been created."
15:18 < kanzure> aaaaa so much is wrong with that
15:18 < fenn> what are the artifacts in the last few seconds?
15:19 < kanzure> here is the data ftp://bigbrain.loris.ca/volumes/
15:19 < fenn> looks like bubbles almost
15:20 < kanzure> damage during parafin fixation?
15:20 < kanzure> or damage during cutting
15:21 < fenn> maybe they injected dye with a syringe 50 times?
15:21 < kanzure> there is a diagrma of their slicer here http://diyhpl.us/~bryan/papers2/neuro/BigBrain:%20An%20Ultrahigh-Resolution%203D%20Human%20Brain%20Model.pdf
15:22 < kanzure> *diagram
15:22 < fenn> "too big for file transfers" is bunk
15:22 < kanzure> of course it's bunk
15:22 < kanzure> "and we are also using super-proprietary software to show off IN THREE DIMENSIONS an animation of stepping through the slices of data"
15:23 < kanzure> i should go taunt some neuroscientists and dare them to attempt to correctly label everything
15:23 < kanzure> pssst archels
15:23 < fenn> the three dimension part is harder than it sounds, because it means the slices have to be distortion-corrected and aligned to 20 micron ish
15:24 -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 250 seconds]
15:24 < fenn> displaying it is relatively straightforward
15:25 -!- Zinglon [~Zinglon@D549A77D.cm-10-1a.dynamic.ziggo.nl] has quit [Quit:  HydraIRC -> http://www.hydrairc.com <- Now with extra fish!]
15:27 < fenn> 100um is 3GB so 20um would be 375GB
15:28 < fenn> but MS-DOS doesn't like it!!!
15:29 < kanzure> check out page 3 figure 1a http://www.unicog.org/publications/1-s2.0-S105381191400487X-main.pdf
15:29 < fenn> do i have to
15:29 < kanzure> essentially, different neuroanatomy atlases are inconsistent with each other
15:30 < kanzure> "surprise! when you have bad data sets and bad labeling you will end up with disagreeing results"
15:30 < fenn> wow
15:31 -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap
15:33 < kanzure> "bah, just use machine learning to fix poorly labeled atlases!" http://www.sciencedirect.com/science/article/pii/S1053811914009483
15:33 < kanzure> "We use the proposed method to combine four brain MRI datasets labeled with different protocols (with a total of 102 unique labeled structures) to produce segmentations of 148 brain regions. Using cross-validation, we show that the proposed algorithm outperforms the generalizations of majority voting, semi-locally weighted voting and STAPLE (mean Dice score 83%, vs. 77%, 80% and 79%, respectively)"
15:33 < kanzure> seems a little silly
15:33 < kanzure> would probably be better to let the machine classify regions on its own
15:33 < kanzure> rather than based on previous human classifications
15:34 < fenn> but then all the accumulated bad results would go away
15:36 < fenn> so if i'm reading this right, each slice is a 200MB image? srsly?
15:37 < kanzure> what's wrong with 200 megabyte slices?
15:37 < fenn> they'd do much better to just use jpg compression at the original resolution, rather than intentionally losing resolution
15:37 < fenn> so i can only speculate as to their motives
15:37 < kanzure> is it possible that they are bad at computers
15:38 < fenn> intentionally created plausible deniability
15:38 < fenn> "we can't give out our data because we're too stupid, i swear"
15:38 < kanzure> "what the fuck's a terrorbyte?"
15:48 < fenn> oops i meant 50MB not 200MB
15:48 < fenn> that's more reasonably
15:48 < fenn> i mean i was looking at the resolution/number of slices for 100um voxels
15:49 < fenn> about a terapixel
15:49 < fenn> teravoxel
15:50 < kanzure> hmm i should send an email to ed boyden about my neuroanatomy complaints
15:51 < fenn> so has anyone done CLARITY on human brains yet?
15:52 < fenn> i know they've done slices
15:53 < kanzure> russell hanson gave me lip for even mentioning that to him
15:53 < kanzure> "rawr you're a moron don't you know that's destructive fuck you"
15:53 < fenn> huh? they're just going in the garbage anyway
15:54 < fenn> i'm having trouble determining the intended meaning behind that statement
15:54 < kanzure> oh i wonder if the "make a 2d map of a 3d structure" stuff would make sense for brain mapping/region labelling/stuff.
15:55 < kanzure> the intended meaning was that someone yelled at me, that's al
15:55 < kanzure> *all
15:55 < kanzure> presumably he was upset and yelling because he wants to focus on non-destructive methods http://brainbackups.com/
15:56 -!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 265 seconds]
15:57 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
15:57 < fenn> there is literally zero technical information on that page
15:58 < fenn> "we are actively developing new methods, yay!"
15:58 < kanzure> maybe their video has more info
15:58 < kanzure> https://www.youtube.com/watch?v=IOfq376Uxok&t=1m
16:00 < kanzure> er... nope.
16:00 < kanzure> "functionalized dna aptamer nanoparticles, smart contrast agents for mri and x-ray imaging, dna barcoding sequencing to map neuronal connections"
16:00 < fenn> i'm highly skeptical that X-ray imaging can produce a connectome and not destroy the brain in the process
16:00 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 240 seconds]
16:01 < kanzure> he said x-ray plus contrast agents
16:01 < fenn> oh in that case the singularity must be here
16:01 < fenn> think about it, contrast agents means you have to shoot more x-rays
16:01 < fenn> more x-rays means more damage
16:01 < kanzure> i have never really heard anyone say "dna aptamer nanoparticles"..
16:02 < kanzure> i guess they are technically smaller than most antibodies
16:02 < fenn> presumably they are attaching nanoparticles to aptamers for use as contrast agents
16:02 < kanzure> i would expect them to be using aptamers to bind to particular targets in the brain itself
16:02 < kanzure> maybe also carrying some payload, sure
16:02 < fenn> anyway how the hell are you supposed to focus on nano-anything from several inches away
16:03 < kanzure> with science and love
16:03 -!- juri_ [~juri@vpn166.sdf.org] has joined ##hplusroadmap
16:03 < fenn> yes the aptamers bind to targets in the brain, and they carry covalently-linked heavy metal nanoparticles so something shows up on the x-ray
16:04 < fenn> but, but, but... where's a buttbot when you need one
16:04 < kanzure> i was really expecting positron emission tomography stuff
16:04 < kanzure> maybe it takes too long to incorporate into an aptamer
16:04 < fenn> the resolution isn't good enough
16:05 < fenn> you have to detect ONE photon
16:05 < fenn> detectors with a gain that high are large, which limits the pixel count, which limits the resolution
16:05 < fenn> or you could have a gigantic megastructure that's perfectly shielded from stray radiation
16:06 < fenn> actually you have to detect two photons, but it has to be the right two photons
16:06 < kanzure> hmph
16:07 < kanzure> you don't need every photon do you?
16:07 < kanzure> like even at a low sampling rate..
