2015-12-22.log

--- Log opened Tue Dec 22 00:00:43 2015
superkuhHeh. Interesting. I feel the same way about people who have had children.00:11
maakuEh.. maybe my point didn't get across. Two points actually:00:15
maaku(1) A part of transhumanism deals with either creating minds (AI) or hive minds, or otherwise enabling or constraining the capabilities of future intelligences, often with concern for desireable outcomes.00:17
maakuThere is a segment of the population who already has experience creating new minds and guiding towards desirable outcomes: parents. Non-parents talking about things like friendlyness in AI or hive minds or whatever is a bit like virgins talking about sex.00:18
superkuhI'm not sure I believe the premise that there'll be an equivalence between strong AI and children. People who have children are compromised in the sense that they'll do almost anything, even irrational things, to do what they perceive as protecting them.00:20
maaku(2) Having the responsibility for caring for and guiding development another human life changes your ethics and priorities. You become a different person with different priorities after bonding with a child you are responsible for..00:20
pompolicwhat if (2) is bad00:20
superkuhYes, exactly #2. They become compromised.00:20
pompolicsounds like you're just switching to a different set of biases00:21
pompolicactually disregard that last sentence; that's insufficient justification00:21
maakuExactly. I am a parent, btw, so I'm suspect of the judgement of those who are not.00:21
maakunot different biases, different base values00:22
maakusuperkuh: the flip side is that a parent would take seriously protective measures to prevent their kids from becoming AI-food00:24
maakuand not take risks a nihilist might00:24
maakuSo yeah, if someone tells me they are "childfree" and just got their tubes tied, yes I'll second-guess their ability to effectively and safely reason about exitential risks00:26
maakuThey've got no skin in the game.00:26
pompolicwhat if they want to live forever00:26
maakupompolic: what's that a response to?00:27
pompoliccounterargument to having no skin in the game00:27
maakupompolic: ah but it is human nature to have a higher threshold for risk taking when it's just yourself on th eline00:28
superkuhAnd for parents to hold back progress in order to create stability for their offspring.00:29
pompolicseems like you want parents if you want to make type 1 errors and nihilists if you want to make type 2 errors00:30
-!- jenelizabeth_ [~jenelizab@cpc76802-brmb10-2-0-cust399.1-3.cable.virginm.net] has quit [Read error: No route to host]00:54
-!- jenelizabeth__ [~jenelizab@cpc76802-brmb10-2-0-cust399.1-3.cable.virginm.net] has joined ##hplusroadmap00:54
fennhttps://en.wikipedia.org/wiki/Proactionary_principle01:07
maakufenn: ?01:16
Diablo-D3and what if we dont want to make mistakes at all?01:22
-!- andares is now known as Esmerelda01:23
Diablo-D3http://zerocarbzen.com/2015/12/21/zero-carb-interview-matt-shepherd/01:23
pompolicDiablo-D3: precisely. i was being flippant01:25
Diablo-D3http://www.biochemsoctrans.org/content/28/6/883.long01:32
Diablo-D3http://www.biochemsoctrans.org/content/ppbiost/28/6/883.full.pdf01:34
Diablo-D3actually there, the full pdf01:34
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap01:43
-!- triggerwarning [~IGLC@unaffiliated/triggerwarning] has quit [Ping timeout: 250 seconds]01:47
-!- dcentral [~IGLC@2601:681:500:165a:ed72:9d1:33a1:7bbd] has quit [Ping timeout: 240 seconds]01:47
-!- dcentral [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has joined ##hplusroadmap01:47
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap02:11
-!- atomical [~atomical@209.153.22.182] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]02:26
-!- fleshtheworld [~fleshthew@2602:306:cf0f:4c20:111d:e0c7:64b3:a0d7] has quit [Read error: Connection reset by peer]03:16
-!- jenelizabeth__ [~jenelizab@cpc76802-brmb10-2-0-cust399.1-3.cable.virginm.net] has quit [Quit: byeeee]03:19
-!- triggerwarning [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has joined ##hplusroadmap03:25
-!- triggerwarning [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has quit [Changing host]03:25
-!- triggerwarning [~IGLC@unaffiliated/triggerwarning] has joined ##hplusroadmap03:25
-!- Gurkenglas [Gurkenglas@dslb-188-103-077-131.188.103.pools.vodafone-ip.de] has joined ##hplusroadmap03:33
-!- Act [uid89656@gateway/web/irccloud.com/x-weubueydguaopvdl] has quit [Quit: Connection closed for inactivity]03:37
-!- jaboja [~jaboja@ery40.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap03:45
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has joined ##hplusroadmap03:54
-!- Houshalter [~Houshalte@oh-71-50-56-224.dhcp.embarqhsd.net] has joined ##hplusroadmap04:10
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-fzwnckehjoctwucx] has quit [Quit: Connection closed for inactivity]04:10
paskymaaku: there's plenty of people who are pretty bad parents and don't care much for their kids, and there's plenty of people who care deeply about this world or other people in it even if they don't have children; and I actually suspect the proportions are so large that parenthood carries little information in the end04:26
maakupasky: I doubt that. But what can I say? "If you were a parent you'd understand." Not helpful so I'll stop engaging.04:27
paskyfair enough :)04:29
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has quit [Ping timeout: 260 seconds]04:42
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has quit [Read error: Connection reset by peer]04:51
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap04:52
-!- jaboja64 [~jaboja@erz41.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap05:00
-!- jaboja [~jaboja@ery40.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 256 seconds]05:03
-!- atomical [~atomical@50.247.130.97] has joined ##hplusroadmap05:07
-!- FAMAS [~FAMAS@unaffiliated/famas] has joined ##hplusroadmap05:10
-!- FAMAS [~FAMAS@unaffiliated/famas] has quit [Max SendQ exceeded]05:11
-!- FAMAS [~FAMAS@unaffiliated/famas] has joined ##hplusroadmap05:11
-!- FAMAS [~FAMAS@unaffiliated/famas] has quit [Max SendQ exceeded]05:16
-!- FAMAS [~FAMAS@103.198.139.33] has joined ##hplusroadmap05:16
-!- FAMAS [~FAMAS@103.198.139.33] has quit [Changing host]05:16
-!- FAMAS [~FAMAS@unaffiliated/famas] has joined ##hplusroadmap05:16
-!- eudoxia [~eudoxia@r167-57-78-97.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap05:31
-!- dcentral [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has quit [Ping timeout: 240 seconds]05:34
-!- triggerwarning [~IGLC@unaffiliated/triggerwarning] has quit [Ping timeout: 260 seconds]05:35
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has joined ##hplusroadmap05:43
-!- jaboja64 [~jaboja@erz41.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 240 seconds]05:46
-!- triggerwarning [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has joined ##hplusroadmap05:46
-!- dcentral [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has joined ##hplusroadmap05:47
-!- yorick [~yorick@oftn/member/yorick] has quit [Remote host closed the connection]05:48
kanzure"Bitcoin's mining difficulty has increased by 41.9% over the last 30 days" https://news.ycombinator.com/item?id=1077420405:55
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has quit [Read error: Connection reset by peer]06:00
-!- Filosofem [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap06:00
-!- JayDugger [~jwdugger@108.19.186.58] has joined ##hplusroadmap06:10
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-plqutunanchzpmgo] has joined ##hplusroadmap06:13
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap06:16
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 264 seconds]06:32
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap06:50
-!- poppingtonic [~Thunderbi@unaffiliated/poppingtonic] has joined ##hplusroadmap07:00
-!- sandeepkr [~sandeepkr@111.235.64.4] has quit [Ping timeout: 256 seconds]07:04
-!- jenelizabeth_ [~jenelizab@cpc76802-brmb10-2-0-cust399.1-3.cable.virginm.net] has joined ##hplusroadmap07:09
-!- Filosofem [~Jawmare@unaffiliated/jawmare] has quit [Quit: Leaving]07:15
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap07:15
gnusha_https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=96844c4b Bryan Bishop: fix sipa's name >> http://diyhpl.us/diyhpluswiki/transcripts/gmaxwell-sidechains-elements/07:20
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 272 seconds]07:24
-!- erasmus [~erasmus@unaffiliated/erasmus] has quit [Read error: Connection reset by peer]07:25
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap07:33
-!- Houshalter [~Houshalte@oh-71-50-56-224.dhcp.embarqhsd.net] has quit [Read error: Connection reset by peer]07:42
-!- eudoxia_ [~eudoxia@r167-56-3-159.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap07:47
-!- eudoxia [~eudoxia@r167-57-78-97.dialup.adsl.anteldata.net.uy] has quit [Read error: Connection reset by peer]07:47
-!- eudoxia_ [~eudoxia@r167-56-3-159.dialup.adsl.anteldata.net.uy] has quit [Client Quit]07:49
-!- atomical [~atomical@50.247.130.97] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]07:52
-!- atomical [~atomical@50.247.130.97] has joined ##hplusroadmap07:52
-!- |node [uid125132@gateway/web/irccloud.com/x-tmazfrxeeusxeylc] has joined ##hplusroadmap08:21
-!- erasmus [~erasmus@unaffiliated/erasmus] has joined ##hplusroadmap08:26
-!- poppingtonic [~Thunderbi@unaffiliated/poppingtonic] has quit [Quit: poppingtonic]08:32
AdrianGkanzure: is it even possible for a nation state to 51% attack?08:35
kanzureselfish mining showed how you don't even need 51%, you need ~33% ish08:39
-!- erasmus [~erasmus@unaffiliated/erasmus] has quit [Read error: Connection reset by peer]08:42
-!- erasmus [~erasmus@unaffiliated/erasmus] has joined ##hplusroadmap08:43
-!- fleshtheworld [~fleshthew@2602:306:cf0f:4c20:15a5:eb3f:cc25:62ea] has joined ##hplusroadmap08:52
AdrianGwell.09:10
-!- FAMAS [~FAMAS@unaffiliated/famas] has quit [Ping timeout: 240 seconds]09:11
AdrianGwe could be being attacked with 33% right now then. there is an "unknown" pool of 18%09:11
maakubitfury is brining online a data center with hashing capability equal to the entire bitcoin network as of a few weeks ago09:15
-!- sandeepkr [~sandeepkr@111.235.64.4] has joined ##hplusroadmap09:19
xrrI would think it is at least easy to recognize the attack. Top of the current chain gets replaced09:21
kanzurethat's not so easy to recognize; that's normal bitcoin blockchain behavior. although lots of reorgs are less normal. but there probably wouldn't be that many reorgs even in an attack scenario like this.09:27
-!- atomical [~atomical@50.247.130.97] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]09:32
kanzure.title https://news.ycombinator.com/item?id=1077750509:33
yoleauxCamera Zapping: Using Lasers to Temporarily Neutralize Cameras | Hacker News09:33
xrrHmm, wouldn't time make a difference though? Normal differences would sync up quickly, wouldn't grow beyond a block or two, while a double spender would bury a transaction with some ten blocks to get a purchase confirmed and then remove the transaction. No?09:39
maakuxrr: there are many more attacks than that09:50
-!- Filosofem [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap10:01
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has quit [Read error: Connection reset by peer]10:01
AdrianG archels: degrees of consciousness? like coma scales?10:04
archelsyeah, exactly that10:11
archels.wik Glasgow coma scale10:11
yoleaux"The Glasgow Coma Scale or GCS is a neurological scale that aims to give a reliable, objective way of recording the conscious state of a person for initial as well as subsequent assessment." — https://en.wikipedia.org/wiki/Glasgow_coma_scale10:11
AdrianGarchels: thats just so you can triage/sort patients10:36
AdrianGit doesnt help to understand conciousness except on a very rudimentary level10:37
Diablo-D3archels: I wonder where I fall on that10:39
-!- Filosofem is now known as Jawmare10:40
xrrmaaku: I guess an attacker can block any transaction from getting confirmations. That would be more difficult to detect10:43
maakuxrr: right10:43
xrrAlso, 51% power can mine 100% of blocks? Fun10:43
