| --- Day changed Sun Aug 24 2008 | ||
| kanzure | A novel pixon-representation for image segmentation based on Markov random field http://www.sciencedirect.com/science?_ob=GatewayURL&_method=citationSearch&_urlVersion=4&_origin=SDVIALERTHTML&_version=1&_uoikey=B6V09-4SDGR1W-5&md5=9870b2fd9ddc1e6547f7c72468614b39 | 02:21 |
|---|---|---|
| ybit | novel indeed | 03:19 |
| kanzure_ | Hi all. | 18:20 |
| nsh | ohai | 18:22 |
| kanzure_ | What I have: four boxes, one laptop, three ethernet PCI cards, one DVD drive, at least two hdd. I think I can throw the 2 hdd and 1 ethernet card into the new box, and then have a gateway with three ethernet ports + 1 onboard port, totalling up to four connected computers at a time. | 18:23 |
| kanzure_ | The catch is that the server would have to come offline for a bit. Any objections? :p | 18:23 |
| nsh | none here | 18:23 |
| kanzure_ | ok, anybody remember what the driver is on debian to make it go vroom? erm, I mean, to serve as a gateway/router thingy? | 18:23 |
| nsh | no idea, sorry | 18:28 |
| kanzure_ | Hm. I'm not good at naming my machines. | 21:02 |
| Phreedom | kanzure_: I think every linux box is a router by default | 21:39 |
| kanzure_ | I had to go through some configuration last time I did it. | 21:49 |
| fenn | are you talking about IP masquerade NAT stuff? | 21:50 |
| kanzure_ | possibly | 21:50 |
| kanzure_ | not sure :) | 21:50 |
| kanzure_ | I used to run a wireless box as a router with its extra ethernet port to another machine | 21:51 |
| kanzure_ | erm, not quite a router | 21:51 |
| kanzure_ | just as a port ghost | 21:51 |
| kanzure_ | also, I don't know what to hostname this box | 21:51 |
| fenn | you'll probably have to set up some iptables rules, it's pretty easy | 21:51 |
| kanzure_ | that's all? | 21:51 |
| * kanzure_ can do that. | 21:52 | |
| fenn | sudo iptables --table nat --append POSTROUTING --out-interface eth1 -j MASQUERADE | 21:53 |
| fenn | sudo iptables --append FORWARD --in-interface eth1 -j ACCEPT | 21:53 |
| fenn | echo 1 | sudo tee /proc/sys/net/ipv4/ip_forward | 21:53 |
| kanzure_ | does irssi log by default? | 21:53 |
| fenn | no | 21:53 |
| kanzure_ | bastards | 21:53 |
| kanzure_ | can I turn on logging from within an irssi session? | 21:53 |
| fenn | yeah | 21:54 |
| nsh | why would anything nnot log by default in this day and age... | 21:54 |
| fenn | terrists | 21:54 |
| nsh | damn terr | 21:54 |
| nsh | must it ruin every good thing | 21:54 |
| fenn | oh, i guess you meant 'can i save whats on the screen to a log file' which is probably "no" | 21:55 |
| kanzure_ | Sort of, I also mean in general | 21:58 |
| kanzure_ | since it's been running for the past month | 21:58 |
| * nsh would suspect you can save the buffer | 21:58 | |
| nsh | oh | 21:58 |
| nsh | then you're probably fuxed | 21:58 |
| kanzure_ | what was that word that I used the other day that meant that one is an egglayer? | 22:03 |
| fenn | oviparous | 22:04 |
| kanzure_ | robot parents maybe | 22:05 |
| kanzure_ | I was trying to figure out how to make it work for humans ;-) | 22:05 |
| kanzure_ | parenting a von neumann probe kind of sucks | 22:05 |
| fenn | you mean more than just the egg-laying, but the child-rearing process too? | 22:06 |
| kanzure_ | well, if you just did egg laying, children without parents are supposedly known to die | 22:07 |
| kanzure_ | however I can't recall citations on this one .. | 22:07 |
| kanzure_ | it might just be a myth | 22:07 |
| fenn | i dont see why they should, with a big enough egg | 22:07 |
| kanzure_ | nutrition isn't the problem | 22:07 |
| kanzure_ | psychobabble stuff | 22:07 |
| fenn | bullshit | 22:07 |
| kanzure_ | could we test with mice? | 22:07 |
| fenn | they might grow up feral and unable to speak, but not dead | 22:08 |
| kanzure_ | what about the Nazi-infant studies? | 22:08 |
| fenn | what nazi infant studies? | 22:08 |