16:07 < kanzure> "Cortical glucose metabolic rate correlates of abstract reasoning and attention studied with positron emission tomography"
16:07 < fenn> you need to be able to find the two photons that came from the same decay event in order to measure where that event was
16:07 < kanzure> http://upload.wikimedia.org/wikipedia/commons/2/20/PET_Normal_brain.jpg
16:08 < fenn> i assume we're talking about synapse scale resolution here
16:08 < fenn> if you want to make pretty colored pictures, that's totally doable
16:09 < kanzure> http://crypto.stackexchange.com/questions/8747/how-to-build-an-electro-mechanical-public-key-cipher-machine
16:09 < kanzure> cc kragen
16:09 < fenn> whole-brain CT with contast agents would definitely help in aligning slices
16:10 < kanzure> i don't know if slice alignment is a serious problem at the moment
16:10 < kanzure> todd never complained about slice alignment to me
16:10 < fenn> it's a problem
16:10 < fenn> also todd is working on ~5mm cubes
16:10 < kanzure> seems like less of a problem than sub-synapse resolution
16:11 < fenn> but you gotta align the whole brain at sub-synapse resolution
16:11 < kanzure> btw what's the back story about CT not explicitly saying x-ray in its name
16:11 < fenn> well people hate hospitals and they love CATs so they called it a CAT scan
16:12 < fenn> and PET scans for dog people
16:12 < kanzure> shit now i'm going to remember this
16:13 < fenn> what's the back-story about MRI not having "nuclear" in its name
16:13 < kanzure> dunno, because it's often pronounced "nuclear magnetic resonance imaging" for some reason
16:13 < kanzure> although not as often as without nuclear
16:14 < fenn> MRI is just NMR with a magnetic field gradient and fancy microwave imaging sensors
16:14 < fenn> they dropped the "nuclear" so patients wouldn't be scared of it
16:15 < fenn> it's bad enough going into a tube with angry machine noise and having to stay perfectly still
16:15 < fenn> no need to add cold-war apocalyptic hysteria and irrational fears about contamination and cancer
16:16 < kanzure> "yep into the cancer machine you go"
16:16 < kanzure> "One study estimated that as many as 0.4% of current cancers in the United States are due to CTs performed in the past and that this may increase to as high as 1.5 to 2% with 2007 rates of CT usage"
16:16 < kanzure> s/cancer machine/Cancer Tube
16:17 < fenn> heh
16:17 < fenn> a CT is relatively open compared to a MRI
16:17 < kanzure> i wonder if a high-resolution whole-body image is worth cancer
16:17 < kanzure> lots of cancers can be killed, so it's prolly worth it
16:18 < fenn> they really ought to just give people antioxidant injections before the CT
16:19 < fenn> pimp my CT scanner: http://brucejonesdesign.com/tag/ct-scanner/
16:19 < kanzure> hey that's a job
16:20 < kanzure> an entire career, even
16:20 < fenn> you could do a goatse theme
16:21 < kanzure> giant menacing monster of some kind
16:23 < fenn> would be much more fun with a transparent cover, you'd get to watch it whirl around at high speed http://i.imgur.com/TovNxVH.jpg
16:23 < fenn> .title https://www.youtube.com/watch?v=2CWpZKuy-NE
16:23 < yoleaux> CT at max speed - YouTube
16:25 < fenn> into the stargate you go
16:26 < kanzure> "expected a star gate to open"
16:26 < kanzure> why are you reading youtube comments?
16:26 < fenn> i wasnt
16:28 < kanzure> what's the spinning for? 
16:29 < kanzure> "Spinning tube, commonly called spiral CT, or helical CT is an imaging technique in which an entire X-ray tube is spun around the central axis of the area being scanned. These are the dominant type of scanners on the market because they have been manufactured longer and offer lower cost of production and purchase. The main limitation of this type is the bulk and inertia of the equipment (X-ray tube assembly and detector array on the ...
16:29 < kanzure> ... opposite side of the circle) which limits the speed at which the equipment can spin. Some designs use two X-ray sources and detector arrays offset by an angle, as a technique to improve temporal resolution."
16:29 < kanzure> "Electron beam tomography (EBT) is a specific form of CT in which a large enough X-ray tube is constructed so that only the path of the electrons, travelling between the cathode and anode of the X-ray tube, are spun using deflection coils. This type had a major advantage since sweep speeds can be much faster, allowing for less blurry imaging of moving structures, such as the heart and arteries. Fewer scanners of this design have been ...
16:29 < kanzure> ... produced when compared with spinning tube types, mainly due to the higher cost associated with building a much larger X-ray tube and detector array and limited anatomical coverage. Only one manufacturer (Imatron, later acquired by General electric) ever produced scanners of this design. Production ceased in early 2006.[86]"
16:30 < kanzure> er, they can't spin fast enough to catch up with their x-rays
16:31 < kanzure> oh right
16:31 < kanzure> put a human in between the emitter and the detector
16:32 < kanzure> move emitter/detector instead of move human
16:32 < fenn> an EBT would be like a gigantic CRT with a thick plastic screen
16:32 < kanzure> electron beam tomography machine http://archives.starbulletin.com/2001/09/06/news/arti.jpg
16:32 < kanzure> from http://archives.starbulletin.com/2001/09/06/news/story10.html
16:33 < fenn> huh it looks just like a helical scanner
16:33 < kanzure> "and a full-body scan [takes] about 20 minutes"
16:34 < fenn> where's the tube
16:34 < kanzure> "full-body scan for $1,095 and a heart scan for $495. For nonresidents the cost is $1,215 for a full scan and $550 for the heart."
16:37 < kanzure> 24 scans/day (well, maybe 20 if you count breaks in between each scan)
16:37 < kanzure> so $20k/day.. not bad..
16:40 < kanzure> would need maybe two people on staff, one in a reception room and one setting victims up and giving them usb sticks with data
16:43 < kanzure> $9k http://www.ebay.com/itm/GENERAL-ELECTRIC-HI-SPEED-NX-I-PRO-CT-scanner-2247006-/121256495320?pt=LH_DefaultDomain_0&hash=item1c3b7348d8
16:43 < fenn> CT scans only take 10 seconds
16:44 < fenn> also they're orders of magnitude less expensive for the machine
16:44 < kanzure> "less expensive for the machine"?
16:44 < fenn> than EBT
16:44 < kanzure> 10 seconds for whole-body?
16:45 < fenn> "sold as is" "for parts not-working"
16:45 < kanzure> pfft how hard could repair be. another $10k?
16:45 < fenn> according to the toshiba website http://medical.toshiba.com/resources/images/ct/aq-one-vision-thumb-trauma.jpg
16:46 < kanzure> huh... well.
16:46 < kanzure> you can't unload patients that fast
16:46 < kanzure> you would need a convey belt of patients ready to be loaded up
16:46 < kanzure> *conveyor
16:46 < kanzure> so maybe 5 minutes per patient... so 90 patients/day.
16:47 < kanzure> $90k/day ain't bad either
16:47 < fenn> .title http://www-formal.stanford.edu/pub/jmc/docdil.html
16:47 < yoleaux> THE DOCTOR'S DILEMMA ( 9-Dec-2002)
16:47 < kanzure> minus cancellations no-shows late-comers arrival-delays-and-jitter
16:48 < fenn> also be sure to read the solution to the dilemma http://www-formal.stanford.edu/pub/jmc/docdilsol.html
16:48 < kanzure> "People keep coming to him until he is exhausted, but there is always an emergency case more touching than all that have gone before and eventually he dies of exhaustion. Write his speech saying that he realizes he can cure more people if he gets some sleep, but true morality requires him to treat the immediate emergency."
16:48 < kanzure> that guy is a fucking idiot. obviously, they should go to him while he is sleeping.
16:48 < kanzure> and if they disturb him during his sleep, he should be allowed to murder them (his choice)
16:49 < kanzure> another option is to pick a fall-guy
16:49 < fenn> the gift is non-transferable
16:49 < kanzure> i don't understand this? "A mechanism should be provided to stop the motion of the finger of the patient momentarily so that it touches the doctor rather than brushes his skin."
16:50 < kanzure> what's wrong with brushing
16:50 < fenn> nothing at all
16:50 < kanzure> "On the basis of the arithmetic the doctor need only spend 1/60 th of his time curing people, i.e. 24 minutes per day." 
16:51 < kanzure> "Imagine further that the doctor, while posessing the gift of healing, is not a super-organizer or super-hero of any sort. How could one make literature of such a situation."
16:51 < kanzure> well, frankly, you would probably have a long period during which you are not optimally utilizing your talents
16:52 < kanzure> you don't just wake up one day and get a human conveyor belt built by moonlight
16:53 < kanzure> i thought my solution was pretty clever ("if he gets tired then let him sleep. so far there's no evidence that touching requires wakefulness.")
16:53 < kanzure> also you can touch more than one person at a time
16:53 < fenn> not that many
16:53 < kanzure> you can double your rate
16:54 < fenn> sure, and more importantly two conveyor belts would be twice as reliable
16:54 < kanzure> so now instead of 24 minutes it's only 12 minutes. thankfully we do not live in a three-dimensional world.
16:55 < fenn> "He forms an organization for curing people and at first works very hard but gradually gets lazy, is corrupted by desire for money, power, fame and women, requires more and more flattery and obsequiousness, eventually strives single-mindedly for power, develops cruel tastes, comes to dominate the country, and is finally assassinated."