maakuyup10:43
maakubut that's detectable10:43
maakuthe censorship angle is a larger systematic risk10:44
maakuit explicitly defeats the entire purpose of bitcoin10:44
-!- augur [~augur@c-73-46-94-9.hsd1.fl.comcast.net] has quit [Read error: Connection reset by peer]10:45
Aurelius_Work2Hmmm10:45
Aurelius_Work2I wonder if I could get a job working on power BI10:45
-!- augur [~augur@c-73-46-94-9.hsd1.fl.comcast.net] has joined ##hplusroadmap10:46
gnusha_https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=f167c7a3 Bryan Bishop: fix two minor typos >> http://diyhpl.us/diyhpluswiki/transcripts/scalingbitcoin/hong-kong/bip99-and-uncontroversial-hard-forks/10:47
-!- yorick [~yorick@ip51cd0513.speed.planet.nl] has joined ##hplusroadmap10:53
-!- yorick [~yorick@ip51cd0513.speed.planet.nl] has quit [Changing host]10:53
-!- yorick [~yorick@oftn/member/yorick] has joined ##hplusroadmap10:53
-!- nmz787_i [~ntmccork@134.134.139.72] has joined ##hplusroadmap10:58
archelsAdrianG: it quasi-maps onto fundamental neurological functions10:58
archelsas such it is appropriate to the mind uploading field10:59
-!- amiller [~socrates1@unaffiliated/socrates1024] has quit [Ping timeout: 250 seconds]10:59
-!- atomical [~atomical@172.56.13.1] has joined ##hplusroadmap11:00
-!- Guest71119 [~socrates1@li175-104.members.linode.com] has joined ##hplusroadmap11:02
-!- nmz787_i1 [ntmccork@nat/intel/x-cbgccyhlzojixnfr] has joined ##hplusroadmap11:02
-!- nmz787_i [~ntmccork@134.134.139.72] has quit [Ping timeout: 260 seconds]11:03
-!- Filosofem [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap11:03
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has quit [Disconnected by services]11:04
-!- Filosofem is now known as Jawmare11:07
-!- erasmus [~erasmus@unaffiliated/erasmus] has quit [Ping timeout: 260 seconds]11:07
-!- nmz787_i1 [ntmccork@nat/intel/x-cbgccyhlzojixnfr] has quit [Quit: Leaving.]11:16
-!- ArturShaik [~ArturShai@195.114.249.170] has quit [Ping timeout: 272 seconds]11:18
-!- atomical [~atomical@172.56.13.1] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]11:30
-!- nmz787_i [ntmccork@nat/intel/x-enjghzjdpoxoujxs] has joined ##hplusroadmap11:35
nmz787_isup11:55
kanzureannoyed about lack of results docs uploaded to clinicaltrials.gov11:56
kanzureconfused about lendingclub claims11:57
nmz787_i/me imagines lendingclub has something to do with baby seals and rental agreements12:03
kanzurewell you're not wrong12:06
-!- Act [uid89656@gateway/web/irccloud.com/x-rnmndqvyhanazakw] has joined ##hplusroadmap12:07
AdrianGkiva default rates sound too good to be true.12:22
Diablo-D3heh, lending club is hilarious12:23
Diablo-D3literally, a bunch of dudes were like HEY PROSPER SOUNDS AWESOME, LETS COPY IT12:23
Diablo-D3and bam they legitimized prosper by cloning it12:23
AdrianGwhat tells you its a clone?12:25
AdrianGtheir scoring/risk management could be entirely different.12:25
Diablo-D3AdrianG: a company that crowd sources their loan funding12:26
Diablo-D3AdrianG: and they both use essentially the same risk management12:27
Diablo-D3both company with federal laws on loan origination12:27
AdrianGDiablo-D3: how do you know about the second part?12:27
Diablo-D3because?12:27
AdrianGrisk management isnt entirely transparent.12:27
Diablo-D3federal law on loans is a gigantic fucking tome12:27
Diablo-D3that no one understands but a special breed of lawyer12:28
Diablo-D3both of them have to meet the minimum of the law, and that minimum is very very very fucking high12:28
Diablo-D3both are lower than if a bank was playing with its own money12:28
Diablo-D3but its still pretty fucking up there12:28
AdrianGi have a hunch. there is probably a direct correlation - the thicker the tome, the more profitable is the economic area being regulated.12:29
Diablo-D3yup12:29
Diablo-D3and subcorralary to that12:29
Diablo-D3the thicker the tome, the more congressional fuckery going on12:29
Diablo-D3but that doesnt mean we shouldnt regulate the almighty fuck out of them12:30
Diablo-D3I mean, look at what one tiny change caused, the whole glass steagal fuckup12:30
Diablo-D3major banks completely fucked themselves and caused the US equivalent to the lost decade12:30
Diablo-D3Im all for major corporations being profitable, but they actually have to _be_ profitable12:31
Diablo-D3temporally local profits at the cost of total profits is not a winning strategy12:31
Alcyiusmaaku, while I can somehow see what you're going at, even people who aren't parents can still tell when someone is being a bad parent. Their thoughts on the matter can't be entirely discarded12:33
AdrianGmajor banks completely fucked up everyone else, and then took their money via treasury too12:33
AdrianGsmall correction ^12:33
AlcyiusThough, there is always a failsafe12:33
Diablo-D3AdrianG: well no, they fucked themselves12:33
Diablo-D3lets say they did it for personal profit12:33
Diablo-D3which they did12:33
Diablo-D3plus or minus12:33
AlcyiusFor instance, create something the AI enjoys more than the real world, and withold it unless the AI performs to our ethical standards12:33
Diablo-D3their wealth is now at a total negative of real value12:34
Diablo-D3because the dollar is devalued12:34
Diablo-D3so the difference between the rich and the poor is greater, ergo they are "richer"12:34
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has quit [Read error: Connection reset by peer]12:34
Diablo-D3but the actual purchasing power of their wealth is now lesser than it was12:34
Diablo-D3theft of money on such a large scale is largely a lose-lose situation12:34
nmz787_ideep UV LEDs are still around $200 a piece (in the mid 200nm range)12:36
nmz787_inot terrible for a high-cost instrument... but still killer for low-cost DNA analysis12:36
AdrianGnmz787_i: what power?12:36
kanzureconvince me that any of this is high-signal12:37
Diablo-D3nmz787_i: define deep12:38
Diablo-D3oh wait you did12:38
Diablo-D3sorry, my brain is fucked up today12:38
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap12:45
AlcyiusMy MRI got insurance approval12:46
AlcyiusFinally12:46
gnusha_https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=94355d2e Bryan Bishop: even more segwit video >> http://diyhpl.us/diyhpluswiki/transcripts/scalingbitcoin/hong-kong/segregated-witness-and-its-impact-on-scalability/12:48
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has joined ##hplusroadmap12:49
nmz787_iAdrianG: the last prices I got were for 5 mA or so12:50
nmz787_iso if you wanted to do a 260nm to 280nm absorbance ratio (which is the standard 'pretty good' routine analysis) you'd need two LEDs, so $400 plus few dollars for a photodiode and arduino clone or whatever, maybe a battery holder and a housing12:51
nmz787_i$500 for a fake nanodrop is still pretty reasonable12:52
nmz787_ibut a real nanodrop can do a whole spectrum, not just the two wavelengths12:52
kanzurehttp://i.kinja-img.com/gawker-media/image/upload/s---SBN_lgf--/c_scale,fl_progressive,q_80,w_800/1462685970574979110.jpg12:54
Diablo-D3kanzure: thats no zaku boy, no zaku!12:55
kanzurego away12:56
-!- fleshtheworld [~fleshthew@2602:306:cf0f:4c20:15a5:eb3f:cc25:62ea] has quit [Read error: Connection reset by peer]12:57
nmz787_ikanzure: is that a real thing from the new movie?13:03
nmz787_iby thing I mean droid13:03
kanzurehttps://pbs.twimg.com/media/CSSTdD5WwAEzy2G.jpg13:04
AdrianGnmz787_i: whats the cheapest FTIR setup ?13:04
Diablo-D3nmz787_i: no, someone put artoo's head on a zaku.13:05
-!- atomical_ [~atomical@172.56.13.1] has joined ##hplusroadmap13:13
-!- atomical [~atomical@209.153.22.182] has quit [Ping timeout: 256 seconds]13:17
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap13:20
-!- atomical_ [~atomical@172.56.13.1] has quit [Ping timeout: 256 seconds]13:23
nmz787_iAdrianG: not sure, the FTIR I have was $50... but it is unsable as-is... though it appears the electronics I did get are fine, I think I just don't have the control unit for it13:29
nmz787_iAdrianG: if you're talking DIY... they basically consist of a photodiode, a voice coil, a nice stable laser (HeNe gas is preferred, though maybe some solidstate are good enough today) and interferometer and or beam combiners, some mirrors, some salt windows... and the IR source (a hot glowing filament of wire)13:31
nmz787_iso you could probably rig one up for $500-1000 not including all the leg work13:31
nmz787_ithe salt windows might be the hardest to find, or DIY... the mirrors also need to be high quality, parabolic or curved, and gold coated I think13:32
nmz787_iand driving the voice coil probably needs to be pretty precise, or at least very repeatable13:32
nmz787_i(since that changes the path length, and creates harmonics)13:33
Diablo-D3woah, people build their own spectrometers?13:33
nmz787_iDiablo-D3: yes, here is one project of mine for UV VIS https://hackaday.io/project/1342-open-spectrometer13:33
nmz787_ithere is another person on hackaday.io that was working on raman13:33
nmz787_i(which I was aiming to produce a detector of sufficiently low noise to enable)13:34
nmz787_i(though we had no collaboration other than a few comments back and forth on their project page)13:34
-!- atomical_ [~atomical@172.56.13.1] has joined ##hplusroadmap13:34
-!- atomical [~atomical@209.153.22.182] has quit [Ping timeout: 276 seconds]13:38
-!- erasmus [~erasmus@unaffiliated/erasmus] has joined ##hplusroadmap13:45
AdrianGnmz787_i: you bought ftir for 50 bucks?13:48
nmz787_iyeah like 4 years ago from a recycler... mainly just to rip it apart... but when you power it on, the laser fires right up and the voice coil starts its reciprocating dance13:55
nmz787_iI was moving books from my garage to my office/electronics-lab this past weekend, and knocked into the FTIR which was under a tarp... didn't realize til the next day that I had a gash on my shin where i knocked into the device13:56
nmz787_iheavy ass beast13:56
-!- Filosofem [~Jawmare@unaffiliated/jawmare] has joined ##hplusroadmap13:58
-!- Jawmare [~Jawmare@unaffiliated/jawmare] has quit [Read error: Connection reset by peer]13:58
kanzureto mitigate "technical" risk in a venture capital funded project, why isn't the trend to give up more equity for the funding? probably there's a floor or ceiling for how much risk you can buy by that method, hm.13:59
nmz787_iequity in something that fails doesn't yield value, so maybe in terms of risk no amount of equity makes things better14:00
kanzurethere is value in equity in failed ventures- usually at least liquidation preferences regarding the physical assets and other stuff....14:00
-!- erasmus [~erasmus@unaffiliated/erasmus] has quit [Read error: Connection reset by peer]14:03
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap14:05
-!- atomica__ [~atomical@172.56.13.1] has joined ##hplusroadmap14:07
-!- erasmus [~erasmus@unaffiliated/erasmus] has joined ##hplusroadmap14:07
-!- atomical_ [~atomical@172.56.13.1] has quit [Ping timeout: 264 seconds]14:09
-!- atomical [~atomical@209.153.22.182] has quit [Ping timeout: 260 seconds]14:11
-!- atomica__ [~atomical@172.56.13.1] has quit [Ping timeout: 276 seconds]14:15
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap14:20
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 240 seconds]14:51
-!- sandeepkr [~sandeepkr@111.235.64.4] has quit [Quit: Leaving]14:57
-!- Madars [~null@unaffiliated/madars] has quit [Quit: Leaving.]15:01
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap15:03
-!- triggerwarning [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has quit [Changing host]15:30
-!- triggerwarning [~IGLC@unaffiliated/triggerwarning] has joined ##hplusroadmap15:30
kanzure"Can financial engineering cure cancer?" http://lfe.mit.edu/wp-content/uploads/2015/08/CanFECureCancer2013.pdf15:31
-!- jcluck [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has joined ##hplusroadmap15:32
xentracthere was a girl a few years back who built a raman spectrometer for her high school science fair, nmz787_i15:33
-!- cluckj [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has quit [Ping timeout: 276 seconds]15:33
kanzure"Commercializing biomedical research through securitization techniques" http://lfe.mit.edu/wp-content/uploads/2015/08/nbt2012.pdf15:34
