2011-03-20.log

--- Log opened Sun Mar 20 00:00:10 2011
-!- augur [~augur@129.2.129.33] has quit [Remote host closed the connection]00:09
-!- augur [~augur@208.58.6.161] has joined ##hplusroadmap00:31
ybitfenn, kanzure, you hate siai01:35
ybitwhy01:35
ybitmaybe hate isn't the correct word01:36
ybit2009-12-01.log:23:24 < kanzure> i hate how the singularity institute is focused on "existential risks"01:44
ybit10:01 < kanzure> they were completely unwilling to listen to me because manufacturing is unreasonable or something01:46
ybit10:02 < Smari> existential risk folks annoy me too.01:46
ybit10:02 < Smari> they're severe doomsayers.01:46
ybit10:02 < kanzure> right and that's what this community is about right now01:46
ybiti'm guessing this is most of it01:47
ybit0:02 < kanzure> and we're trying to change that01:47
ybit10:02 < Smari> and won't be told that doom can be averted, because that means they're out of a job.01:47
ybit10:03 < fenn> that's why any reasonable plan of action involves dismantling the global employee state we've got going01:47
ybiter, guess that was unrelated :)01:48
JayDuggerMy comment or fenn's quote about global employee state?01:49
ybitJayDugger: you are Smari?01:51
-!- devrandom [debian-tor@gateway/tor-sasl/niftyzero1] has quit [Ping timeout: 246 seconds]03:42
-!- ThomasEgi [~thomas@pppdyn-0e.stud-ko.rz-online.net] has joined ##hplusroadmap03:59
-!- ThomasEgi [~thomas@pppdyn-0e.stud-ko.rz-online.net] has quit [Changing host]03:59
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap03:59
-!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined ##hplusroadmap07:28
JayDugger...07:50
-!- enki-][ is now known as enki-[quit]07:55
-!- jrabbit [~babyseal@unaffiliated/jrabbit] has quit [Ping timeout: 240 seconds]08:01
-!- jrabbit [~babyseal@unaffiliated/jrabbit] has joined ##hplusroadmap08:22
jrayhawkSIAI's main concern, AI takeoff, is potentially the most significant event in human history, and, among significant events, it has a rare distinction in that we can forsee and influence its initial circumstances such that we might effect its eventual results. The SIAI has taken upon itself to be the stewards.08:33
jrayhawkThis responsibility is essentially unfathomable, so they have two real goals at this point:08:33
jrayhawk1) attempt to find a way of comprehending their undertaking08:34
jrayhawk2) avoid fucking anything up until 1) is done08:34
-!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 246 seconds]08:35
jrayhawkBryan and Fenn, who make tangential progress on the singularity, do not fit in well with these goals.08:37
jrayhawkBryan seems to intuit that a goal system such as the SIAI's will result in nothing being accomplished, and a lot of smart people being used to accomplish it.08:41
jrayhawkWhich is certainly plausible.08:41
jrayhawkI'm not really all that resentful of it; SIAI's and FHI's rationality blogs (training and recruitment centers for goal number one) seem like a rising tide raising all ships.08:59
jrayhawkand rationality fanfic, of course09:00
JayDuggerHarry Potter fanfic, if you please.09:01
-!- devrandom [debian-tor@gateway/tor-sasl/niftyzero1] has joined ##hplusroadmap09:24
JayDuggerGood night, everyone.09:27
-!- JayDugger [~duggerj@pool-173-74-118-239.dllstx.fios.verizon.net] has left ##hplusroadmap []09:27
-!- jrabbit [~babyseal@unaffiliated/jrabbit] has quit [Ping timeout: 276 seconds]10:22
-!- Technicus [~Technicus@DSLPool-net209-116.wctc.net] has joined ##hplusroadmap10:24
-!- mjr [mjr@2001:1bc8:102:60d4:21b:fcff:fe32:5eda] has quit [Read error: Operation timed out]10:30
-!- jrabbit [~babyseal@unaffiliated/jrabbit] has joined ##hplusroadmap10:40
-!- mjr [mjr@2001:1bc8:102:60d4:21b:fcff:fe32:5eda] has joined ##hplusroadmap10:50
-!- uniqanomaly__ [~ua@62.87.224.60] has joined ##hplusroadmap11:10
-!- uniqanomaly_ [~ua@dynamic-78-8-85-57.ssp.dialog.net.pl] has quit [Ping timeout: 240 seconds]11:12
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Ping timeout: 255 seconds]11:46
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap11:59
-!- Juul [~Juul@c-24-130-50-84.hsd1.ca.comcast.net] has joined ##hplusroadmap12:02
-!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined ##hplusroadmap12:17
-!- jmil [~jmil@2001:468:1802:e148:223:32ff:feb1:9dfc] has quit [Quit: jmil]12:29
-!- mjr [mjr@2001:1bc8:102:60d4:21b:fcff:fe32:5eda] has quit [Read error: Operation timed out]12:37
-!- mjr [mjr@2001:1bc8:102:60d4:21b:fcff:fe32:5eda] has joined ##hplusroadmap12:37
-!- Juul [~Juul@c-24-130-50-84.hsd1.ca.comcast.net] has quit [Quit: Leaving]12:40
-!- mjr [mjr@2001:1bc8:102:60d4:21b:fcff:fe32:5eda] has quit [Read error: Operation timed out]12:44
-!- mjr [mjr@2001:1bc8:102:60d4:21b:fcff:fe32:5eda] has joined ##hplusroadmap12:48