16:55 < fenn> this could happen even with the conveyor belt
16:56 < kanzure> q: does his medical talents work on psychological illness and can it work on himself?
16:56 < fenn> touching yourself is morally wrong
16:57 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
16:57 < kanzure> even if it saves a billion people?
16:57 < kanzure> what was it, 70 million per year?
16:57 < kanzure> that adds up, you know
16:59 < fenn> this story is obviously a metaphor for automated production and the waste created by human irrationality
16:59 < kanzure> was it because i mentioned a conveyor belt or was there some other meaning here
16:59 < kanzure> like, "you will become corrupt if you let people see inside their bodies and make money from that"
17:00 < fenn> can the machine feed itself? i'd say yes, because otherwise (if it were too inefficient) it wouldn't be able to produce enough for the entire planet
17:01 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 264 seconds]
17:01 < fenn> it was probably because you were envisioning shooting patients as fast as possible through a machine
17:01 < kanzure> oh good point you can probably use a projectile system
17:01 < fenn> "no no, we cant have efficiency, this is AMERICA"
17:02 < kanzure> "There are 28 healthcare providers who offer CT Scan procedures in Austin. The cheapest list price of an CT Scan in Austin is $390.00 while the highest price is $8,600.00. There are 22 different types of CT Scan provided in Austin."
17:03 < kanzure> "CT Neck Cost Average $1,300" wtf
17:03 < fenn> in addition to actually scanning you, someone has to look at the images
17:03 < kanzure> "Heart CT Scan Cost Average $6,900" wtf x2
17:03 < kanzure> yeah well fuck that part. why should that be coupled to that service?
17:04 < fenn> heart requires EBT (i guess) because it's moving around
17:04 < kanzure> iirc they often do not have your primary physician looking at the scan in real-time
17:04 < kanzure> just some technician making "hmm" sounds
17:04 < fenn> your primary physician isn't trained to look at images
17:04 < fenn> the technician isn't the guy who looks at the images either
17:05 < kanzure> whatever, i still think having images is more useful than not having images
17:05 < fenn> me too
17:05 < kanzure> especially for giant tumor cases
17:05 < fenn> i think it's insane that you have to purchase the services of the guy looking at images bundled with the actual scan
17:05 < fenn> nevermind the high cost, and they won't tell you the cost, and they won't give you a scan without a doctor's order, etc
17:06 < Lemminkainen> if you give uneducated people more choice fenn how do you prevent fraudulent claims and corruption in their medical lives?
17:06 < kanzure> "According to Petsbest.com, the cost for a scan ranges from $800 to $1,200 for each diagnostic screening.  The total cost may be more than this if multiple scans may be needed.
17:06 < kanzure> Read more: http://www.howmuchisit.org/dog-ct-scan-cost/#ixzz3KPoRpr5z"
17:06 < kanzure> ffff
17:06 < Lemminkainen> most people don't connect the dot between "4 donuts per day" and "diabeetus"
17:07 < kanzure> oh is that what it takes to get diabetes? i've been wondering.
17:07 < fenn> kanzure you didn't know?
17:07 < fenn> soda works too
17:07 < Lemminkainen> if you want to get there faster you can quaff soda by the liter
17:07 < fenn> quaff, my pretties, quaff!
17:08 < kanzure> yeah i've often wondered how much junk i would have to eat to get diabetes
17:08 < Lemminkainen> well how active are you?
17:08 < Lemminkainen> I'm sure we can optimize this process
17:08 < fenn> it's partly genetic, and it depends on how much you exercise, and whether you get real nutrition mixed in with the junk or not
17:08 < kanzure> if you get real nutrition, the 4 donuts/day matter less?
17:08 < kanzure> that sucks
17:08 < fenn> yes
17:09 < kanzure> i was guestimating maybe these people are eating 2-3 gallons of ice cream or something per day
17:09 < kanzure> soda makes sense
17:09 < fenn> any juice really
17:09 < fenn> or what passes for juice in this country
17:09 < kanzure> "In fact, some CT units cost over $100,000" haha
17:10 < kanzure> so you make your money back in like a week of full use
17:10 < kanzure> let's do this
17:10 < Lemminkainen> you just want a CT scanner or do you want a full clinic?
17:11 < kanzure> "More than 70 million CT scans per year are now performed in the US"
17:11 < kanzure> depends on what you mean full clinic
17:11 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Ex-Chat]
17:11 < kanzure> i'm thinking: buy a ct machine, maybe two or three interns, rent some property and stick a website up
17:12 < kanzure> so 70M scans/year at $1k/scan is $70B/year.. nice.
17:12 < kanzure> good market size
17:13 < fenn> i'm seeing average scanner price is #1-3 million
17:13 < fenn> $
17:13 < kanzure> veterinary or no?
17:13 < fenn> hospital http://www.beckershospitalreview.com/hospital-key-specialties/12-statistics-on-ct-scanner-costs.html
17:14 < Lemminkainen> full clinic = insurance, certs, actual reimbursement relationships in place with insurers (who is going to pay for these scans), radiologist on staff
17:14 < kanzure> certs?
17:14 < kanzure> nope, no insurance please
17:14 < Lemminkainen> the various certifications necessary for a clinic, will vary by location
17:14 < kanzure> $1k out of pocket or bust
17:14 < fenn> to make sure you're not zapping people with deadly radiation
17:14 < fenn> how about a cruise ship in international waters
17:15 < Lemminkainen> OK, so you want to sell scans as a consumer product outside of insurance so you'll only be able to target those consumers who can dish out $1K on top of their insurance premiums
17:15 < kanzure> you would have to negotiate with the cruise ship company, so no thanks
17:15 < kanzure> Lemminkainen: yep
17:15 < Lemminkainen> talk to Blueseed
17:15 < kanzure> blueseed doesn't actually have a cruise ship yet
17:15 < kanzure> nice try
17:15 < kanzure> and yes the number of people who can afford it sans insurance is smaller than the total market size, but that doesn't sound like a big deal to me
17:16 < fenn> "I had a full body PET scan as well as a chest and brain CT. I got the bill today. 26K before insurance, 600 after."
17:16 < kanzure> these guys provide a scanner attached to a semi-truck for short-term rentals http://www.magnaserv.com/ct-scanner-mobile-rentals.html
17:17 < fenn> yep i've seen one of those parked; they have little feet/jacks that take the load off the tires for stability
17:17 < kanzure> what was it being used for?
17:18 < kanzure> $400k http://www.alibaba.com/product-detail/16-16-Slice-CT-scanner_1529474927.html
17:18 < fenn> implanting mind-control devices into innocent american citizens
17:18 < Lemminkainen> they're mobile health clinics that can be taken to places that need them most; such as retirement homes and rural Tennessee
17:19 < Lemminkainen> and Blueseed may not have a ship yet, but Dan will be easier to convince than Carnival Cruise Lines
17:20 < kanzure> here's one for "$2k-$30k" http://www.alibaba.com/product-detail/MY-D055-Dual-slice-CT-scanner_60085344975.html
17:20 < fenn> that can't be right
17:20 < kanzure> "life-time maintenance"
17:21 < fenn> the pictures of the cardboard box are too small
17:24 < fenn> life-time maintenance: it works until it doesn't
17:25 < kanzure> http://mayamedical.en.alibaba.com/search/product?_csrf_token_=1e9eedcmcow7f&IndexArea=product_en&SearchText=ct+scanner
17:28 < kanzure> here's one for $200? http://www.alibaba.com/product-detail/CT-SIEMENS-EMOTION-16-Slice_138968175.html
17:28 < kanzure> "2009 Siemens Emotion 16 Slice CT Scanner / System Scan Time: 162207.3 sec"
17:28 < kanzure> oh.. "199.000,00$"
17:29 < kanzure> $50k http://www.alibaba.com/product-detail/CT-Siemens-Somatom-Emotion-1ch-_50004200272.html
17:29 < kanzure> finding stuff on alibaba is not as straight forward as you would think
17:30 -!- rayston [~rayston@ip68-106-242-42.ph.ph.cox.net] has joined ##hplusroadmap
17:30 < fenn> siemens doesn't sell on alibaba
17:31 < fenn> neither does toshiba, GE
17:32 < kanzure> $5k, working but disassembled http://www.ebay.com/itm/Dissasembling-TOSHIBA-CT-Scanner-TCT-600HQ-for-parts-X-Ray-Tube-CXB-350A-/121489498242?pt=LH_DefaultDomain_0&hash=item1c4956a082
17:32 < fenn> anyway as i said before, one of these things would be really useful for doing hardware reverse engineering of the mesoscale variety
17:33 < fenn> "for parts" doesn't mean "working"
17:33 < fenn> hm ok
17:33 < kanzure> for all you know they think it means "working means that they don't need to put things together that might seem complex"
17:34 < kanzure> i would have to write some extra software huh :(
17:35 < kanzure> what's all this shit? http://i.ebayimg.com/00/s/MTYwMFg5MDA=/z/sPAAAOSw~FNUaQbE/$_57.JPG
17:35 < fenn> he shows the console
17:36 < kanzure> right but for all you know that's proprietary lockdown software
17:36 < superkuh> Looks like a Rockwell Retro Encabulator.