| kanzure_ | dunno, just something I've heard | 22:08 |
| kanzure_ | the infants died when only provided nutrition | 22:08 |
| kanzure_ | don't know if this included diaper changing or sanitation and so on | 22:08 |
| fenn | well, if wolves can raise a baby then robots can probably do a decent job | 22:09 |
| kanzure_ | wolves ? | 22:09 |
| kanzure_ | thought that was just a legend :) | 22:09 |
| fenn | nope | 22:09 |
| kanzure_ | well then. | 22:09 |
| kanzure_ | if you can get me a ref on that, that'd be awesome | 22:09 |
| * kanzure_ finds snopes .. | 22:09 | |
| fenn | http://www.feralchildren.com | 22:10 |
| fenn | lol | 22:10 |
| fenn | i wonder if there's a sort of goth subculture around feral children | 22:10 |
| kanzure_ | "Rome wasn't made in a day, true, but it was made by wolf children." | 22:12 |
| fenn | i saw a documentary where this leopard killed a monkey and then felt sorry for the baby money and raised it | 22:12 |
| kanzure_ | this would be a simple experiment to do with a mouse, really | 22:14 |
| kanzure_ | baby mouse, wire it up with one of those subdermal medicine dispension systems if we want | 22:14 |
| fenn | a man is not a mouse, it doesn't prove anything | 22:14 |
| kanzure_ | for periodic nutrition releasal | 22:14 |
| kanzure_ | (refill during sleepcycles) | 22:15 |
| kanzure_ | hm | 22:15 |
| fenn | but if you had a hyperintelligent mouse, it might be good enough to skip the humans altogether :P | 22:15 |
| fenn | syrian gazelle boy! | 22:17 |
| kanzure_ | let's mount a DNA synthesizer, sequencer, ectogenetic tank / artificial/synthetic/otherwise-normal womb, and wearable computer on an exoskeleton | 22:19 |
| kanzure_ | it's not quite what I've been thinking | 22:20 |
| kanzure_ | you know how I got all hyper about the "recursive public" phrase and the way twobits.net says it? recursive public == SKDB == recursive self-improvement, except it's not really 'improvement', it's about architectures of these 'publics' or 'selves'. such as an exoskeletal architecture .. or an architecture where you're not going to completely, totally die if you die. By "you" I mean something else than consciousness though, so don't interpret this p | 22:21 |
| kanzure_ | *removable/portable womb ;-) | 22:21 |
| kanzure_ | no reason to emulate humans there .. | 22:21 |
| fenn | "so don't interpret this p" | 22:22 |
| kanzure_ | poorly | 22:22 |
| * kanzure_ was being too verbose | 22:22 | |
| fenn | "recursive public" sounds like in-crowd jargon to me | 22:23 |
| kanzure_ | there's a concept of a 'public' | 22:23 |
| kanzure_ | normally it's just "whoever comes by to pick something up" and it's literally an "anybody" | 22:23 |
| kanzure_ | whereas when you say "public" it's more like "family" or "people that are in the same context" | 22:23 |
| fenn | i know what it means, i just hate using jargon because it makes the person who doesn't know it feel inferior/ignorant | 22:23 |
| kanzure_ | wait, what's the alternative wording? | 22:23 |
| kanzure_ | does "recursive societies" mean the same thing? | 22:24 |
| fenn | yes, but "recursive" doesn't carry the right meaning on its own | 22:24 |
| fenn | also i think 'public' in 'recursive public' refers to the infrastructure as well as the people | 22:25 |
| fenn | but whatver | 22:25 |
| kanzure_ | people are infrastructure | 22:25 |
| kanzure_ | is anything ever secure/private? | 22:25 |
| fenn | um, that's a rather distorted view of the world | 22:25 |
| fenn | which is unfortunately very widely prevalent in the modern era | 22:26 |
| kanzure_ | pardon? | 22:26 |
| kanzure_ | it's true that a person is a biological system | 22:26 |
| fenn | people are not infrastructure, a man is worth more than what use he is to you | 22:26 |
| fenn | otherwise why bother to do anything at all? | 22:27 |
| kanzure_ | oh, well, "people are infrastructure" could be interpreted to mean that I'm a nasty businessman making broad generalizations | 22:27 |