-!- jebba_ [~jebba@ip-216-17-181-80.rev.frii.com] has quit [Ping timeout: 264 seconds]13:39
fennbertrand russell fanfic13:40
ybiti never really thought about agi seriously until yesterday13:43
ybitsomething that i guess is hard to admit13:44
fennwhat do you mean thought about seriously?13:47
ybityeah, that's something that need clarification, but i'm having a tough time focusing13:48
ybitprobably should brew some coffee13:48
ybiti'm writing my thoughts out right now, one sec13:48
ybitwell, i could just think aloud in here13:49
ybitat some point we're going to try expanding our brain's current limitations with nanotechnology, before we do this we need to better understand the brain.13:49
ybitthe logical way of doing this would be to create an aritifical brain13:49
ybitonce you've done that, you've created the first agi13:50
ybitthis is where i'm concerned13:50
ybiti never really considered that agi would come before we used nanotech in our brains, but it makes sense13:51
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has joined ##hplusroadmap13:54
ybitthis puts into question if creating an artificial brain and publishing the results openly is a good idea13:54
-!- jebba_ [~jebba@ip-216-17-181-80.rev.frii.com] has joined ##hplusroadmap13:55
ybiti put all this in here, because i'm assuming that others in here have given it more thought13:55
fennpeople are already creating artificial brains, albeit small ones13:55
fennwe're nowhere close to "real" nanotech13:56
ybitright13:56
fennhowever artificial brain != agi13:56
ybityes13:56
fenni.e. your cat can't write C++13:56
fennpublishing the results openly is a good idea.13:57
ybitwhy?13:57
fennyou aren't smart enough to hog it13:57
fennthe combined efforts of everybody will produce more net benefit than whatever you personally (or your in-group) could come up with13:58
fennalso, if there is indeed any danger, others will be made aware of it13:59
fennif SIAI is actually developing an AGI in secret like they were discussing years ago, i would be very suprised13:59
ybiti suppose the first agi would be built using supercomputers, and the only persons that would have the capabilities would  large corps and extremely wealthy individuals13:59
fenni guess14:00
fennyou can do a lot with a modern GPU14:00
ybiti don't know enough to say if it could or couldn't be written to thrive on a botnet14:00
mjrbotnet distribution is hard14:00
ybiti'm concerned about safeguards these organizations would use when building the agi14:01
fenni'm still in the camp that AGI is trivial once you know how to do it14:01
ybitalso about agi taking off and obliterating humanity14:02
fennok, well, write a concerned letter to your nearest university computer science department14:03
* fenn goes and reads some web comics14:04
ybitit would almost seem logical to (a) not allow for this agi to be localized, being fed only the information needed to solve a particular problem and (b) to have the plug pulled after a given time, maybe to prevent it from getting out of control. maybe b isn't necessary14:04
fennyou mean embodied?14:04
ybitlocalized, er, as in restricted access to the internet14:05
fennpff14:05
fenngood luck14:05
ybitright, i figured a response would come14:05
ybiti want more thoughts on the matter, i've given it little thought14:05
fenni thought it was interesting that watson wasn't connected to the net14:05
fennseems they did it because the net contains mostly junk14:06
ybithrm14:06
ybiti guess that is silly14:07
ybiti just assumed access to the internet meant access to anything it wanted or needed to expand and (appeal to pathos) destroy humanity14:08
ybitand i haven't read the theories behind seed ai14:10
ybitcurious to know the proposed methods for recursive improvement14:11
fenni havent found anything beyond vague "it would design a better AI"14:14
fennlet me know if you do14:14
jrayhawkhttp://yudkowsky.net/singularity/aibox14:26
-!- devrandom [debian-tor@gateway/tor-sasl/niftyzero1] has quit [Ping timeout: 246 seconds]14:30
-!- wrldpc2 [~benny@58-89-241-19.nttmil.net] has quit [Quit: wrldpc2]14:32
-!- augur [~augur@208.58.6.161] has quit [Remote host closed the connection]14:39
kanzure"yes but it's 100000 times smarter than you, so you can't possibly do anything about it!!!"14:45
kanzure"nuh uh"14:45
kanzure"i love mom MORE"14:46
kanzure"no i love mom more times 5"14:46
kanzure"no i do, time 20"14:46
kanzure"no me, times infinity!!"14:46
QuantumGfrom what I've seen of AI researchers, that multiplier brings the AI up to Big Bang Theory smarts.14:48
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has quit [Ping timeout: 240 seconds]15:06
-!- augur [~augur@129.2.129.33] has joined ##hplusroadmap15:16