xentracnmz787_i: interesting the choice of the Propeller to get deterministic timing easily15:34
xentracI also didn't know about hackaday.io.  thanks!15:36
-!- jcluck [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has quit [Ping timeout: 256 seconds]15:37
-!- cluckj [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has joined ##hplusroadmap15:38
xentrachow much did the TCD1304AP cost?  digi-key doesn't list it15:38
-!- jcluck [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has joined ##hplusroadmap15:39
-!- Jonathan [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has joined ##hplusroadmap15:40
xentracheh, "With increased CAD and computational ability, optical design has been able to take the mirrors and the grating and merge them into one optical element."  that one optical element sounds challenging to fabricate?15:41
-!- andytoshi [~andytoshi@unaffiliated/andytoshi] has quit [Ping timeout: 260 seconds]15:41
Diablo-D3Sounds like they're just 3D printing the object at least partially using computer aided problem solving15:42
-!- cluckj [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has quit [Ping timeout: 256 seconds]15:42
-!- jcluck [~cluckj@pool-108-16-231-242.phlapa.fios.verizon.net] has quit [Ping timeout: 256 seconds]15:43
xentracDiablo-D3: we are not in the Diamond Age and therefore you cannot "3D print" optics15:44
xentracyet15:44
Diablo-D3I dunno15:45
Diablo-D3you might actually be able to15:45
xentracessentially all of the instrument-grade diffraction gratings in the world come from the Grating Lab, which cuts masters in glass with three grating engines, one built by the hands of Michelson himself15:45
Diablo-D3ahh, theres that15:45
Diablo-D3that sounds like that entire industry is...15:45
xentracsomeday, yes; today, no15:45
* Diablo-D3 puts on sunglasses15:45
xentracyour humor is not adding to the signal level in this channel and is likely to result in you being banned15:46
kanzurecorrect15:46
Diablo-D3fine =/15:46
Diablo-D3I couldn't think of a good pun anyways15:47
xentracthe actual gratings that are distributed are molded from the glass masters through several intermediate stages15:47
xentracthe process is very demanding because it must preserve the surface detail of the grating down to a small fraction of the wavelength of light15:47
Diablo-D3xentrac: okay so....15:48
Diablo-D3its basically integrated circuit manufacturing15:48
xentracthe necessary capabilities have a lot in common with IC manufacturing, but the process is completely different15:48
Diablo-D3well, I was thinking along the lines of how features are now measured in low single digit nanometers15:49
Diablo-D3and UV lithography is basically boned because features are now smaller than the wavelength of the UV light they're using15:51
kanzurexentrac: let me know when i should ban him.15:51
xentrackanzure: I was trying to prevent him from being banned by giving him an extra warning so he could correct himself15:53
Diablo-D3?15:53
Diablo-D3So you're going to ban me for talking about nano-scale fabrication techniques?15:55
Diablo-D3Especially since, apparently, they use/will be using EULV lithography for diffraction grating manufacturing?15:56
xentracnow that sounds interesting15:56
-!- Act [uid89656@gateway/web/irccloud.com/x-rnmndqvyhanazakw] has quit [Quit: Connection closed for inactivity]15:57
xentracI haven't heard of it though15:57
xentraclink?15:57
Diablo-D3https://en.wikipedia.org/wiki/Nanolithography15:58
Diablo-D3surprisingly there15:58
Diablo-D3it doesn't have a ref though, so meh15:58
-!- Houshalter [~Houshalte@oh-71-50-56-224.dhcp.embarqhsd.net] has joined ##hplusroadmap15:59
-!- drewbot_ [~cinch@ec2-54-204-89-58.compute-1.amazonaws.com] has quit [Remote host closed the connection]16:00
-!- drewbot [~cinch@ec2-54-163-118-160.compute-1.amazonaws.com] has joined ##hplusroadmap16:01
Diablo-D3https://en.wikipedia.org/wiki/Electron-beam_lithography16:01
Diablo-D3now that is pretty interesting16:01
Diablo-D3xentrac: so those diffraction grates you're interested in are for IR?16:04
-!- nmz787_i [ntmccork@nat/intel/x-enjghzjdpoxoujxs] has quit [Quit: Leaving.]16:05
Diablo-D3why is that the only company that produces them?16:05
Diablo-D3it looks like something they could repurpose old semiconductor equipment for, doesn't it?16:06
xentracDiablo-D3: the relevant criteria are "bandwidth" and "signal-to-noise ratio".  we all probably already know that UV lithography is running into problems due to wavelength, for example, and that the current process node is 14nm (not yet low-single-digit nanometers)16:06
xentracso saying those things is not super endearing :)16:07
Diablo-D3Okay? Normally its considered rude to start talking about such things without context.16:09
xentracRichardson Gratings (the current owner of the Grating Lab) is the only company that produces research-grade gratings because reproducing more gratings from the master is very cheap, but making the master is very difficult16:09
xentracoh, I'm answering this: 23:55 < Diablo-D3> So you're going to ban me for talking about nano-scale fabrication techniques?16:09
kanzurethose megafunds would probably have better results by just buying the fda itself, or influencing the political process to make drug approval cheaper overall. once you accumulate $30 billion, a bunch of options open up.16:10
Diablo-D3Yes, and threatening to ban people for trying to understand the problem by comparing it to other problems also isn't good for the snr.16:10
kanzurealthough getting new laws passed costs way less than $30 billion16:10
xentracthere is a company doing specialty gratings with a scanning-beam process: http://www.plymouthgrating.com/Products/Products%20Page.htm16:11
xentracbut I still want to know how nmz787 is planning to produce the optics for their spectrophotometer, because if they have a better way, that's super awesome16:11
xentracI mean, it suggests that there are important manufacturing capabilities that I don't know about, which is a really important thing for me to find out about :)16:12
xentracthe PGL results I linked above are fairly crude by some measures compared to the blazing achieved by Richardson16:14
Diablo-D3xentrac: well, Im pretty sure nmz787 can't do IR plates in his basement and/or garage... so yeah, I'm wondering too16:16
xentracyou might be surprised at what nmz787 can do in his basement ;)16:18
Alcyius*sigh*16:18
AlcyiusWhy16:18
xentracthis is the nearest semiconductor manufacturing process to the process used for gratings: https://en.wikipedia.org/wiki/Nanoimprint_lithography16:21
-!- Act [uid89656@gateway/web/irccloud.com/x-ikhmcdeneaizkpvg] has joined ##hplusroadmap16:28
-!- nmz787_i [ntmccork@nat/intel/x-gtayaabfmrxxxrsr] has joined ##hplusroadmap16:34
kanzuremegafund proposal relies on an awfully large amount of intellectual property law ("research-backed obligations").16:38
xentracit would probably work better to start a research cult16:40
xentraconly problem is that it would be banned in China16:40
kanzurexentrac: http://lfe.mit.edu/wp-content/uploads/2015/08/nbt2012.pdf this one16:40
kanzurei am not particularly impressed by existing answers to high-risk project financing... "high-risk" is sorta handwavy too; was the laser "high risk"?16:41
xentracright16:41
xentrachow big is Harvard's endowment?16:42
kanzure36 billion16:43
xentracso basically they're proposing a university funded by its tech transfer office?16:44
xentracor does Harvard not issue bonds?16:44
nmz787_ixentrac: TCD1304 series are about $10, easily had on aliexpress and ebay... I think there is some licensing thing that prevents their sale on digikey or something, I'm not too sure... I once heard they might have been discontinued and all sales were of old-stock... but that was years ago and hard to believe since they're still just as easily available (and also there is no replacement/substitute product)16:44
nmz787_ixentrac: 3D printed optics has been demonstrated... not great, but OK16:45
xentracnmz787_i: really?  including gratings?16:45
xentrac$10 is pretty amazing. we are living in the future16:46
kanzurexentrac: they are proposing a megafund structure where the bonds have a high chance of high yield by funding ~150 pre-clinical-trial products.16:46
xentracyes, so I see16:47
xentracalthough "high chance of high yield" is almost a contradiction ;)16:47
kanzurewell, i mean, high chance of 10-12% yield16:47
nmz787_ixentrac: http://hackaday.com/2014/12/13/3d-printed-lenses-open-up-possibilities/16:48
nmz787_iI think that is the one I was thinking of16:48
nmz787_ialso e-beam lithography is old news16:48
nmz787_ias is ion beam litho16:48
xentracwell, it still might turn out to be useful16:48
kanzureeven if harvard endowment fund did issue regular bonds or something, not sure how to answer your question16:49
xentrac"all lenses require some form of polishing to become optically clear. It was printed with a 50 micron resolution"16:49
nmz787_iand while many gratings come out of Richardson (now absorbed by.... Edmund optics I think) there are other (read non-US) suppliers/manufacturers16:49
xentracso that's about three orders of magnitude too coarse to print a grating with16:49
nmz787_iright, that's why the post is title "lenses"16:49
nmz787_inot gratings ;)16:49
xentrac(or nine orders if you're counting voxels)16:49
nmz787_ibut gratings have been made DIY too16:49
xentracyeah, but I don't think the other suppliers have a significant fraction of the research market, do they?  even overseas16:50
xentracindeed16:50
nmz787_ihttp://diyhpl.us/~bryan/papers2/optics/photolithography/High%20resolution,%20low%20cost%20laser%20lithography%20using%20a%20Blu-ray%20optical%20head%20assembly.pdf16:50
xentracI mean, one of the three ruling engines in the Grating Lab was a Do It Yourself by Michelson in 191016:50
nmz787_ithat isn't really sub-micron... but gratings nonetheless16:50
xentrac450nm feature size could work if you were only doing infra4red16:51
xentracbut for infrared spectroscopy everyone does FTIR anyway16:51
nmz787_ixentrac: well gratings/grisms etc are far from consumer items... so the market isn't very transparent or open16:51
kanzurexentrac: i would say one of the differences is that harvard isn't chasing yield16:51
kanzureand harvard isn't chasing commercialization16:52
xentrackanzure: I am sure that Harvard's administrators are quite enthusiastic about maximizing the future size of their endowment16:52
xentracbut the bonds are a big difference16:53
kanzurethe administrators might be, but as an institution they are not studiously applying their knowledge towards optimizing themselves as a technology development and infrastructure deployment superleague.16:53
kanzurewhy big difference?16:53
xentracalso, Harvard is more careful about its survival than the yield16:53
kanzureif their endowment is 36 billion then they probably only have a few billion/year to play with16:54
xentracbecause if you can actually get pension funds to accept 3% yields on your research-backed bonds, you can bring to bear enormous amounts of money on the problem16:55
xentracwhatever the problem is16:55
xentracand that's why Elon Musk owns the majority of the rooftop solar panels in California16:55
kanzurethese papers are looking at 10-12% returns16:55
kanzureover 10-15 year period16:55
kanzure(for $30 billion-ish. does not seem to be accessible for lower amounts of capital)16:56
Diablo-D3Okay, so, what exactly are you doing with these?16:56
xentracyeah, I know they're proposing higher returns16:56
Diablo-D3the FTIR spectrometers16:56
xentracbut you aren't going to get pension funds to invest in bonds at 10%16:56
kanzurexentrac: whynot?16:56
Diablo-D3uh, what bond is going to be 100% failproof and pay 10%?16:57
xentracyields and risks are inversely related16:57
Diablo-D3exactly16:57
kanzurehave you read the paper?16:57
xentracI haven't finished it yet16:57
kanzurethey are aiming for a default probability of 0.4%16:58
Diablo-D3this is why most funds like this just invest in government bonds, most of them federal16:58
xentracbut the general idea is that if your bonds are really low-risk, you won't have to pay 10% to get people to buy them16:58
xentracand if you are paying 10% to get people to buy them, then they will conclude that they aren't really low-risk16:58