17:36 < kanzure> oh look it has a custom dashboard -_- http://i.ebayimg.com/00/s/OTAwWDE2MDA=/z/O-UAAOSw0vBUaQeI/$_57.JPG
17:36 < kanzure> is that a floppy disk slot?
17:37 < fenn> that's a 5 inch floppy slot
17:38 < fenn> worst case scenario you intercept the video signal
17:38 < superkuh> The tank in the bottom of ..._57.JPG might be the high voltage transformer under oil.
17:39 < kanzure> they are all $_57.JPG
17:39 < superkuh> ...oh.
17:39 < superkuh> http://i.ebayimg.com/00/s/MTYwMFg5MDA=/z/sPAAAOSw~FNUaQbE/$_57.JPG
17:40 < fenn> probably runs on 3-phase power
17:40 < kanzure> what sorta clinic or doctor's office has 3-phase?
17:40 < fenn> i'm just looking at the wires, man
17:43 < fenn> looks like a fun project, and dangerous in all the wrong ways
17:46 < fenn> put a mouse through and see if it develops cancer
17:46 < kanzure> don't they always develop cancer?
17:46 < kanzure> seems like an unfair test
17:47 < fenn> not if they die of protein deprivation first
17:49 < fenn> the insulating tape wrapped around HV electrical cables doesn't inspire confidence
17:51 < kanzure> i wonder how long their xray tubes last
17:57 -!- CheckDavid [uid14990@gateway/web/irccloud.com/x-nnucrjeoemeeawjo] has quit [Quit: Connection closed for inactivity]
17:57 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
17:59 -!- lichen [~lichen@c-50-139-11-6.hsd1.or.comcast.net] has quit [Quit: leaving]
17:59 -!- lichen [~lichen@c-50-139-11-6.hsd1.or.comcast.net] has joined ##hplusroadmap
18:01 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 272 seconds]
18:05 < fenn> the freestanding microstructures thing has "nano cantelievers" which could be useful for portable mass spectrometry
18:06 < fenn> a molecule binds to the tip and changes the resonant frequency; the frequency shift determined by the mass, and we can measure frequency very accurately
18:07 < fenn> i guess the trick is getting it to stick to the very end and not somewhere else
18:15 < fenn> if the cantilever were charged, electric field concentration would attract ions toward the tip
18:16 < kanzure> nano cantilevers are also useful for other cantilever things like afm
18:24 -!- maaku [~quassel@173-228-107-141.dsl.static.fusionbroadband.com] has joined ##hplusroadmap
18:24 -!- maaku is now known as Guest76468
18:25 < kanzure> pfft that's not how you publish a retraction http://www.cs.princeton.edu/~stevenag/bitcoin_threshold_signatures.pdf
18:27 -!- gab_ [~gab@151.66.150.206] has joined ##hplusroadmap
18:27 < kanzure> hello gab_ 
18:28 -!- gab_ [~gab@151.66.150.206] has quit [Client Quit]
18:28 < kanzure> :(
18:29 -!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 250 seconds]
18:31 -!- ebowden [~ebowden@147.69.179.205] has joined ##hplusroadmap
18:32 < kanzure> "Imaging at depth in tissue with a single-pixel camera" http://arxiv.org/pdf/1411.2731v1.pdf
18:32 < kanzure> i asked russell about aptamer imaging and he gave that link
18:35 < kanzure> fenn: still around?
18:35 -!- Guest76468 [~quassel@173-228-107-141.dsl.static.fusionbroadband.com] has quit [Remote host closed the connection]
18:35 -!- maaku_ [~quassel@173-228-107-141.dsl.static.fusionbroadband.com] has joined ##hplusroadmap
18:38 -!- Russell [~textual@cpe-74-73-107-82.nyc.res.rr.com] has joined ##hplusroadmap
18:38 < Russell> Hi
18:38 < kanzure> welcome back to the 90s
18:38 < Russell> hey hey kanzure
18:38 -!- Russell is now known as Guest26212
18:38 < kanzure> we were looking at patch-clamp things earlier today:
18:38 < kanzure> .title http://journal.frontiersin.org/Journal/10.3389/fncel.2014.00145/full
18:38 < yoleaux> Frontiers | Patch-clamp recordings of rat neurons from acute brain slices of the somatosensory cortex during magnetic stimulation | Frontiers in Cellular Neuroscience
18:40 -!- maaku__ [~quassel@173-228-107-141.dsl.static.fusionbroadband.com] has joined ##hplusroadmap
18:40 -!- maaku_ [~quassel@173-228-107-141.dsl.static.fusionbroadband.com] has quit [Read error: Connection reset by peer]
18:41 -!- ebowden [~ebowden@147.69.179.205] has quit [Remote host closed the connection]
18:43 -!- Guest26212 is now known as russell0
18:45 < kanzure> fenn: superkuh: this is the same russell0 from brainbackups.com
18:45 < kanzure> he has promised a 90% discount on all brainbackup services to us
18:46 < russell0> http://arxiv.org/pdf/1411.2731v1.pdf
18:48 -!- Irssi: ##hplusroadmap: Total of 75 nicks [0 ops, 0 halfops, 0 voices, 75 normal]
18:51 < kanzure> russell0: so are you aiming for sub-synapse resolution in the short term?
18:51 < kanzure> short term i guess means <5 years
18:52 < yottabit> "Brain Backups is actively developing new methods and has developed a number of proprietary techniques for imaging neurons using synthetic biology, MRI, smart contrast agents and sensors, and aptamer-based technology."
18:52 < russell0> sub micron, synapse receptor densities
18:52  * yottabit is curious about the synbio techniques
18:52 < kanzure> so the storage cost estimates are based on receptor resolution?
18:52 -!- HEx2 [~HEx@hexwab.plus.com] has quit [Ping timeout: 256 seconds]
18:53 < russell0> connectivity and synaptic weight resolution yes
18:53 < yottabit> also, hi russell0
18:53  * yottabit is around for a few mins until the gf tells me to get off the laptop and eat
18:53 < russell0> hi yottabit!
18:55 < Lemminkainen> hey russell0
18:55 < yottabit> i'm too intrigued
18:55 < kanzure> .title https://www.youtube.com/watch?v=nJpFvQ0YZLk
18:55 < yoleaux> Ultra high Resolution 3D Human Brain Model (BigBrain) in Atelier3D Viewer - YouTube
18:55 < kanzure> "The bigbrain is the brain of a 65yo woman with no neurological or psychiatric diseases in clinical records at time of death. The brain was embedded in parafin and sectionned in 7404 coronal histological sections ( 20 microns), stained for cell bodies. The bigbrain is the digitized reconstruction of the hi-res histological sections (20 microns isotropic)."
18:56 < kanzure> obviously this is not cellular resolution or anything
18:56 < kanzure> but they have data up on ftp ftp://bigbrain.loris.ca/
18:56 < yottabit> is mri = fmri in the case of brainbackups?
18:57 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
18:57 < yottabit> hm, i'm told i have to be afk again
18:58 < kanzure> besides receptor density is there anything else on your list of things you suspect are necessary to get a workable backup?