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has joined ##hplusroadmap15:19
-!- phreedom [~quassel@109.254.17.41] has quit [Ping timeout: 250 seconds]15:45
-!- phreedom [~quassel@109.254.17.41] has joined ##hplusroadmap15:47
-!- uniqanomaly__ [~ua@62.87.224.60] has quit [Read error: Connection reset by peer]16:00
-!- uniqanomaly [~ua@62.87.224.60] has joined ##hplusroadmap16:00
-!- Jappe2 [jappe2@kapsi.fi] has quit [Ping timeout: 260 seconds]16:15
-!- Jappe2 [jappe2@kapsi.fi] has joined ##hplusroadmap16:15
-!- strages [~strages@wifi.makerslocal.org] has joined ##hplusroadmap16:25
-!- devrandom [debian-tor@gateway/tor-sasl/niftyzero1] has joined ##hplusroadmap16:51
-!- JayDugger [~duggerj@pool-173-74-79-43.dllstx.fios.verizon.net] has joined ##hplusroadmap16:59
-!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has joined ##hplusroadmap17:11
JayDuggerGood evening, everyone.17:38
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Remote host closed the connection]17:42
-!- JayDugger [~duggerj@pool-173-74-79-43.dllstx.fios.verizon.net] has left ##hplusroadmap []17:59
-!- phreedom [~quassel@109.254.17.41] has quit [Ping timeout: 276 seconds]19:25
-!- phreedom [~quassel@109.254.17.41] has joined ##hplusroadmap19:29
kanzureback19:45
-!- wrldpc2 [~benny@58-89-241-19.nttmil.net] has joined ##hplusroadmap19:47
kanzuresubmitted my ycombinator application20:07
-!- jmil [~jmil@c-68-81-252-40.hsd1.pa.comcast.net] has quit [Quit: jmil]20:35
wrldpc2if anyone stumbles across artilect war pl0x holler20:46
-!- wrldpc2 [~benny@58-89-241-19.nttmil.net] has quit [Quit: wrldpc2]20:47
-!- strages [~strages@wifi.makerslocal.org] has quit [Ping timeout: 248 seconds]20:50
-!- eridu [debian-tor@gateway/tor-sasl/eridu] has quit [Ping timeout: 246 seconds]20:53
-!- klafka [~textual@cpe-66-66-10-44.rochester.res.rr.com] has quit [Ping timeout: 248 seconds]21:19
-!- klafka [~textual@cpe-66-66-10-44.rochester.res.rr.com] has joined ##hplusroadmap21:22
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has joined ##hplusroadmap21:52
-!- eridu [debian-tor@gateway/tor-sasl/eridu] has joined ##hplusroadmap21:53
-!- timschmidt [~timschmid@h184-61-46-42.mdsnwi.tisp.static.tds.net] has quit [Read error: No route to host]22:04
-!- klafka [~textual@cpe-66-66-10-44.rochester.res.rr.com] has quit [Quit: Computer has gone to sleep.]22:20
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has quit [Ping timeout: 246 seconds]22:24
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has joined ##hplusroadmap22:25
-!- mheld [~mheld@pool-173-76-224-45.bstnma.fios.verizon.net] has quit [Quit: mheld]23:05
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has quit [Ping timeout: 260 seconds]23:14
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has joined ##hplusroadmap23:15
-!- sebastienb [~sbailard@206.248.158.253] has joined ##hplusroadmap23:16
sebastienbHi all.  kanzure, you about?23:18
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has joined ##hplusroadmap23:22
sebastienbAny gitduino devs about?23:24
sebastienb:D23:24
sebastienbjrayhawk, ping?23:40
jrayhawkhello!23:45
sebastienbHow goes it?23:45
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has quit [Remote host closed the connection]23:46
sebastienbI was wondering how gitduino is coming along?23:46
sebastienb:D23:46
-!- strages [~strages@c-71-207-215-204.hsd1.al.comcast.net] has joined ##hplusroadmap23:47
jrayhawkI'm not well informed about gitduino; its relationship to my project management stuff is confusing and I don't keep track of it anyway. It's possible Bryan has the codebase up somewhere, let me see...23:48
sebastienbI've been figuring that self-hosting will be one of the major waypoints.23:49
sebastienbI'm hoping that we'll be doing up a blog.reprap.org intro to it soon.23:51
sebastienbOne of the requirements we need to switch over is co-maintainers, so there's a payoff to announcing it, once it's got the usability to make it interesting.23:52
jrayhawkI don't see a published repo for it, so I'm probably about as informed as you. It's possible Bryan has specific features he's waiting on from me; if he does, you can feel free to bug me on his behalf or whatever.23:53
sebastienbI emailed Bryan yesterday.  I'll try pinging him again tomorrow.23:54
sebastienb(The mediawiki has been driving me up the wall for ~5 years ...)23:54
sebastienbjrayhawk, I think I'll be signing off. It's a bit late, and I've got a busy day tomorrow.23:57
jrayhawkLast I heard, gitduino is a sort of a nice wrapper for piny anyway, so if you want something primitive working now and have a VPS lying around, I can install that for you.23:57
jrayhawkinstall piny, i mean23:57
--- Log closed Mon Mar 21 00:00:03 2011

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