Diablo-D3if the US government defaults, well, we have muuuuch larger problems to deal with than a couple kids going to college16:58
-!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-cmxmsdplanbkppno] has joined ##hplusroadmap16:58
kanzure"This default probability is comparable to the historical realized 10-year default rates of the highest-rated category of debt (Aaa) from 1920 to 2010, according to the bond-rating agency Moody’s16:58
kanzure[25]."16:58
xentracand if they are wrong about that risk, then someone else will do the same thing and make a lot more money by paying 3% on similar bonds16:59
Diablo-D3btw, why is the fund trying to manage that risk itself?16:59
Diablo-D3isn't that a waste of the fund's time and money?16:59
xentracDiablo-D3: I am reading http://lfe.mit.edu/wp-content/uploads/2015/08/nbt2012.pdf17:01
xentracthe first five or so pages are about that17:01
Diablo-D3So this is about, what, forming a fund that just uses low risk investments to fund research?17:02
Diablo-D3Oh wait, no17:02
Diablo-D3Its about making a portfolio of projects to reduce risk17:02
Diablo-D3That is more of the silicon valley angel investor shotgun effect than I'd like to see17:03
Diablo-D3xentrac: I added it to my read list17:03
kanzurewell, idea is to see if there is a feasible way of funding large-scale high-risk projects like orbital telescopes, dyson spheres, large-scale cryonics facilities, whole brain emulation work, brain uploading depositories and scanning tech, etc.17:04
kanzureif we are having such significant problems with basic medicine then even highly-likely-to-eventually-be-feasible things like dyson spheres are going to be even more painful to make happen17:05
Diablo-D3Well, the problem with that is most people are simply not well enough educated to handle such concepts17:05
Diablo-D3And I'm saying that from personal experience17:05
xentrackanzure: interestingly this paper mentions MIT raised US$0.75B in 100-year bonds in 201117:05
kanzuregah please go the fuck away17:05
xentracat 5.6%17:05
kanzureDiablo-D3: you are very very low-signal and i would like you to leave17:05
Diablo-D3What the hell man17:06
Diablo-D3You can't just say everything you disagree with is low signal17:06
kanzurefuck you for assuming i disagree with any of that shit, wtf17:06
kanzureno i want you to leave because you are low-signal, not because i fucking disagre3e17:06
Diablo-D3Okay, so, how are you going to convince investors to invest into moonshots when reasonably average people won't even buy electric cars17:07
Diablo-D3That isn't "low signal", that is a relevant problem17:07
justanotheruserwhat17:07
Diablo-D3What the paper xentrac linked to seems to more cover "what to do with the money, and how to best protect it against individual project failure" and less about actually getting money to invest into such projects in the first place17:08
kanzurethat is not a relevant problem; it's not even a problem.17:08
nmz787_itbh I didn't understand why xentrac mentioned in the first place that Diablo-D3 was low signal... I mean I thought the joke about '... day' and "puts on sunglasses" was funny and not noisy... if anything it made the subject interesting because it was humourous17:09
kanzureno he has been here for days smelling up the place17:09
nmz787_iDiablo-D3: FTIR is for doing chemical analysis17:09
kanzurexentrac: i wonder why they wanted to do 100 year bonds17:10
Diablo-D3kanzure: I feel like I know you from somewhere else17:10
xentrackanzure: this paper's "megafunds" is sounding more and more like a higher-risk version of a research university17:10
kanzurexentrac: but again though, harvard isn't focused on tech development, engineering and deployment. and yet they are a research university. so i don't think the comparison works out?17:11
xentracwell, that isn't all they do, but they do do that17:11
Diablo-D3Harvard does make money off of some of the things they have funded research into17:11
xentracI think the bigger issue is that 0.4% risk of failure over 10 years is a 25-year MTBF, which is pretty short for a research university17:12
kanzureDiablo-D3: yes we all know about tech transfer offices. please, please go away.17:12
Diablo-D3kanzure: look, I'm normally a nice guy, but if you're going to keep doing that, I'm going to put you on ignore17:12
kanzurei don't care if you are nice. i'll ban your ass.17:12
xentracand of course Harvard gets more funding from its alumni than from its tech transfer office, rigght?17:12
-!- kanzure [~kanzure@bryan.fairlystable.org] has quit [Changing host]17:13
-!- kanzure [~kanzure@unaffiliated/kanzure] has joined ##hplusroadmap17:13
Diablo-D3Yes, and continually interrupting the conversation about how you don't like what I'm saying is not really helping the snr is it?17:13
-!- mode/##hplusroadmap [+o kanzure] by ChanServ17:13
-!- mode/##hplusroadmap [+q Diablo-D3!*@*] by kanzure17:13
-!- mode/##hplusroadmap [-o kanzure] by kanzure17:13
kanzurexentrac: i have no idea if that's the case or not17:14
Jonathanhah17:14
-!- Jonathan is now known as cluckj17:14
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving]17:15
xentrac(harvard may not be the best university to look at, but clearly the proposed "megafunds" are more like Harvard or MIT than they are like Antioch College)17:16
xentractotal cash receipts from giving (mostly from alumni I suppose) in 2015 were US$1B: http://finance.harvard.edu/files/fad/files/harvard_financialreport_fy15.pdf?m=144612434717:17
kanzuretech transfer office surely pulls more revenue than $1B ? maybe not......17:17
xentracI was thinking "surely less"17:18
kanzureother issue is that in hyper-competitive environment of low-risk investment opportunities at many scales and magnitude, high-risk probably never has a chance. government funding doesn't really solve this either.17:18
xentracbut so far I am not finding any mention of their tech transfer revenues in this report at all17:19
kanzurewhy would you go for surely less ? i mean maybe because harvard, but iirc stanford for example makes a buttload from patents.17:19
kanzurea dirty nasty buttload17:19
xentracI don't think it's that large of a buttload though17:19
xentracand Stanford itself is a bit of an outlier there, much as is Harvard on endowment size17:19
nmz787_iDiablo-D3: I agree that kanzure's comments about snr actually degrade it17:20
kanzurewell i mean we should be okay with hand-waving about a slightly imaginary university that has similar traits17:20
xentracsorry for contributing to that, nmz787_i17:20
kanzurepage 8 http://www1.hw.ac.uk/kescotland/media/Technology%20Transfer%20at%20Stanford.pdf17:20
nmz787_ibut I guess I am not here lately enough to smell 'the stink' as he put it17:20
kanzure"$100 billion dollars in revenue or approximately half of the Silicon Valley revenue is spin-off from Stanford University"17:20
xentracone thing that may not be captured by these figures is that Stanford (and maybe Harvard) may demand equity in spinoffs, rather than direct payment17:21
xentracsure, but how much of that value do they capture?17:21
kanzureoh hell17:21
kanzurehm17:21
xentracDiablo-D3 points out http://www.statnews.com/2015/12/21/harvard-faculty-scientific-research/ which says US$13M annually17:22
xentracanyway, so the "megafunds" idea is kind of to run a biotech-focused research university entirely off its securitized tech transfer revenues with an MTBF of about 25 years?17:23
xentracno, what am I saying?17:23
xentrac0.4% / 10 years is .04% per year, or 2500 years MTBF, which to me sounds like somebody's been hitting the crack pipe17:24
xentracain't nobody got a 2500-year-MTBF organization out there.  not the Pope, not the Japanese Emperor, nobody.17:24
xentracbut that's okay if we're just talking about 0.4% for the lowest-risk tranche or something?17:26
nmz787_iwait, isnt the catholic church pretty close to 2500 years?17:28
nmz787_ithey're the most successful business when viewed in a certain way17:29
xentracit's close, but do you really think its likely lifespan from today is *another* 2500 years?17:29
xentracassuming no "institutional wearout", which may be unreasonable17:29
xentracI mean there were a lot of similar institutions around when it was born, most of which have since been destroyed17:30
xentracso you have to correct for survivorship bias17:30
xentracI don't think we can count Saudi Arabia as a survival of the Nestorian church, for example17:30
nmz787_iI can't find the article I was thinking of17:31
xentrac(also, they lost most of their assets a few times, which would have been a problem if they had securitized those assets to investors...)17:31
nmz787_ithe article was more of a comparison17:32
nmz787_italking about steady growth (using membership instead of revenue)17:32
nmz787_ibut that basically growth has been consistent for like 2000 years17:32
xentracI can believe that.  and Pope Bergoglio seems like he might correct its recent decades of falling below the trend line17:33
xentracbut what about the Nestorians and the Byzantines?17:34
xentracor for that matter the Zoroastrians and the Mithraists17:34
nmz787_iyou're speaking "foreigner" to me17:36
nmz787_iI suck at history pretty much17:36
nmz787_iI am pretty sure the non-US guy I work with knows more about US presidents than I do17:37
cluckjalso jews17:38
cluckjbeen around for kind of a while17:38
cluckjthe catholic church isn't 2500 year MTBF either, they've had a few17:41
xentracJudaism still exists as a religion, but not a church in the sense of an institution that can hold assets17:42
xentracthis segment about drug-royalty investment companies is really interesting; I didn't know these existed17:43
-!- Gurkenglas [Gurkenglas@dslb-188-103-077-131.188.103.pools.vodafone-ip.de] has quit [Ping timeout: 250 seconds]17:44
-!- AmbulatoryCortex [~Ambulator@173-31-155-69.client.mchsi.com] has joined ##hplusroadmap17:44
cluckjhmm17:45
xentracit's strange that they think that Royalty Pharma invested US$5.8 billion over 8 years and yet their proposed "megafund" could be viable with much higher-risk assets with only US$35B17:45
xentracyou'd think it would be more like a factor of 10 difference17:45
cluckjhttps://en.wikipedia.org/wiki/List_of_oldest_companies17:47
xentracif we take "companies" in general as our reference population, I think the MTBF is about 30 years17:48
xentracbut the hazard function is far from constant, with much higher risk in the first few years17:48
-!- nmz787_i [ntmccork@nat/intel/x-gtayaabfmrxxxrsr] has quit [Ping timeout: 240 seconds]17:49
kanzurewhat is origin of interest in 2500-year-MTBF or MTBF?17:49
xentrac0.4% risk of default over 10 years implies MTBF ≥ 2500y17:50
cluckjrisk calculations? which are a recent thing17:50
xentracp.1117:50
kanzureyea i suppose more generally, diversification is a lousy risk mitigation scheme. ultimately you have to mitigate actual risks through methods other than weird financialization.... like actual study, engineering, hard work, that sort of thing.17:53
kanzurewhich... uh, requires funding.17:53
kanzurewait i think i'm stuck in an infinite loop17:54
-!- atomical [~atomical@209.153.22.182] has quit [Max SendQ exceeded]17:54
xentraca more interesting question is whether diversification might *increase* risk17:56
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap17:56
xentracI mean essentially the researchers at the "megafund" would be employees of a largish company (thousands of employees)17:56
kanzuredepends on what diversification really means--- ultimately many engineering problems have similar underlying principles or solutions17:56
xentracthe organizational techniques to make such large organizations manageable are somewhat at odds with the needs of innovation17:57
kanzurethere has not been much experimentation with top-down strategic planning for scientific achievement17:57
kanzurethere are very few research orgs that have tried this; myelin repair foundation is one of the more prominent examples that comes to mind.17:57
kanzuresens foundation is another example sorta- although they sorta suck at strategy and funding compared to myelin repair foundation i think? not sure.17:58
cluckjbig government research projects?17:58
kanzurehttp://diyhpl.us/wiki/transcripts/open-science-summit-2010/scott-johnson-myelin-repair-foundation/17:58