18:58 < kanzure> or even s/receptor density/receptor distribution
18:59 < kanzure> you could probably reconstruct individual synapses just from receptor location data
18:59 < russell0> Allen Brain Atlas is another good, Mouse, mesoscale connectome
19:00 < russell0> we're looking at histological type data but getting it non-destructively using nanoCT with 200-400nm voxel sizes
19:00 < kanzure> what sort of tissue depths?
19:00 < russell0> Anything that **doesn't** involve serial sectioning, etc.
19:00 < russell0> nanoCT can do any depth, it's just X-ray
19:01 < kanzure> this was from earlier commentary today:
19:01 < kanzure> 16:00 < fenn> i'm highly skeptical that X-ray imaging can produce a connectome and not destroy the brain in the process
19:01 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 264 seconds]
19:01 < russell0> A mesoscale connectome of the mouse brain www.ncbi.nlm.nih.gov/pubmed/24695228
19:02 < kanzure> maybe you pump the brain with antioxidants first though
19:02 < kanzure> .title
19:02 < yoleaux> A mesoscale connectome of the mouse brain. - PubMed - NCBI
19:02 < russell0> was that 16:00 today?
19:02 < kanzure> 16:00 PST
19:02 < kanzure> (today)
19:02 < russell0> we're not 100% invested in CT of course
19:03 < russell0> just running a couple of experiments using that imaging modality now
19:03 < kanzure> heh so i was looking at ct scanners on ebay today
19:03 < kanzure> http://www.ebay.com/itm/Dissasembling-TOSHIBA-CT-Scanner-TCT-600HQ-for-parts-X-Ray-Tube-CXB-350A-/121489498242?pt=LH_DefaultDomain_0&hash=item1c4956a082
19:03 < kanzure> for different purposes (whole body scans)
19:03 < russell0> We're running some experiments imaging on this one: http://www.ge-mcs.com/en/radiography-x-ray/ct-computed-tomography/nanotom-s.html
19:04 < kanzure> how long do those xray tubes last?
19:04 -!- ebowden [~ebowden@147.69.179.205] has joined ##hplusroadmap
19:05 < russell0> A long long time ==  it's a high quality instrument
19:05 < kanzure> neat
19:05 < russell0> This is definitely what a brain looks like in CT without any specific contrast: http://i.ebayimg.com/00/s/OTAwWDE2MDA=/z/qQAAAOSw~FNUaQaI/$_57.JPG
19:06 < kanzure> 16:01 < kanzure> he said x-ray plus contrast agents 
19:06 < kanzure> 16:01 < fenn> oh in that case the singularity must be here
19:06 < kanzure> 16:01 < fenn> think about it, contrast agents means you have to shoot more x-rays 
19:06 < kanzure> 16:01 < fenn> more x-rays means more damage 
19:07 < russell0> "think about it, contrast agents means you have to shoot more x-rays" nonsense
19:12 < kanzure> are there any blood brain barrier issues with aptamers?
19:15 < russell0> of course -- but aptamers can be selected to cross the BBB
19:15 < kanzure> hmm so the setup there would be,
19:15 < kanzure> make some aptamer selection process that has a facisimile to BBB?
19:15 < kanzure> because the test cycle would be way too long testing against real BBB... i think.
19:16 < russell0> no
19:16 < kanzure> (i spent some time in an aptamer lab (ellington))
19:16 < russell0> For instance: In vivo SELEX for Identification of Brain-penetrating Aptamers www.nature.com/mtna/journal/v2/n1/full/mtna201259a.html
19:16 < kanzure> (but didn't get to work with aptamers....)
19:16 < kanzure> oh that is a neat one
19:17 < russell0> There are many other examples...
19:17 < kanzure> "e found that the aptamer library employed here required more than 15 rounds of in vivo selection for convergence to specific sequences."
19:17 < kanzure> *we found
19:17 < kanzure> 15 is not too bad
19:18 < russell0> Aptamers and antibodies, nanobodies that participate in receptor mediated endocytosis are another route
19:19 < kanzure> so you will eventually want a library of different aptamers per receptor variant
19:19 < kanzure> hmm
19:20 < kanzure> and would you dose a brain with the full library, or just one receptor type at a time?
19:20 < russell0> we have a multi-functionalization scheme
19:20 -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Remote host closed the connection]
19:20 < russell0> so one NP can hit multiple receptors
19:20 < kanzure> oh, do you care about their types though?
19:21 < russell0> of course!!
19:21 < kanzure> also do you care about ion channels
19:22 < kanzure> so one nanoparticle hits multiple receptor types.. how do you distinguish that signal?
19:22 < russell0> sure for instance glutamate gated ion channels
19:22 < kanzure> cool
19:23 < russell0> So in case anyone on here cares: we're raising money to build the imaging agents for the full list of neuroreceptors and neurotransmitters at https://www.indiegogo.com/projects/brain-backups/
19:23 < russell0> you should definitely get involved if you believe in non-destructive non-invasive brain imaging
19:24 < russell0> 8)
19:24 < kanzure> hmm
19:24 < kanzure> i'd rather take equity
19:24 < russell0> Or invest...
19:25 < russell0> Two videos that explain a lot of the background are linked from the Indiegogo page
19:25 < russell0> One given a month ago at SingularityNYC
19:25 < kanzure> we have seen both videos :)
19:26 < russell0> all of you?? :)
19:26 < kanzure> i kickban people that don't follow along
19:27 < kanzure> helps maintain signal/noise
19:27 < lichen> you havent banned me yet
19:27 < kanzure> that's because nmz787 is trying to hire you
19:27 < lichen> fair
19:28 -!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap
19:28 < russell0> I'm prob. gonna take off from so active typing soon.  Kanzure -- hope I filled in some info you wanted.
19:29 < kanzure> yep absolutely
19:29 < kanzure> feel free to idle around
19:29 < russell0> cool
19:29 < kanzure> there will be more questions
19:54 < kanzure> russell0: which things are currently working, with the CT scanner?
19:57 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
20:01 < kanzure> http://en.wikipedia.org/wiki/Scanning_joule_expansion_microscopy
20:01 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 264 seconds]
20:08 < fenn> guh why do people only appear while i'm away
20:09 < kanzure> he hasn't completely disappeared
20:09 < russell0> One of the problems with CT is we haven't figured out how to barcode the nanoparticles
20:09 < russell0> Barcoding in other imaging modalities like MR, NIR, etc. is comparably easier
20:10 < russell0> we can barcode the RNA or DNA on the NP and get it off after imaging using CT, but not *during* imaging at present
20:10 < russell0> which I find annoying
20:11 < russell0> If there are any materials physics wizards who can figure out barcoding NP in nanoCT do get in touch
20:12 < russell0> If you're following this thread, this paper might be interesting: Sequencing the Connectome www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001411
20:12 < kanzure> we have at least one ab initio chemistry person in here
20:12 < kanzure> or we did, i wonder where she went
20:12 < fenn> that mouse paper was based on 469 mouse brains; i'm not sure what "TissueCyte STP tomography image" is
20:15 < russell0> It's a $900K two-photon serial tomography machine from a company called TissueVision
20:16 < russell0> Founded by a guy Tim Ragan
20:16 < kanzure> what makes it $900k :)
20:16 < russell0> It breaks so often! ;)
20:16 < kanzure> ha
20:16 < russell0> STP = serial two-photon
20:18 < kanzure> any chance of capturing data about individual cell bodies? axons, etc
20:20 < fenn> how does barcoding work in MRI and NIR?
20:21 < russell0> From that paper: "Recombinant adeno-associated virus (AAV), serotype 1, expressing EGFP optimally was chosen as the anterograde tracer to map axonal projections19,20. We also confirmed that AAV was at least as efficient as, and more specific than, the classic anterograde tracer biotinylated dextran amine (BDA) (Extended Data Fig. 1), as described separately21.
20:21 < russell0> EGFP-labelled axonal projections were systematically imaged using the TissueCyte 1000 serial two-photon (STP) tomography system22, which couples high-speed two-photon microscopy with automated vibratome sectioning of an entire mouse brain."
20:21 < fenn> am i right in thinking that the "bar code" is an absorption/fluorescence spectrum?
20:24 < fenn> also why is injecting large amounts of heavy metals not considered destructive?