| kanzure_ | 'infrastructure' has connotations related to goals, doesn't it? hm | 22:28 |
| kanzure_ | but on the other hand I could argue to you a von Neumann probe that contains 20 different phenotypes of peoples | 22:28 |
| kanzure_ | bah, this is just semantics. sorry. | 22:29 |
| fenn | some people think the whole universe belongs to them personally, at the expense of everything else | 22:29 |
| fenn | i don't think that's right | 22:29 |
| kanzure_ | heh, remember my troubles with 'property rights' | 22:30 |
| kanzure_ | (although I'd like to figure out something better than "first come first serve" for matter acquisition) | 22:31 |
| nsh | how bout | 22:31 |
| fenn | say your embryos on the probe grow up and decide they don't really care about exploring the universe after all, and blow up your probe, is this 'good' or 'bad'? | 22:31 |
| nsh | everyone grows the fuck up and learns to share | 22:31 |
| nsh | i think that might work | 22:31 |
| nsh | whoa, embryos on a probe?!? | 22:32 |
| * nsh wonders what is happening in this conversation | 22:32 | |
| fenn | nsh: kanzure is considering using humans or modified humans to direct the replication of von neumann probes in other star systems | 22:32 |
| fenn | because it's easier (maybe) to grow humans than to grow computer chips or whatever | 22:33 |
| nsh | oh | 22:33 |
| * nsh is dubious | 22:33 | |
| * fenn to | 22:33 | |
| nsh | how much intelligence does a von neumann probe really need? | 22:33 |
| kanzure_ | well, you have to remember what I'm doing here | 22:33 |
| nsh | you could use a drosophila brain | 22:33 |
| fenn | nsh: more than you think.. could you bootstrap civilization on your own? from a bunch of interstellar debris | 22:34 |
| kanzure_ | the reason why I'm thinking about humans at the moment is because I'm interested in extending the human architecture to perhaps be its own sort of von neumann probe and so on | 22:34 |
| nsh | fenn, good point | 22:34 |
| kanzure_ | it's not really that; I have trouble wording it | 22:34 |
| fenn | you want to change the definition of human, simple enough | 22:35 |
| kanzure_ | heh | 22:35 |
| fenn | present biological humans are unsuited for interstellar travel | 22:35 |
| kanzure_ | but but | 22:35 |
| kanzure_ | no, that's not what I want | 22:35 |
| kanzure_ | robots are for interstellar travel, really | 22:35 |
| kanzure_ | I mean, that's what we should use for that | 22:35 |
| fenn | eh, why? | 22:36 |
| * fenn wonders what we're having a discussion about | 22:36 | |
| kanzure_ | nothing really | 22:36 |
| kanzure_ | hold on | 22:36 |
| * kanzure_ fixes | 22:36 | |
| * fenn hands kanzure a can of "aspie's best friend" the new brain stimulator drink | 22:37 | |
| kanzure_ | did we already slash down my idea of updating one's all-too-human architecture? i.e., grow a few clones, keep them adequately nutritioned, etc. ? Plus the usual ideas on apt-get for genes and technologies to throw into the exoskeleton or one's (mobile? please??) fablab | 22:43 |
| kanzure_ | erm -> re: growing clones. It occurs to me that we should be smart enough to pre-fetch all of the food necessary to sustain a child for 20 years. (egg yoke) (yolk?) | 22:44 |
| fenn | 20 years is a lot of food | 22:45 |
| kanzure_ | let's calculate it | 22:45 |
| fenn | easier to just graft a few leaves on :) | 22:45 |
| kanzure_ | take the numbers from intravenuous feeding tubes for patients per day | 22:45 |
| kanzure_ | on a healthy diet | 22:45 |
| kanzure_ | a few leaves is hardly enough for autotrophy and humans | 22:45 |
| kanzure_ | there's a guy out on the net who wants to become photosynthetic | 22:45 |
| kanzure_ | I think he started Orion's Arm. | 22:45 |
| fenn | not if you have a nuclear reactor providing plenty of light and cooling | 22:46 |
| kanzure_ | fusor.net :p | 22:46 |