kanzure"The reason why we show these circles.. is that basic science is pretty random. There's no research plan. No disease organization actually has a research plan. They put out a request for proposals, they peer review those, and those that get the highest ratings, that's what gets funded. Most academics would say that it should be random, since you can't know where they are going. Well, in some cases you want to be outcome directed"17:59
kanzureso in particular i mean the random grant submission process heh17:59
kanzureby picking a strategy you can focus on some very particular questions and hypotheses and work towards either validating some models or rejecting the entire premise, as quickly as possible18:00
kanzurethis requires you to be good at speculation and theory and science and knowledge, but that does not seem like an overly burdensome requirement.18:01
cluckjit is, seemingly, all about the benjamins18:01
xentrac01:57 < kanzure> there has not been much experimentation with top-down strategic planning for scientific achievement18:02
kanzuredarpa tends to have some strategy, though. so that's an interesting place to look.18:02
xentracseriously?  what about DARPA, SBIR grants, the Manhattan project...?18:02
cluckj^18:02
kanzureyes yes i just said darpa, cool your hyperjets :P18:02
xentracI was typing too slow!18:02
kanzuresbir grants though- not sure why you bring those up here?18:02
cluckjthose are outcome-based though?18:02
cluckjDARPA is like "we need a killer robot to do X"18:02
kanzureno darpa also does broader stuff18:03
cluckjor "we need a big-ass bomb"18:03
xentraca lot of SBIR grant RFPs are for scientific achievements18:03
kanzurelike they did something approximating "make me 'sudo make me a sandwhich'" (their ifab initiative)18:03
kanzureand they have their 100 year colonization starship thingy18:03
xentracthe problem is that most of that stuff is kind of incremental18:03
kanzurewhy is that problematic?18:03
cluckjsomething that is coincidentally outcome-based, but is in practice very focused on basic research?18:03
xentracI mean, the outcomes of scientific discoveries 10 or 20 years out are really hard to predict18:04
xentracso it's hard to top-down plan what you need to research *today* with objectives half a century out18:04
kanzurewell yes but- i mean, presumably you don't focus on magical thinking, you focus on good speculation with good assumptions and good mental arithmetic18:04
xentracright, but that may not actually work18:04
kanzurei think that's what post mortems are for, "oops we forgot that we needed a null hypothesis"18:05
xentracI mean you could have argued that the Correspondence Principle showed that lasers were impossible18:05
kanzure"You want to prove the total task will work even if the subtasks fail, and otherwise abandon it. Then you want to prove each subtask is impossible, and replace it appropriately and re-plan integration as quickly as possible (etc etc). It's not as big a deal to structure things perfectly if you have infinite resources and can parallelize everything, which is how the space shuttle and particle colliders are built."18:06
xentraceven if you knew that the confocal microscope would be a really great way to use lasers if they could exist18:06
kanzure"The big danger is doing the non-failfast steps first with one person. If one component has a major problem, that means one node is unexpectedly big. In practice, people replace that component with another component rather than delay, or engineer around it, or just accept the delay. But the overall delay is not due to delay along a specific path--it's due to multiple delays, some on every critical path."18:06
kanzure("with one person" as in the person architecting and implementing the damn thing (needs to look at the actual details and make sure there's no large splotches of infeasibility))18:07
cluckjthe upcoming scientific funding paradigm is going to be changing from big institutional science to individual entrepreneurship, which might be an improvement18:07
xentracbut there's no way that you would have predicted in 1952 that lasers would lead to supermarket barcode scanners, CD-ROMs, high-bandwidth fiber-optic communication, and consumer-level laser-cut wood18:07
kanzurein the 1920s there was someone that wanted a telescope-based optical internet based on libraries that had lots of mirrors and mechanical switches18:08
kanzurepneumatic relay networks were around since forever18:08
xentracor for that matter atomic clocks18:08
cluckjnow we have fiber networks and DLPs18:08
xentrac(and GPS...)18:08
xentracsure, optical long-distance communication has a history spanning millennia18:09
xentrac.wik heliograph18:09
yoleaux"A heliograph (Greek: Ἥλιος helios, meaning "sun", and γραφειν graphein, meaning "write") is a wireless solar telegraph that signals by flashes of sunlight (generally using Morse code) reflected by a mirror. The flashes are produced by momentarily pivoting the mirror, or by interrupting the beam with a shutter." — https://en.wikipedia.org/wiki/Heliograph18:09
kanzurei'm not sure if you would have needed to assume or know about lasers though18:09
xentracno, what I'm saying is that even if you could know that lasers were *possible* in 1952, instead of being tricked by a naive reading of the Correspondence Principle into labeling them as "magical thinking"18:10
kanzurenon-laser alternatives seem plausible18:10
kanzurewhat's so bad about avoiding lasers?18:10
xentracyou wouldn't have been able to predict the effects of actually achieving lasers18:10
kanzurepresumably lasers would still happen (just not because of you :p)18:10
xentracsure, possibly18:10
kanzureok so you mean trying to come up with arguments for working on lasers18:11
xentracalthough if you take that perspective, in the limit we should not worry about anything and just let other people do all the work18:11
xentracbut I'm not saying avoiding lasers is bad18:11
kanzurei think that framing of the problem is sorta wrong though-- like, "how do we fund the development of brain uploading/scanning/whatever" is a much different question than "hmm how do we commercialize this theoretical physics result?".18:11
xentracI'm saying that it's hard to plan scientific achievements top-down because you don't know in advance what's possible and what's not, and for that matter what will happen18:12
kanzurei think you can have very strong boundaries around what you know is possible18:12
kanzurethere's a lot that we know is possible, we just don't have exact answers on how to get there, but we can definitely be systematic in our approach for identifying reliable paths or routes18:12
xentracsure, in the late 1800s Kelvin showed that it was impossible for the Sun to be more than a few million years old18:12
xentracI agree about the brain-uploading question --- we know some of the things we will need to figure out18:13
kanzurethat's not a helpful example- the analogy is sorta useless to me, how am i supposed to make a tech thingy that requires the sun to be less than a million years old...18:13
* cluckj eats his Thomas Kuhn book, whole18:13
xentracheh18:13
kanzureor s/supposed to make/supposed to imagine18:13
xentracwell, a working tokamak or even Farnsworth fusor requires the "impossible" things to happen that allowed the sun to be more than a few million years old18:14
kanzurebtw what was his reasoning for impossibility18:14
xentrache knew the sun's mass from the period of the Earth's orbit, the Earth's mass, and the gravitational constant18:15
xentracso he calculated the potential energy of that mass falling in from infinity, which was surely where the Sun's energy output came from, and its output18:16
kanzureso he was looking at sun mass burn rate?18:16
xentracwell, he didn't know of a mechanism by which the sun could be burning mass18:16
xentracthat wasn't discovered for several decades more18:16
xentrache knew the answer was wrong, FWIW18:16
kanzurebecause astronomy and stars and light years more than a few million?18:17
xentracbecause it was known that the Earth was several billion years old and had been populated by plants for a good fraction of that18:17
kanzureoh that is a less fun reason that he knew he was wrong. oh well.18:17
xentracno, actually the Milky Way was all that was known of the universe at the time, and it's pretty small18:17
xentracthey didn't realize there were other galaxies18:17
xentraccf. Kuhn above18:18
xentracanyway so we can work on the problems we think will help with brain uploading18:18
xentracbut it might turn out that the things that actually help most are the ones we can't yet tell will help18:18
kanzureso actually i am not so interested in brain uploading of any particular individual..... i'd be fine with very rough approximation that seems to boot up anything between bonobo-level general cognitive ability and human-level general cognitive ability. the specific memories i'm much less interested in.18:19
xentracwhich means that a non-top-down-planned invisible college would get to them first18:19
cluckjthere are some that help a lot, and predictably, like refrigeration18:19
xentracyes18:19
cluckjor vaccinations18:19
xentracthe best modern example is probably the Japanese Fifth Generation Computing Project18:19
cluckjor (currently) antibiotics18:19
cluckjyeah, fuckin' computers too18:19
xentracwhere top-down planning prevented Japan from achieving much of anything in computing in the 1980s18:20
kanzuremaybe they just had poor planners though?18:20
kanzureget fucking john von neumann in charge of this shit, damn18:20
xentracsure18:20
xentracheh18:20
xentracI don't know if you heard but the good doctor is dead18:21
xentracantibiotics is a good example of the uncertainty: we have this new family of antibiotics that might turn out to work well and last long enough to allow us to design new ones18:21
kanzurethere's lots of alternatives in antibiotics land; they just take a while to get worked on. igem keeps unturning new ways to do antibiotics. i wouldn't worry aobut that.18:21
cluckjantibiotics is a great example too, because there's more than one 'paradigm' to kill bacteria18:21
xentracbut it might turn out that bacteria develop resistance to them faster than we expect and that phage therapy or even diamondoid nanobots would have been a better approach18:22
kanzurenah there's lots of stuff you can do to make sure that resistance is not easily-achievable.18:22
cluckjbacteriophages are one way too, and that approach died with the soviet union18:22
xentracor I don't know some kind of immune supplementation18:22
xentracI think there are stll Russian clinics that do phage therapy, cluckj18:22
kanzurebacteria have high parallelism but they have zero ability to surpass super-large energy gaps on the evolutionary landscape18:22
cluckjI think you can still buy them OTC in russia18:22
xentrackanzure: that seems likely to be correct, but bacteria keep surprising us18:23
kanzurewhen have they surprised us18:24
kanzureantibiotic resistance has not been surprising18:24
xentrachorizontal gene transfer was discovered in 1951 but it wasn't until the 1990s, for example, that it was recognized as a major driving force of bacterial evolutio18:25
kanzurewhat?18:25
kanzureso they put the estimate an order of magnitude too low. so what?18:25
xentracand we keep finding bacterial species that are progressively more extremophilic, surpassing previously hypothesized limits on extremophilia18:26
xentracalthough fortunately those are rarely pathogenic18:26
cluckjhaha18:26
cluckjTaq polymerase18:26
kanzurefusion answer was a good answer earlier18:26
cluckjspeaking of unintended outcomes of extremophile research18:26
xentracyes, that's a good answer too18:27
kanzurebuuuut i don't think kelvin was trying to build a nuclear reactor18:27
cluckjit's an unknown unknown thing18:27
xentracno, just like I'm not trying to build an antigravity or time-travel or FTL machine18:28
kanzureengineering shows us that planning actually works and we can achieve results, even when we don't know the exact format of all of the components (such as in software, where you don't have full clarity on every single detail at the very beginning).18:28