20:24 < fenn> i didn't watch the video
20:26 < kanzure> https://www.dropbox.com/s/ww9x698xd75n5vd/A%20mesoscale%20connectome%20of%20the%20mouse%20brainnature13186.pdf?dl=0
20:26 < fenn> (gadolinium is considered a heavy metal right?)
20:26 < russell0> gold NPs have very little toxitiy.  Iron or iron oxide is ingested in the multi-gram amounts as a nutritional supplement.  Colloidal silver is an age-old anti-bacterial.
20:26 < fenn> woah woah woah
20:27 < fenn> just because some misguided soul ingests something doesn't mean you can put it in your brain directly
20:27 < russell0> If you can show any toxicity of a 15-50nm e.g. Au NP let me know
20:28 < fenn> .title http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2988217/
20:28 < yoleaux> Toxicity and cellular uptake of gold nanoparticles: what we have learned so far?
20:29 < fenn> i haven't read this yet
20:32 < fenn> "use of the CTAB molecules is essential and thus the gold nanorods are “born” with bound surfactant, ... CTAB alone is a quite toxic to cells at sub-micromolar dose" (CTAB is a surfactant)
20:34 < fenn> "possible size-dependent toxicity of gold nanoparticles ... In particular, gold nanoparticles less than 2 nm in diameter show evidence of chemical reactivity that does not occur at larger sizes"
20:36 < fenn> if you talk about nanoparticles as "micromolar" does that mean moles of gold or moles of particles?
20:36 < russell0> Thanks for the paper.  Toxicity in mice leading to cellular damage and inflammation in the liver, but at what dose?  No toxicity in zebrafish.  It's complicated!
20:37 < Lemminkainen> particles, fenn
20:39 < russell0> Investigating the optimal size of anticancer nanomedicine www.pnas.org/content/111/43/15344
20:39 < fenn> looking at table 1 it seems that particles around 1.4 nm are the most toxic
20:39 < fenn> er, inprevious paper
20:39 < russell0> Investigating the Impact of Nanoparticle Size on Active and Passive Tumor Targeting Efficiency pubs.acs.org/doi/abs/10.1021/nn500299p
20:40 < fenn> blarg
20:40 < fenn> russell0: sorry all our paper paywall-crossing infrastructure went offline like two days ago
20:40 < russell0> lol get a login! ;)
20:40 < kanzure> one sec booting it up
20:40 < russell0> ehh sorry
20:41 < russell0> anyway 50 nm is a good size for low toxicity
20:41 < russell0> and good activity in cancer applications
20:41 -!- paperbot [~paperbot@unaffiliated/kanzure/bot/paperbot] has joined ##hplusroadmap
20:41 < kanzure> paperbot: http://pubs.acs.org/doi/abs/10.1021/nn500299p
20:41 < russell0> "Here, we systematically evaluated the size-dependent biological profiles of three monodisperse drug–silica nanoconjugates (NCs; 20, 50, and 200 nm) through both experiments and mathematical modeling and aimed to identify the optimal size for the most effective anticancer drug delivery. Among the three NCs investigated, the 50-nm NC shows the highest tumor tissue retention integrated over time"
20:44 < kanzure> guess paperbot is still broken
20:44 < fenn> ok i am willing to concede that 50nm gold particles are probably not toxic
20:45 < russell0> nitroxides are another neat non-NP contrast agent
20:49 < fenn> assuming you have a brain full of nanoparticles stuck in the right places, how do you get an x-ray image of individual particles?
20:50 < russell0> So many question Fenn... so many questions
20:50 < russell0> Do you think I can just tell you **everything**??
20:50 < fenn> yes
20:50 < fenn> what comes after exa
20:50 < fenn> is it yocto
20:51 < kanzure> russell0: if you get hit by a bus we'll probably be the only ones left to build it
20:51 < russell0> "assuming you have a brain full of nanoparticles stuck in the right places, how do you get an x-ray image of individual particles?"  what do YOU think
20:51 < fenn> ok a brain (20cm cube) would have 1 zettavoxel at 20nm resolution
20:51 < kanzure> one way i can think to get an x-ray image of individual nanoparticles is to put a scanner up really close
20:52 < fenn> there's no way to image that many pixels at once so you need to scan
20:52 < fenn> here's where my knowledge of x-ray optics fails
20:53 < fenn> presumably you can build something approximating a lens with a nanoscale zone plate
20:54 < russell0> basically the x-ray scintillator is 4000px X 4000px
20:54 < russell0> you can zoom in the field of view sort of as much as you want
20:54 < fenn> but focusing on something a few nm in size, several inches away, would require a very precise lens
20:54 < russell0> this gives a computed tomographic image in 3D
20:54 < russell0> no
20:54 < fenn> is it pinhole optics?
20:54 < russell0> no
20:55 < fenn> um... x-ray lasers don't exist afaik
20:55 < fenn> x-ray mirrors are ridiculously hard to fabricate and not very precise even if you do
20:56 < russell0> "The detector in the nanotom is a flat panel digital detector. The model is a DXR 500L and it is made by GE healthcare. It has a Cesium Iodide scintillator on an amorphous Silicon plate."
20:57 < fenn> that doesn't tell me anything about what the emitter looks like, or the optical path
20:57 < fenn> a caveman could make a scintillator
20:57 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
20:58 < russell0> http://d32ogoqmya1dw8.cloudfront.net/images/research_education/geochemsheets/techniques/_1172960774.png
20:59 < fenn> ok so that point source, if it's produced by an impinging electron beam, will heat up whatever it's hitting and spall away material
20:59 < fenn> or sputter i guess
21:00 < fenn> the focal plane will change, or the source will change if you aim the beam at a different place
21:00 < russell0> What are you trying to get at?
21:00 < fenn> how to image nanoscale objects with x-rays
21:01 < russell0> just look at the design of nanoCT machines from GE, siemens, etc.
21:01 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 256 seconds]
21:03 < fenn> xradia nanoxct-100 "Produces high-resolution 3D X-rays as fine as 50 nm; depth of focus up to 65 microns
21:03 < fenn> - Zone plate and capillary based X-ray optics"
21:03 < russell0> lol anyone except xradia
21:04 < fenn> GE nanoCT "sample sizes of up to 250 x 240 mm and the combination of proprietary GE technology in terms of X-ray tube, detector, generator and CT software ensures that a voxel size of down to 300 nm (0.3 µm) can be achieved."
21:04 < russell0> yeah!!!
21:05 < kanzure> "proprietary magic" is not an answer
21:05 < kanzure> quick, to the patent machine!
21:05 < kanzure> https://www.google.com/search?num=100&safe=off&site=&tbm=pts&source=hp&q=nanoCT&oq=nanoCT&gs_l=hp.3..0l2j0i10l8.457.1011.0.1252.7.6.0.0.0.0.297.775.0j2j2.4.0.cprnk%2Cekomodo%3Dtrue%2Ckpnr%3D100...0...1.1.58.hp..4.3.477.0.ae3JqWK-PSA
21:07 < fenn> 240mm/3072px = 78 micron so they're doing some kind of scanning to get down to 0.3 micron
21:08 < russell0> right.  in fact the machine will not do 200nm voxel at 240mm field of view, it will do something like 78 um
21:11 < fenn> am i missing something here? the xradia machine had better specs
21:11 < fenn> Zernike phase contrast mode at 8 keV with 50-nm resolution using Cu rotation anode X-ray source
21:12 < fenn> so they rotate the electron beam target to prevent it from melting
21:12 < fenn> .title http://www.nature.com/nphys/journal/v6/n11/abs/nphys1765.html
21:12 < yoleaux> Zernike phase contrast in scanning microscopy with X-rays : Nature Physics : Nature Publishing Group
21:13 < kanzure> paperbot: http://www.nature.com/nphys/journal/v6/n11/abs/nphys1765.html
21:13 < paperbot> http://libgen.org/scimag/get.php?doi=10.1038%2Fnphys1765
21:13 < kanzure> there we go
21:13 -!- nmz787 [~nmz787@131.252.130.250] has joined ##hplusroadmap
21:13 -!- nmz787 [~nmz787@131.252.130.250] has quit [Client Quit]
21:13 < kanzure> jrayhawk_: ^
21:13 < ebowden> paperbot: http://www.nature.com/nature/journal/vaop/ncurrent/full/nature13886.html
21:13 < russell0> yeah I'm sure there's some machine out there with better specs than yours, doesn't mean you have that machine with better specs ;P
21:13 < paperbot> http://libgen.org/scimag/get.php?doi=10.1038%2Fnature13886
21:14 < russell0> heh
21:14 < russell0> https://itg.beckman.illinois.edu/microscopy_suite/equipment/nano_ct/
21:14 < kanzure> .title
21:14 < yoleaux> ITG : Xradia NanoCT (nanoXCT-100)
21:15 < fenn> oh i thought you were laughing at it because it was bad
21:15 < ebowden> \:D/
21:15 < ebowden> Thanks paperbot. :D
21:16 < russell0> http://www.lao.cz/data/ke-stazeni/nanoXCT_100_eu_1_.pdf voxel: 50nm, field of view 15 um
21:19 < ebowden> paperbot: http://www.cell.com/current-biology/abstract/S0960-9822%2814%2900123-7
21:19 < fenn> i am wondering what the "objective lens" is made from
21:20 < russell0> as a matter of fact, however, we're moving away from X-ray to NIR and the so-called NIR windows
21:20 < russell0> Au Nanorod Design as Light-Absorber in the First and Second Biological Near-Infrared Windows for in Vivo Photothermal Therapy pubs.acs.org/doi/abs/10.1021/nn401187c
21:20 < russell0> In the NIR region, two biological transparency windows are located in 650–950 nm (first NIR window) and 1000–1350 nm (second NIR window)
21:21 < russell0> Because of the difficulty of fast scanning on nanoCT systems
21:21 < fenn> russell0: to be clear, you are intending to scan living brains?