| fenn | there's a guy in india who stares at the sun for two hours a day, and hasnt eaten in twenty years | 22:46 |
| kanzure_ | a statue? | 22:46 |
| fenn | no, he's a jainist | 22:46 |
| kanzure_ | I haven't heard of this one. | 22:47 |
| kanzure_ | Anyway, 20 years of food might turn out to be a few thousand gallons of nutritional slurp | 22:47 |
| nsh | he's a jainist? the guy with the hilarious website? | 22:48 |
| nsh | i thought he was central european and christian | 22:48 |
| nsh | (inverted commas) | 22:48 |
| fenn | Hira Ratan Manek | 22:48 |
| nsh | ah, different guy | 22:49 |
| kanzure_ | so we didn't slash it down? | 22:49 |
| kanzure_ | good | 22:49 |
| kanzure_ | that's what we're talking about :p | 22:49 |
| fenn | what is your idea, in detail? | 22:51 |
| kanzure_ | is there a way to explain an idea without making it sound goal-oriented? | 22:54 |
| fenn | yes, most scientific theories are like this | 22:56 |
| kanzure_ | hm. since I suck at describing things when I try to be coherent, I'm going to try to be as incoherent as possible | 22:56 |
| kanzure_ | what I want to do is build my typical fablab setup, get my DNA synthesizer and my metal cutting tools and so on | 22:57 |
| fenn | hint: remove i/we/you words and use passive tense | 22:57 |
| kanzure_ | oh, maybe calling it a 'family' is not too terrible | 22:58 |
| nsh | Eliezer is a lot more stomachable when he's writing about physics | 22:58 |
| nsh | (than AI) | 22:58 |
| kanzure_ | I think what I'm making might be a 'transhuman transitional system' ;-) but I'm not sure | 22:59 |
| kanzure_ | namely because there's no "self-improvement" necessarily in it | 22:59 |
| kanzure_ | i.e., if a user had apt-get install for genes, that's no guarantee of any phenotypic "improvement" | 22:59 |
| kanzure_ | (apt-get install in either (1) pre-embryonic stages of development or (2) viral gene therapy as well -- both are technologically feasible) | 23:00 |
| kanzure_ | but it's also a 'family' in that the user is actively backed up with little experiments running around | 23:00 |
| kanzure_ | hrm, this isn't making any more sense | 23:02 |
| kanzure_ | actually, it really should ever since I mentioned that it's skdb applied on a smaller scale | 23:04 |
| kanzure_ | and with the mentions of 'becoming a system administrator of the self' from a while back | 23:04 |
| kanzure_ | damn, why does it keep turning off at 70% completion of the debian installation? | 23:06 |
| fenn | i had a computer that got hit by lightning that did that.. it also turned off if you moved the mouse too much or plugged anything into USB | 23:07 |
| kanzure_ | I think it might be overheating | 23:07 |
| kanzure_ | I asked the ##hardware guys whether I could have applied too little thermal paste | 23:08 |
| fenn | it had weird voltages and would do it more when it was cold in the room | 23:08 |
| kanzure_ | and they sai no .. | 23:08 |
| kanzure_ | *said no | 23:08 |
| kanzure_ | I think they might be telling lies, so I'll go pile the thermal paste on it | 23:08 |
| fenn | usually less thermal paste is better than more | 23:08 |
| fenn | you want as much metal to metal contact as possible, paste just gets in the way | 23:08 |
| kanzure_ | even if it's not metal? | 23:08 |
| kanzure_ | oops | 23:08 |
| kanzure_ | even if it's not even? | 23:08 |
| fenn | uh, what are you pasting? | 23:09 |
| kanzure_ | on top of the processor? | 23:09 |
| fenn | processor is metal | 23:09 |
| kanzure_ | right | 23:09 |
| kanzure_ | I asked if "even if it's not even" | 23:09 |
| fenn | er.. i dont know what you're asking | 23:09 |
| kanzure_ | i.e., maybe I didn't apply it smoothly | 23:09 |
| fenn | thermal paste just fills in the cracks so it's even | 23:09 |
| fenn | ideally you'd lap the heatsink so it's perfectly flat and use no paste | 23:10 |
| kanzure_ | lap? | 23:10 |
| fenn | a grinding process | 23:11 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!