kanzurewell we know how to build a time traveling machine damn it18:28
xentrac(although Kelvin was a great scientist and I am just a punter)18:28
cluckjjust gotta go FTL18:28
kanzureno you don't need faster-than-light for time travel, sigh18:28
cluckjthere are probably easier ways of time traveling, but we can't even think of what they are because of our current understanding of physics18:29
cluckjI won't say epistemology because 'philosophy'18:30
kanzurehuh? no, they are known.18:30
kanzurethere are at least two widely known ways to do forward-only time travel18:30
kanzurethree, if you are okay with the slow method joke answer18:30
xentracsure, I meant backward time travel :)18:30
cluckj^18:30
kanzurecheaters.18:30
xentracI think talking about "forward-only time travel" in the context of "time travel" also kind of qualifies as a "joke answer" :)18:30
cluckjlol18:31
kanzuremaybe i'm a time clown18:31
kanzureanyway i still tihnk that planning and strategy seems to work18:33
kanzureand focusing on correctness seems to help (i really do think john von neumann would have conducted a better research strategy)18:33
xentracyeah, I think focusing on correctness and focusing a good part of our effort on things that are strategically important is a good idea18:35
kanzureit's a shame that we have to bother to actually say such generic obvious things18:35
kanzurewell i guess they are not always obvious18:35
cluckjI agree it's important too18:36
cluckjI think finding ways to increase and foster serendipity is better :)18:36
xentracI think we need both18:38
cluckj^18:38
xentracwhich I think is not obvious to kanzure :)18:38
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has quit [Ping timeout: 240 seconds]18:39
cluckjthe latest scientific knowledge production regime is about focusing on expanding current understandings in order to profit from them18:39
kanzurenot really... it's very poorly structured for that.18:40
xentracfortunately the Invisible College of Hackademia is here to fill the gap :)18:40
cluckjhah18:41
cluckjit's not so invisible anymore18:41
kanzurehackademia sounds like a rare blood disease18:41
-!- yashgaroth [~yashgarot@2602:306:35fa:d500:f5e0:f867:a11d:8d52] has joined ##hplusroadmap18:41
kanzurei hope there's a treatment18:41
kanzureiirc not even the soviet five year plans were based on merits :-/18:42
-!- erasmus [~erasmus@unaffiliated/erasmus] has quit [Read error: Connection reset by peer]18:45
xentracwell, that's good; it means your idea might not be disastrous ;)18:45
kanzurehuh?18:47
cluckjMTBF on soviet five-year plans is about 5 years18:47
xentracif you went back and discovered that the Soviets had been doing the kind of top-down innovation planning you want to do, and it turned out to be a disaster, then you would have significant though not definitive evidence that it was a bad idea18:49
kanzurewhy would you have to go back to find that out, though?18:50
xentracI mean, they certainly didn't leap and bound ahead of the US on innovation; they did make some advances, like the phage therapy mentioned above and a few other things, but mostly they were behind the US18:50
kanzureinstead of going back in time let's just check the old trusty soviet patent database instead http://patents.su/18:51
xentracby "went back" I meant in the sense of reading old records, not literally using time travel18:51
cluckjlol18:51
kanzureoh right, those were economic plans18:55
kanzurenot necessarily science18:55
cluckjcentralized planning wasn't just for economics iirc18:55
kanzureor engineering18:55
kanzuretrying to find.18:55
xentracunfortunately I don't even read Cyrillic, much less Russian, so things like http://patents.su/1976 are hard for me to make much sense of18:55
kanzureamateur18:56
xentracheh18:56
kanzurelaser cutter status?18:57
cluckjscientific research is kinda part of the economy :\18:58
kanzureyes but i dunno level of actual planning of science18:58
kanzuremaybe it was just per-institution funding18:58
kanzure"more science towns. yeah, that sounds good." - stalin 118:59
kanzureso wikipedia does not seem to have a copy of any of the actual plans??19:00
-!- C0RVUS [~C0RVUS@cpc7-hava2-2-0-cust1017.6-1.cable.virginm.net] has joined ##hplusroadmap19:00
xentrac"katushekki made of dvuhdiski mandrel vnuny leash skpri aid tube sliding iCal Katushev ayuscha I am smiling possibilities 5 us, the shaft frames cal pipes, and the Rhone are equipped with one another nom soedinenIzvestna sliding mandrel for winding electrical coils of the authors."19:00
kanzuresigh of course they don't.19:00
kanzurei bet these were "proprietary".19:00
kanzurehow can you execute the plan if you can't even get a copy19:01
xentracmy laser cutter status is that I have an apparently-working workflow from PostScript to MDF objects but I need to try a bunch more things19:01
kanzurehow many fires so far?19:02
xentracI lost most of the last week to an unexpected digestive illness19:02
-!- C0RVUS [~C0RVUS@cpc7-hava2-2-0-cust1017.6-1.cable.virginm.net] has quit [Client Quit]19:02
xentracseveral, but none that didn't blow out immediately19:02
kanzurehm.19:02
xentracI also hacked together some stuff with JS and SVG because my PostScript interaction loop time is only going to get worse19:05
-!- jdqx [~jdqx@2602:306:cc94:1950:3919:2d4c:eb64:715f] has quit [Read error: Connection reset by peer]19:07
-!- mode/##hplusroadmap [+o kanzure] by ChanServ19:09
-!- mode/##hplusroadmap [-q Diablo-D3!*@*] by kanzure19:09
-!- mode/##hplusroadmap [-o kanzure] by kanzure19:09
xentracI find PostScript a little frustrating in that, despite pathforall, your access to the generated geometry is still fairly limited19:13
xentracif you want to find out, for example, whether and where two paths intersect, your options for doing so are limited and inefficient, despite the fact that the RIP has to implement that operation in order to implement clippath and fill19:14
xentracit's not quite as bad as the situation with OpenSCAD19:14
xentracI guess I should quit whining and try other systems, like Antimony and Blender19:16
xentracor BRL-CAD, as nmz787 suggested19:16
maakuxentrac: how was it known the Earth was several billion years old before radioisotope testing?19:19
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has quit [Ping timeout: 265 seconds]19:19
-!- Houshalter [~Houshalte@oh-71-50-56-224.dhcp.embarqhsd.net] has quit [Ping timeout: 276 seconds]19:20
-!- atomical [~atomical@209.153.22.182] has quit [Max SendQ exceeded]19:20
xentrachmm, I'm not sure19:21
cluckjwas it?19:21
xentracit was19:21
cluckjprobably thermodynamics?19:22
maakuI'm not questioning it -- I seem to recall the same trivia that Kelvin knew his gas-law estimation of the Sun's age was wrong19:22
maakubut never thought to ask why he knew that19:23
xentrachttp://www.scientificamerican.com/article/the-age-of-the-sun-and-the-earth/ talks about it a bit19:23
xentracalso says Kelvin's estimate was 32000 years!19:23
xentracit was the sun's future that he thought was limited to 300,000 years, not its past19:24
xentracthen Thomson derived 22M years as its age19:25
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap19:25
xentracI think this Scientific American article may date from before the discovery of fusion19:27
xentrac"if the sun consists of a fraction of one per cent of radium, this will account for and make good the heat that is annually lost by it"19:30
xentrac"As yet this substance has not been found in the sun, but the presence of helium, combined with the fact that helium may be obtained from radium, renders the presence of radium in the sun quite probable"!19:33
xentracso that's the state of knowledge in 190819:34
xentracand it seems rather that Kelvin had persuaded the geologists that their estimate of billions of years as the age of the earth was clearly wrong19:34
kanzurenot sure whether radium/helium was sufficient reason19:35
-!- triggerwarning [~IGLC@unaffiliated/triggerwarning] has quit [Ping timeout: 240 seconds]19:35
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has joined ##hplusroadmap19:35
xentracwell, you would think that the miraculous energy production of radium would have been enough to persuade him19:37
xentracbut no, Kelvin bent over backwards to defend the idea that radium (and by extension other similar "radio-active" phenomena) probably didn't heat either the Earth or the Sun19:38
-!- atomical [~atomical@209.153.22.182] has quit [Max SendQ exceeded]19:41
doclthe ironic thing is that it might be easier to make a Dyson Sphere than a lot of the things we're trying to do, such as small-scale nuclear fusion or curing cancer. the Dyson Sphere only really needs a self replicating robot plus a scalable orbital organizational scheme. who knows how complicated curing cancer or fusing atoms will turn out to be?19:45
-!- triggerwarning [~IGLC@2601:681:500:165a:5007:743a:99a9:5995] has joined ##hplusroadmap19:47
xentracI worry that part of my enthusiasm for your point, docl, stems from my personal enthusiasm for bootstrapping self-replicating machinery :)19:47
xentracsince the other two parts of the problem, distributed-system design and launching the machinery into a solar orbit, are eminently feasible19:48
kanzurei am not actually interested in curing cancer19:49
AlcyiusBesides the fact that "curing cancer" is like saying that you want to "cure virus"19:49
doclI feel like the concept of nanoassemblers has gotten way too much press. macroscale robotic replicators are just as possible, and depend a lot less on unknown hand wavy stuff.19:49
kanzurewhy worry about amount of press?19:50
cluckjjust cure diabetes19:50
kanzurebrain uploading conveniently cuts out all of these problems19:50
kanzurebut you get different problems instead19:51
kanzurepossibly easier problems19:51
kanzureespecially if you are willing to chuck out memory and other junk like that19:51
cluckjI like my brain right where it is19:52
kanzurebrain uploading isn't necessarily about *your* brain... let's set the sights a little bit lower. any brain would do, really.19:52
xentracdocl: a lot of problems go away if you can assemble things atom by atom19:53
xentracI mean, there's a reason the SAE defines criteria for dozens of different kinds of steel19:53
docltrue enough19:54
kanzurethink of the health insurance savings of brain uploading19:54
kanzureand healthcare savings19:54
xentracnearly each of them is optimal for some use, and often its merit exceeds that of other kinds of steel for that use by an order of magnitude or more19:54
xentracif you're assembling things out of bulk materials, you have to confront all of that complexity19:55
nmz787i saw some mention of postscript19:55
xentrac(and this is before you get to annealing vs. tempering vs. case-hardening...)19:55
doclit's just that whenever the topic of self-replicating robots comes up people immediately jump to nanobots. and that's technically a different question. I mean, maybe we actually need to do terascale experimentation before we can get to the point of nanoassemblers.19:55
-!- atomical [~atomical@209.153.22.182] has joined ##hplusroadmap19:55
nmz787I am trying to load some postscript into inkscape to convert to vectors to try out on my small laser etcher19:56
xentracyeah, that's my old-fashioned masochism, nmz78719:56
kanzurereally tiny small laser etcher? or conventional consumer laser etcher.19:56
xentracmaybe try running it through ps2eps first if you're having trouble, nmz787?19:56
nmz787kanzure: both19:56
nmz787well i had to install the latest inkscape version19:56
nmz787lets see what happens now19:56
xentracI'm pretty sure even sodipodi had EPS import19:57
nmz787i also don't know how to tell if inkscape actually outputs vectors or not19:57
kanzureone of my mentors was involved in original postscript specification stuff, i think.19:57
nmz787since I see this: http://blog.linuxgrrl.com/2013/08/12/how-to-produce-vector-eps-with-cmyk-color-using-free-software/19:57
xentrac(pstoepsi is an alternative if ps2eps doesn't work)19:57
nmz787i rememeber inkscape not producing vectors reliably years ago too19:57
kanzurebtw i told xentrac about your svg laser cutter test file19:58