21:21 < kanzure> paperbot: http://pubs.acs.org/doi/abs/10.1021/nn401187c
21:22 < russell0> yup
21:22 < fenn> did you know that they move around with heartbeat?
21:22 -!- nmz787_w [83fc82fa@gateway/web/freenode/ip.131.252.130.250] has joined ##hplusroadmap
21:22 < russell0> and breathing and ...
21:23 < kanzure> they also grow
21:23 < russell0> and die!
21:23 < kanzure> they grow entire nanometers per day i think
21:23 < kanzure> at least
21:23 < kanzure> i knew a number once...
21:23 < fenn> well that's interesting but moving around like a bowl full of jelly seems like a problem for any kind of scanning approach
21:23 < kanzure> it is somewhere in http://diyhpl.us/~bryan/papers2/neuro/
21:24 < russell0> Fenn - how would you measure synaptic weights using NIR?
21:24 < russell0> or Kanzure
21:25 < fenn> you'd need to do something like PALM/STORM but with IR wavelengths and biocompatible phosphors
21:25 < fenn> since NIR wavelength is so far above synaptic scale, you obviously need superresolution techniques
21:26 < fenn> you could bind antibodies to whatever you're measuring, say receptor proteins, at low enough concentration that it doesn't make you go nuts
21:27 < russell0> Or send RNA down the axon and cause a contrast event when it reaches its companion synapse
21:27 < fenn> a weak binding constant would allow them to detach float around and reattach to other proteins
21:27 < fenn> contrast has nothing to do with it
21:27 < russell0> Our microscopy core told us we can get 0.5um resolution in NIR without superresolution techniques
21:27 < fenn> in PALM the dye absorbs light at one wavelength, then randomly decays and emits another wavelength
21:28 < russell0> ever heard of upconverting NPs?
21:28 < nmz787_w> or raman
21:28 < nmz787_w> tip-enhanced raman maybe
21:28 < fenn> if you get the number of dye molecules down to around 1 per airy circle radius, you can calculate the exact center coordinates (in 3d) of the emission
21:28 < fenn> um, that was wrong
21:29 < fenn> the numeber of emission events*
21:29 < fenn> nmz787_w: anti-stokes scattering?
21:29 < russell0> Are we talking about CT anymore?
21:30 < russell0> A new website that someone is putting together not live yet http://brainbackups.flubula.net/
21:30 < fenn> uh, not sure what the terminology is, but PALM is usually done on a 2d microscope slide and gives you 3d coordinates
21:31 < russell0> we're not really interested in 2d microscope slide anything
21:31 < fenn> anyway regarding "upconversion" you can't do that because the emitted high frequency light will be blocked by the skull; the output has to be in NIR also
21:31 < fenn> actually it will be absorbed by the brain tissue
21:31 < fenn> scattered
21:31 < russell0> no -- I have data right now showing UC NPs emitting through tissue
21:31 < kanzure> .wik http://en.wikipedia.org/wiki/Photoactivated_localization_microscopy
21:31 < yoleaux> "Photo-activated localization microscopy (PALM or FPALM) and stochastic optical reconstruction microscopy (STORM) are widefield (as opposed to point scanning techniques such as laser scanning confocal microscopy) fluorescence microscopy imaging methods that allow obtaining images with a resolution beyond the diffraction limit." — http://en.wikipedia.org/wiki/Photoactivated_localization_microscopy
21:31 < fenn> i imagine NIR will be scattered by brain tissue too
21:32 < russell0> Does anyone have anything constructive to say?
21:32 < russell0> Like using these tools how would you approach this?
21:32 < dingo_> TITS AND BEER
21:32 < fenn> i just told you?
21:33 < russell0> Try again
21:33 < dingo_> I just implemend xmodem for my telnet bbs ....? :)
21:34 < fenn> russell0: this has been an entertaining game but you've provided zero input
21:34 < russell0> I bet
21:34 < russell0> Fenn: "how to image nanoscale objects with x-rays"
21:37 < fenn> .title http://www.nature.com/nphoton/journal/v8/n9/full/nphoton.2014.166.html
21:37 < yoleaux> Through-skull fluorescence imaging of the brain in a new near-infrared window : Nature Photonics : Nature Publishing Group
21:37 < nmz787_w> fenn: http://diyhpl.us/~nmz787/pdf/Apertureless_near-field_far-field_CW_two-photon_microscope_for_biological_and_material_imaging_and_spectroscopic_applications.pdf
21:38 < russell0> Yeah we found that through-skull paper a month ago
21:38 < nmz787_w> ,wik optical coherence tomography
21:38 < nmz787_w> .wik optical coherence tomography
21:38 < yoleaux> "Optical coherence tomography (OCT) is an established medical imaging technique that uses light to capture micrometer-resolution, three-dimensional images from within optical scattering media (e.g., biological tissue). Optical coherence tomography is based on low-coherence interferometry, typically employing near-infrared light." — http://en.wikipedia.org/wiki/Optical_coherence_tomography
21:39 < nmz787_w> fenn: I believe I common real-time-clock crystal might suffice
21:39 < fenn> nmz787_w: topic is "how would you recover synaptic density and connectome from a living brain without hurting it"
21:39 < nmz787_w> fenn: ah, ok
21:39 < kanzure> receptors, not just synapses
21:40 < nmz787_w> there must be a JTAG-like way
21:40 < fenn> :)
21:40 < nmz787_w> through the eyes
21:40 < fenn> anselm's idea was to infect the brain with genetically engineered protozoa that leave fiber-optic poop trails behind them going back to the source
21:40 < kanzure> i wonder if russell0 knows anselm
21:40 < kanzure> is anselm considered a 3scan person?
21:41 < fenn> that way they could pick up light emitted by the cells before it was scattered
21:41 < russell0> I think synaptic weights and cell subtypes are more important than synaptic densities
21:41 < kanzure> russell0: btw anselm just raised $9M the other day for his dna synthesis startup, cambrian genomics
21:41 -!- yottabit [uid36770@gateway/web/irccloud.com/x-lcrydekpayeqaqbw] has quit [Quit: Connection closed for inactivity]
21:41 < kanzure> well, 10
21:42 < fenn> i'm sure they've met
21:44 < fenn> russell0: your upconverting nanoparticles, do they emit x-rays?
21:44 < russell0> Some do!
21:44 < fenn> hrm that's a thing
21:48 < russell0> "infect the brain with genetically engineered protozoa that leave fiber-optic poop trails behind them going back to the source" sounds worse than Parkinson's and Alzheimer's at the same time
21:49 < russell0> Tau proteins stabilize axons but in over-abundance kill neurons, causing Alzheimer's, etc.