doclI think maybe there's a problem with trying to do everything "optimal"... you need it to be functional, but not necessarily optimal in the early stages of bootstrapping. you can bother with optimizing later on when you have a lot of factories working in parallel with each other.19:58
kanzuredis one http://diyhpl.us/~bryan/papers2/diybio/mccorkle_tomkins-tinch_microchannels.svg19:58
xentracyeah, I looked at the test file and didn't think it helped with what I wanted to find out19:58
xentracdocl: agreed19:58
nmz787kanzure: yep... coreldraw was much more reliable19:58
kanzuredocl: we should ping freitas and complain about lack of progress. maybe he has something since AASM and kinematic self-replicating machines.19:58
kanzurecoreldraw--- i actually used that with my laser cutter. i completely forgot about that.19:59
xentracyeah, everybody runs their laser cutters with coreldraw19:59
nmz787damn, I am in windows and ps2eps is easier to get on linux.... (i got here because I couldn't mount my windows partition from linux as it was in hibernate state)19:59
xentrackanzure: it would be interesting to hear what freitas is up to20:00
xentracnmz787: install it in cygwin or a virtual machine20:00
doclkanzure: that would be awesome. Dani Eder is the only person I'm aware of that actively developing this area20:00
kanzuredani eder keeps ignoring me, i hate him20:00
* xentrac gives kanzure a hug20:00
docldarn20:00
xentracI have no idea what else to recommend for vector format conversions on MSWindows, nmz78720:00
kanzureall i demand is total domination over space matter and time what is so hard to appease about this20:01
nmz787xentrac: rebooting will be easier than messing with cygwin20:01
xentracps2eps (and friends), pstoepsi (and friends), pstoedit, and pstools are the things I have found useful in the past20:01
kanzuredocl: i think that eventually we'll figure out a good way to do http://gnusha.org/skdb but so far it has been hard to find people willing to review and speculate about easier ways to solve these problems.20:01
kanzuredocl: skdb was originally a way to get a bunch of hardware packages and then do search over the graph to find self-replication20:02
xentracyeah, I know20:02
kanzureturns out though that self-replication sort of randomly happens if you just randomly populate a board with different operations,20:02
kanzurehttp://diyhpl.us/~bryan/papers2/ai/The%20evolution%20of%20self-replicating%20computer%20organisms%20-%20A.%20N.%20Pargellis.pdf20:02
xentracwe're just so far from that right now20:02
kanzurewhich might have interesting implications for self-replication20:03
kanzureerm i mean hardware self-replication20:03
kanzure... hardware self-replication *bootstrapping* in particular20:03
xentracyes.  unfortunately the hardware world is full of inconvenient conservation laws, plus the Third Law, which make that harder20:03
kanzureorigin of life stuff (like via andrew ellington etc) has shown that many rna origins were of the type to support simple self-replication precursors similar to that computational simulation (er, except, with actual molecules and polymer chemistry and stuff)20:04
kanzurebut that probably benefited from immensely massive parallelism20:05
kanzurehooking up random bits of hardware until self-replication precursors emerge... does not seem productive.20:05
kanzures/emerge/happen20:05
xentracI don't think self-replicating hardware from natural materials will be that hard20:06
xentracI feel like I kind of have a handle on the problem now20:06
kanzurealso, we already have (biological) self-replication so we should probably continue hijacking that20:06
kanzurewhat sorta handle?20:06
-!- Act [uid89656@gateway/web/irccloud.com/x-ikhmcdeneaizkpvg] has quit [Quit: Connection closed for inactivity]20:07
kanzurei wonder if freitas' book cites nick szabo's clay clanking replicator.20:07
xentracwell, you probably want planar semiconductor fabrication, just because it's so many orders of magnitude ahead of the alternatives20:08
kanzurealright. so that probably requires spatial light modulation somewhere...20:08
kanzureand lithography things.20:09
xentracsure, but SLM is just planar semiconductor fabrication20:09
xentracyou might be able to get by at low efficiency with direct laser writing too20:09
kanzuredo you have additional parts of the answer? or just that :p20:09
xentracbut you need to be able to smelt bulk silicon and aluminum20:09
xentracaluminum can answer nearly all structural and conduction needs adequately, except springs20:10
xentrac6061 aluminum is aluminum alloyed with magnesium and silicon, so you need magnesium for that; I'm not sure if a plain aluminum-silicon alloy is adequately strong, but it's certainly better than just plain aluminum20:11
kanzurei agree that planar semiconductor fabrication is a useful piece of the puzzle. planar manufacturing in general seems to simplify many parts of the overall problem space.20:11
kanzurealthough i haven't figured out how to do planar manufacturing of multi-kilo-newton engines or actuators20:12
xentracwell, most of the fabrication of electric motors or generators is already planar20:13
xentracJacques Mattheij's automated fabrication tool for his windmill project was just a CNC plasma torch table20:14
xentraclow-frequency magnetic machines like that benefit enormously from high-permeability alloys, though, and those are almost all iron-based20:15
xentracso if you can avoid that entirely, it simplifies the picture a lot20:15
xentracyou can get actuators or motors out of piezoelectric or magnetostrictive materials, but you can also get them entirely out of coils of aluminum wire — at the cost of low efficiency20:16
xentracyou can do all kinds of structural and machinery stuff just cutting shit out of sheets of aluminum and jamming it together20:18
doclseems like having cheap access to a large vacuum chamber would make it a lot easier to design stuff optimized for space/lunar conditions.20:19
kanzurein 2007 for some reason i spent a tremendous amount of time looking into the design and fabrication of ultra-high vacuum chambers.20:20
kanzurethis was actually how i ended up meeting and then forgetting azonenberg's roommatte long before he met azonenberg..... oops.20:20
kanzure*roommate20:20
xentracso you need aluminum smelting cells, for which you need a reactor to produce cryolite, and graphite; silicon smelting and zone refining; general-purpose furnaces for materials processing; high-precision robot arms; and cutting torches20:21
xentracI'm not sure if you can get by without lasers20:21
xentracI mean, practically20:21
kanzure((context: it wasn't until much later that i realized that azonenberg's roommate person was someone i had met previously; i thought i only knew azonenberg because of azonenberg things.))20:22
-!- erasmus [~erasmus@unaffiliated/erasmus] has joined ##hplusroadmap20:22
xentracnormal silicon smelting consumes charcoal in an arc furnace20:22
doclkanzure: did you find any good designs for making really big hard vacuum chambers? maybe it would help to put remote controlled robots on the inside and let hackers remotely try to create projects using them.20:23
xentracI'm handwaving a bit about the chemical plant process design aspects, but I feel like they are fairly well understood20:23
doclxentrac: it seems like lasers let you get out of a lot of stuff, like arc welders or drill bits.20:24
kanzuredamn... no results: http://web.archive.org/web/*/http://heybryan.org/projects/atoms/atom_holography_notes.html20:24
xentracyeah, although plasma torches get you out of drill bits to some extent too ;)20:24
xentracyou need a pretty big fiber laser to not need a drill bit for aluminum though20:24
xentracI think a drill bit might be a better bargain20:25
kanzure16:18 < pmetzger> doing SPM based nano bootstrap requires UHV I think, though Merkle has talked about using systems with pure argon atmospheres during bootstrap.20:25
kanzure16:19 < kanzure> hehe with a good page on how to construct an atom laser: http://heybryan.org/projects/atoms/atom_holography_notes.html20:25
doclhmm. 3d printing the preforms for fiber optics is a thing now20:25
kanzurefrom http://gnusha.org/logs/2010-06-15.log20:25
kanzure16:17 < kanzure> i was looking into building UHV chambers a few years ago20:26
kanzure16:17 < kanzure> i was super interested in bose-einstein condensates20:26
kanzure16:18 < kanzure> for some reason it turned into a von neumann probe page for me http://heybryan.org/projects/atoms/20:26
kanzure16:18 < kanzure> because i was interested in atom holography20:26
kanzure16:18 < kanzure> has some good uhv howto links on that page though :)20:26
kanzurehuh.20:26
xentrac(also apparently zone refining won't give you electronics-grade silicon)20:26
xentrac(so a bit of what I said above is wrong)20:27
kanzurehuh did i have a data loss event20:27
xentrac(hmm, maybe it will give it to you and is just not used nowadays because the other methods are more efficient)20:28
kanzurei should read the logs20:29
maakuyeah was about to tell you regarding the 40420:30
xentrackanzure: I'm not finding Szabo's clay-clanking-replicator proposal20:31
kanzure199x post to some usenet newsgroup20:32
xentracwell, I did find a 199x Szabo clanking replicator post20:32
xentracbut it was about legos20:32
kanzurei seem to mention it here https://groups.google.com/d/msg/openmanufacturing/Sdwr1W1JqNo/zcKmFrgCRjQJ20:33
kanzuremaybe not20:33
xentracclay is pretty nice if you can get it; you can build furnaces for firing clay out of clay20:33
kanzure"Excerpts from KSRM on "closure engineering" -- http://groups.google.com/group/openmanufacturing/msg/4ff7a92e2425dde2 "20:33
kanzure"szabo: no results found" (for openmanufacturing) whaaaat20:33
kanzuregod damn time slips wtf20:34
xentracI don't think it will work very well on orbit between Mars and the asteroid belt20:34
kanzure"In 1998, Chris Phoenix suggested a general idea for a macroscale replicator on the sci.nanotech newsgroup, operating in a pool of ultraviolet-cured liquid plastic, selectively solidifying the plastic to form solid parts. Computation could be done by fluidic logic. Power for the process could be supplied by a pressurized source of the liquid."20:34
kanzure"In 1995, Nick Szabo proposed a challenge to build a macroscale replicator from Lego(tm) robot kits and similar basic parts. Szabo wrote that this approach was easier than previous proposals for macroscale replicators, but successfully predicted that even this method would not lead to a macroscale replicator within ten years."20:34
kanzurechris phoenix proposal http://groups.google.com/groups?hl=en&selm=6f0nui%248ih%241%40news.nanospace.com20:34
kanzurenick szabo proposal http://www.lucifer.com/~sean/N-FX/macro.html20:34
kanzure404....20:35
xentracthat's the one I just read20:35
kanzurei have a backup of lucifer.com but it's just something i stole in 2014 so it wont have this20:35
kanzurehttp://web.archive.org/web/20000305141204/http://www.lucifer.com/~sean/N-FX/macro.html20:35
xentracyeah20:35
kanzureTo: extropians@waterville.warwick.com, if-sci@lucifer.com20:36
kanzurei was not aware of an extropians mailing list on that server20:36
maakudocl: maybe work on sand-crawlers across the sahara, shitting out roads and fiber optic cables?20:36
kanzuremerely these things http://diyhpl.us/~bryan/irc/extropians/20:36
maakudocl: more interesting as a target is a self-replicating lunar base imho20:36
xentracwouldn't be surprising if *.warwick.com aliased to a single mail routing system20:37
maakubut the digital matter of nanotech is so much easier to work with20:37
xentracthe desert is more accessible20:38
kanzureyeah i don't see anything about a clay clanking replicator20:38
kanzurethis seems like something fenn will remember though20:38
xentracsubterranean realms also20:38
doclwell, if you can 3d print the preforms and the drawing tower, and also create a laser from the fiber, then all you need is for everything else to be dependent on the laser(s).20:39
xentracagreed20:39
xentracyou need erbium refining or something for the fiber, don't you?  or are there less exotic phosphors that can work adequately?20:40
erasmusI want a 3D printer20:42
erasmusbut the large ones are still expensive20:42
erasmusif I had one then I could make this  https://github.com/OpenBCI/Ultracortex/tree/master/Mark_320:42
xentracI'm less and less enthusiastic about 3D printers, even as they get better20:43