21:50 < fenn> "Metallic nanoparticles (MNP) are able to release localized x-rays when activated with a high energy proton beam by the particle-induced x-ray emission (PIXE) effect."
21:50 < fenn> that's all i've found so far
21:51 < russell0> Or you could just image the densities or NR and NT's throughout a whole brain without killing it, or using CLARITY
21:51 < nmz787_w> pixe is not gonna work, how do you get the particles in there?
21:52 < fenn> the "particle" in PIXE is the proton
21:52 < nmz787_w> it's any particle with sufficient inertia
21:52 < nmz787_w> (in my experience gallium)
21:52 < fenn> .wik anatoli bugorski
21:52 < yoleaux> "Anatoli Petrovich Bugorski (Russian: Анатолий Петрович Бугорский; born 1942) is a Russian scientist who was struck by a particle accelerator beam in 1978." — http://en.wikipedia.org/wiki/Anatoli_Bugorski
21:52 < nmz787_w> but even so, how do you accelerate the particles?
21:52 < russell0> listeria vs. protozoa
21:53 < russell0> energy
21:53 < nmz787_w> that won't help anyone practically though
21:53 < nmz787_w> need a /bit/ more definition of the system
21:54 < fenn> russell0: having trouble decoding your acronyms, what's "NR and NT's"
21:54 < russell0> neuroreceptors and neurotransmitters
21:54 < fenn> ok i didnt realize "neuroreceptor" was a word
21:55 < ebowden> paperbot: http://link.springer.com/referenceworkentry/10.1007%2F978-1-4614-4788-7_108
21:56 < russell0> For example: dopamine receptors include D2R D3R, or D1-D5
21:56 < russell0> Dopamine transporter is imaged using DaTscan etc.
21:56 < russell0> there are 5 dopamine receptors
21:57 < fenn> yes we were talking about trying to do PET scan at high resolution earlier
21:57 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
21:58 < fenn> my impression was that it was infeasible due to detector size vs noise
21:58 -!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 264 seconds]
21:59 < fenn> for PET you need to correlate TWO photons, and picking them out against background radiation is a task
22:00 < fenn> "A PET scanner detects these emissions "coincident" in time, which provides more radiation event localization information and, thus, higher spatial resolution images than SPECT (which has about 1 cm resolution)."
22:01 < ebowden> paperbot: http://www.ncbi.nlm.nih.gov/pubmed/22173844
22:01 < paperbot> ConnectionError: HTTPConnectionPool(host='libgen.org', port=80): Max retries exceeded with url: /scimag/librarian/form.php (Caused by <class 'socket.error'>: [Errno 104] Connection reset by peer) (file "/usr/local/lib/python2.7/dist-packages/requests/adapters.py", line 375, in send)
22:01 -!- Beatzebub__ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 264 seconds]
22:02 < fenn> in order to reliably pick up a single photon (which you'd need to do for PET) you need large cascading detectors, which could be scintillators or avalanche diodes
22:03 < fenn> hm this is cool https://en.wikipedia.org/wiki/Silicon_photomultiplier
22:09 < fenn> bah basically no info online from srs.fbk.eu
22:10 < fenn> "Punch-through biasing"
22:11 < fenn> so anyway the gamma ray has to interact with regular matter in such a way that it gives off light; obviously traditional scintillation vials are out because they are so large that the pixel count would be too low
22:12 < fenn> maybe adding gold nanoparticles on top of the silicon avalance photodiodes would allow sufficiently compact detectors to localize the event to within a nanometers, but i dunno how to do the math for that
22:13 < fenn> or some other nanoparticle that has a reasonably high x-ray cross section
22:16 < fenn> "It has been calculated that only some 4% of the energy from a β emission event is converted to light by even the most efficient scintillation cocktails."
22:16 < fenn> (4%)^2 = 0.16%
22:17 < nmz787_w> PIN photodiodes can detect gamma and xray
22:17 < nmz787_w> they're cheap too
22:17 < fenn> oh 4% of the energy, nevermind
22:17 < nmz787_w> but they're just single pixel
22:17 < fenn> nmz787_w: directly?
22:19 < nmz787_w> yes
22:19 < nmz787_w> http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_2170812_-1
22:19 < nmz787_w> .title
22:19 < yoleaux> CJKIT-20255: JAMECO KITPRO: Education & Hobby Kits
22:20 < nmz787_w> http://cfi.lbl.gov/instrumentation/Pubs/LBNL-34873.pdf
22:20 < nmz787_w> PET something something .... PIN photodiode
22:20 -!- yashgaroth [~ffffff@2606:6000:cb85:6a00:cdbf:9d7c:fcfd:b08d] has quit [Quit: Leaving]
22:21 < nmz787_w> http://www.ifm.umich.mx/~villasen/AIP-Com_Vol/039-Ramirez-Jimenez.pdf
22:22 < nmz787_w> 'Application of PIN diodes in Physics Research'
22:23 < fenn> the LBNL one uses a scintillator
22:23 < fenn> big fat crystal
22:25 -!- Boscop_ [me@188.126.91.226] has joined ##hplusroadmap
22:27 < fenn> every time i read about x-ray spectroscopy i forget half of it
22:28 -!- nmz787_w [83fc82fa@gateway/web/freenode/ip.131.252.130.250] has quit [Ping timeout: 246 seconds]
22:31 -!- nmz787__ [83fc82fa@gateway/web/freenode/ip.131.252.130.250] has joined ##hplusroadmap
22:31 < nmz787__> http://diyhpl.us/~nmz787/pdf/Spectral_Response_of_Gamma_and_Electron_Irradiated_Pin_Photodiode.pdf
22:31 -!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap
22:31 < nmz787__> sorry I missed out
22:31 < nmz787__> browser crashed
22:32 < fenn> i hear ya. my laptop has started turning off randomly after being unplugged
22:33 < fenn> it's one way of closing tabs at least
22:34 < nmz787__> i tried buying an asus x205 (quad core atom with 1366x768 screen) earlier for $100 at staples but I went too late :/
22:34 < fenn> oh today is capitalism day
22:35 -!- nmz787___ [~nmz787@131.252.130.250] has joined ##hplusroadmap
22:37 -!- Netsplit *.net <-> *.split quits: Boscop__
22:38 -!- nmz787__ [83fc82fa@gateway/web/freenode/ip.131.252.130.250] has quit [Quit: Page closed]
22:43 < fenn> nmz787___: i can see you
22:46 -!- ebowden [~ebowden@147.69.179.205] has quit [Remote host closed the connection]
22:58 -!- Beatzebub [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap
22:58 < nmz787___> how do i change my nick from the irc command window?
23:01 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has quit [Ping timeout: 258 seconds]
23:05 < fenn> use /nick
23:06 < fenn> also you should set up sasl and get an unaffiliated cloak
23:10 < fenn> god damn robots crawling all over gnusha
23:10 < nmz787___> idk what sasl is, sounds like cryption
23:10 < nmz787___> i read some faq on cloaking but it didn't help me
23:11 < fenn> you need to go in #freenode and ask for a cloak
23:11 < nmz787___> hmm, says my nick nmz787 is already in use :(
23:11 < fenn> msg nickserv ghost nmz787
23:12 -!- nmz787___ is now known as nmz787
23:13 -!- nmz787 [~nmz787@131.252.130.250] has quit [Disconnected by services]
23:14 -!- nmz787 [~nmz787@131.252.130.250] has joined ##hplusroadmap
23:15 < nmz787> hmm, ok I guess i got it
23:22 -!- Lemminkainen [uid2346@gateway/web/irccloud.com/x-bwpumoubtwkdzvwh] has quit [Quit: Connection closed for inactivity]
23:32 -!- Viper168_ [~Viper@unaffiliated/viper168] has joined ##hplusroadmap
23:34 -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 258 seconds]
23:36 -!- ebowden [~ebowden@147.69.179.205] has joined ##hplusroadmap
23:49 -!- russell0 [~textual@cpe-74-73-107-82.nyc.res.rr.com] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
23:57 -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has joined ##hplusroadmap
23:58 -!- Beatzebub_ [~beatzebub@S0106b81619e8ecee.gv.shawcable.net] has joined ##hplusroadmap