kanzurecapthindsight is all about 2d printers and 6-axis 2d printers. so you're in good company. (oops i guess not; where did he go?)20:43
xentrachonestly that particular thing looks like you could do a nearly equally good job by cutting holes in a baseball cap20:43
xentracthe dry electrodes are the interesting part, and they aren't printable20:44
xentrac(and the circuit board, which likewise)20:45
doclwhat purpose does the erbium serve? does it simply generate light of the desired frequency?20:47
doclit seems like you could simply use an optical filter to grab the desired frequency from sunlight.20:48
xentracit fluoresces20:48
xentraca laser works by having a "population inversion" among atoms capable of emitting light at the desired frequency20:48
xentracnormally most of those atoms are in their ground state, so photons of that energy are more likely to be absorbed (exciting an atom) than to be emitted (allowing an excited atom to relax to the ground state)20:49
xentracwhich means that as light travels through regular glass, for example, gradually more and more of the red and blue parts of the spectrum are absorbed by different excitable atoms20:49
xentracif you "invert" these populations, so that more of the atoms are in the excited state than in the ground state, emission becomes more likely than absorption, and the object glows instead of absorbing light20:50
xentracin a laser, this happens in a process called "stimulated emission", which was super obvious when I read Feynman but now I can't explain adequately20:51
xentracsuggesting a fluency illusion I need to debug20:51
-!- atomical [~atomical@209.153.22.182] has quit [Quit: My Mac has gone to sleep. ZZZzzz…]20:51
xentracso what you do is that first you "pump" the emitting atoms, so that most of them are excited, and then you either wait or you hit the laser with a small stimulus pulse20:52
xentracyou can pump them in lots of ways20:52
xentracsunlight is one20:52
maakuxentrac: there isn't an economic need for self-replicating robots anywhere on Earth though, even the desert20:53
xentracbut the intensity of the sunlight matters20:53
maakuit makes tremendous sense for lunar resource development, however20:53
xentracmaaku: not until about 2025 maybe?20:53
maakuwhy 2025?20:53
kanzureseems to be large need for biological self-replicating robots on earth. non-biological self-replication would also be very helpful, such as for tooling and equipment reasons.20:54
maakukanzure: helpful, but i'm talking opportunity costs because tech is expensive20:54
kanzurealthough previously many manufacturing implements were the result of dead biological self-replicators20:54
kanzureoh right opportunity cost. right, right.20:54
maakunote price to delivery to lunar surface is about $40k/lb20:54
maakuand only in small quantities20:54
maakuand thanks to the wonder of impact dynamics, the regolith is a fine powder of various ores20:54
xentracthat's about the time photovoltaic production starts to amount to a significant part of human energy production20:55
maakuxentrac: so have a human built factory building tons of solar panels, and people setting them up ... not sure why the robotic complications are necessary20:55
doclif you could actually make them work, there be plenty of demand for Earth based self-replicating robots.20:55
xentracor I guess human marketed energy production20:56
doclit's basically free money, power, goods, equipment...20:56
xentracmaaku: did you see the wired article about the single-purpose CNC mill whose purpose I won't mention?20:56
maakudocl: you are significantly more optimistic than me at its ability to operate without intervention :)20:56
kanzurewas the purpose to annoy me? i bet it was to annoy me.20:56
xentracheh, no, it was to avoid triggering keyword filters20:57
maakuhahaaha (and no, I didn't)20:57
xentracthe guy talked about the enormous difference between trying to do the machining by hand and doing it on the CNC mill that did the same thing20:57
xentracbasically because he isn't a master machinist he had no hope of completing the task by hand20:57
xentraca Santa Claus Machine could eliminate human poverty if people aren't denied access to it20:58
xentracalthough given how Wikipedia is less popular than cat photos I'm not sure how well people will use it20:58
kanzureso what you're saying is the lego instruction booklet generator needs to include cat pics?21:00
maakuxentrac: you seem young. youthful enthusiasm is good. don't get dissuaded by old farts like me that are too cynical to think such ideas would work21:00
kanzurexentrac is older than dirt21:00
xentracheh, I'm 3921:00
doclmaaku: if we're talking about partially automated factories, the reason there's demand on earth is because nobody wants to launch an untested concept to the moon.21:00
xentracand that's only in this body!21:00
maakuhah! ok. I'm only 3121:01
kanzurebuncha old farts in here21:01
maakudocl: well yeah, obviously you test it here on earth21:01
maakuI'm just doubtful an economic argument can be made for use on earth vs alternatives21:02
xentracthe flip side of the santa claus machine is that if it can self-replicate solar panels then you can manage a lot more concentration of wealth if it doesn't require employees to man the factory21:02
-!- ArturShaik [~ArturShai@37.218.157.155] has joined ##hplusroadmap21:02
maakui mean obviously if the tech was fully developed and bug-free, it'd get used21:02
maakubut you'd run out of funding getting to that point21:02
xentracwhich is also an economic argument: as founders and investors of the startup, you get to keep all of the value it produces, instead of paying the majority of it to employees21:02
xentracthis is the kind of economic structure that has historically led to disasters like Anaconda Copper and Nauru21:03
maakuwhereas even if you have to use remotely operated vehicles driven by people at mission control to debug a lunar self-replicator, and a fully-staffed regolith analog lab where you try stuff out, it's still cheaper than sending smelters and other industrial equipment to the moon21:03
doclwell you can't necessarily test everything on earth because stuff that relies on low gravity doesn't work here. Also, vacuum chambers of the size that were talking about tend to be kind of expensive. at some point it might actually be a lot cheaper to simply test it in space.21:04
maakuxentrac: I'll buy that argument for molecular nanotech, which I consider a simpler proposition21:04
xentracMNT would also have that effect, yes21:04
xentracwell, all of those effects21:04
xentracthe distribution of access to it will make all the difference21:05
maakuIt is my informed opinion that the only thing holding up MNT is funding at this point. There's a clear path forward, just needs lots of money to do it. (So I work on bitcoin)21:05
xentracI suspect that groups that manage to get many people to participate in the process of invention and debugging will do better than groups that don't21:05
xentracpresumably Satoshi would be funding MNT if he cared ;)21:06
maakufor macro self-replication, I'm not so sure ... there's a lot of complexity to be tamed, and it seems to me a much more unsolved problem as a result21:06
maakumaybe that is my own ignorance, however21:06
kanzurehal should have setup a bitcoin/crytonics fund on his way out21:06
xentracno, I think you're right that there's a lot of complexity and unsolved problems involved21:06
xentracI don't really know whether it's easier or harder than the Nanorex angle on MNT21:07
maakumacro self-replicating machines seems like an AGI-complete problem, just given the complexity of unconstrained environments (and if you constrain the environment too much, it gets less interesting)21:07
xentracwhich I assume is what kanzure means21:08
maakuxentrac: Nanorex angle?21:08
kanzurenanorex was https://github.com/kanzure/nanoengineer21:08
maaku(Although -- last thing I'll say on this -- lunar environment is naturally highly constrained.)21:08
xentracright21:08
xentracI think the Atacama is too21:08
xentracthe temperature range is even lower!21:09
maakuright just not sure what "Nanorex angle on MNT" meant21:09
maakuxentrac: so long as sand provides what you need. sand on earth is basically silica. regolith is an even mixture of highly enriched ores.21:09
xentracI should let kanzure explain Nanorex's plan in more detail if he feels like it, since I have only a glancing acquaintance21:10
xentracyes, that is true21:10
xentracbut lacking, for example, hydrogen21:10
kanzureno- please continue21:10
kanzurealso i am afk21:10
maakuxentrac: that's why you'd operate at the poles21:11
maakucold is a problem though21:11
maakui'm about to be afk too21:12
xentracyeah, I should be too21:12
maakuI will mention I have a business idea for jump-starting private space industry and self-replicating lunar bases via commodity markets using Blockstream technology and the clever application of legal jurisdictions21:13
maakuIf anyone wants to do that, let  me know. I sincerely doubt I'll get around to it given priorities.21:14
-!- jdqx [~jdqx@2602:306:cc94:1950:71:46e0:d4e:3e89] has joined ##hplusroadmap21:16
-!- FAMAS [~FAMAS@103.198.139.33] has joined ##hplusroadmap21:17
-!- FAMAS [~FAMAS@103.198.139.33] has quit [Changing host]21:17
-!- FAMAS [~FAMAS@unaffiliated/famas] has joined ##hplusroadmap21:17
-!- jaboja [~jaboja@erz41.neoplus.adsl.tpnet.pl] has quit [Remote host closed the connection]21:36
-!- jtimon [~quassel@74.29.134.37.dynamic.jazztel.es] has quit [Ping timeout: 240 seconds]21:36
-!- poppingtonic [~Thunderbi@unaffiliated/poppingtonic] has joined ##hplusroadmap21:47
-!- FAMAS [~FAMAS@unaffiliated/famas] has quit [Read error: Connection reset by peer]21:55
-!- FAMAS [~FAMAS@103.198.139.33] has joined ##hplusroadmap21:55
-!- FAMAS [~FAMAS@103.198.139.33] has quit [Changing host]21:55
-!- FAMAS [~FAMAS@unaffiliated/famas] has joined ##hplusroadmap21:55
erasmusmy cats breath smells like catfood21:56
-!- jdqx_ [~jdqx@108-201-65-149.lightspeed.ftwotx.sbcglobal.net] has joined ##hplusroadmap22:00
-!- FourFire [~FourFire@192.40.88.18] has joined ##hplusroadmap22:02
-!- jdqx [~jdqx@2602:306:cc94:1950:71:46e0:d4e:3e89] has quit [Ping timeout: 260 seconds]22:04
doclwhat information do we have about creating ultrahard vacuums?22:06
poppingtonichttps://xkcd.com/1619/22:07
doclhttp://www.cloudynights.com/topic/466140-diy-aluminizing-chamber/22:17
justanotheruserultra high vacuum?22:22
justanotheruseror do you actually mean ultrahard vacuum22:23
doclhttp://www.belljar.net/basics.htm22:25
doclis there a difference?22:25
doclI've heard the vacuum of space referred to as hard vacuum, but ultra high vacuum seems to be a common term as well.22:28
justanotheruserI have never heard of it myself22:28
doclI'm thinking hard is probably a less formal term when applied to vacuum22:30
docl"ultrahard vacuum" might just be a malapropism on my part22:31
-!- yashgaroth [~yashgarot@2602:306:35fa:d500:f5e0:f867:a11d:8d52] has quit [Quit: Leaving]22:32
-!- FourFire [~FourFire@192.40.88.18] has quit [Read error: Connection reset by peer]22:33
-!- AmbulatoryCortex [~Ambulator@173-31-155-69.client.mchsi.com] has quit [Read error: Connection reset by peer]22:33
-!- FourFire [FourFire@cm-84.215.195.59.getinternet.no] has joined ##hplusroadmap22:38
-!- CheckDavid [uid14990@gateway/web/irccloud.com/x-plqutunanchzpmgo] has quit [Quit: Connection closed for inactivity]22:49
doclhttps://en.wikipedia.org/wiki/Ultra-high_vacuum22:51
-!- errrasmus [~erasmus@unaffiliated/erasmus] has joined ##hplusroadmap22:58
-!- erasmus [~erasmus@unaffiliated/erasmus] has quit [Ping timeout: 255 seconds]22:59
-!- errrasmus [~erasmus@unaffiliated/erasmus] has quit [Read error: Connection reset by peer]23:08
-!- Madplatypus [uid19957@gateway/web/irccloud.com/x-falznafctrcdjkkh] has joined ##hplusroadmap23:23
-!- nmz787_i [ntmccork@nat/intel/x-baoczqttcjasbcqf] has joined ##hplusroadmap23:26
-!- nmz787_i [ntmccork@nat/intel/x-baoczqttcjasbcqf] has quit [Quit: Leaving.]23:40
-!- |node [uid125132@gateway/web/irccloud.com/x-tmazfrxeeusxeylc] has quit [Quit: Connection closed for inactivity]23:41
--- Log closed Wed Dec 23 00:00:44 2015

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