| --- Day changed Wed Jul 30 2008 | ||
| wrldpc | eh | 00:23 |
|---|---|---|
| wrldpc | intrigue. | 00:23 |
| kanzure | "the wayforward machine" | 02:30 |
| wrldpc | i got golden helix | 02:43 |
| kanzure | eh? | 03:00 |
| kanzure | hey, uh | 03:08 |
| kanzure | joining #singularity on oftc.net wouldn't be a terrible idea right about now | 03:09 |
| kanzure | anyone awake? | 03:15 |
| -!- wrldpc [n=wrldpc@pool-71-174-89-247.bstnma.fios.verizon.net] has quit [Read error: 113 (No route to host)] | 04:43 | |
| -!- wrldpc [n=wrldpc@pool-71-174-89-247.bstnma.fios.verizon.net] has joined #hplusroadmap | 04:47 | |
| -!- Phreedom [n=freedom@ip-194-50-167-184.mir.dn.ua] has joined #hplusroadmap | 04:52 | |
| -!- jm [n=jm@p57B9CF9D.dip.t-dialin.net] has joined #hplusroadmap | 08:47 | |
| -!- wrldpc [n=wrldpc@pool-71-174-89-247.bstnma.fios.verizon.net] has quit [] | 09:39 | |
| -!- Phreedom [n=freedom@ip-194-50-167-184.mir.dn.ua] has quit [Read error: 110 (Connection timed out)] | 10:52 | |
| -!- Phreedom [n=freedom@ip-194-50-167-184.mir.dn.ua] has joined #hplusroadmap | 11:06 | |
| -!- jm|earth [n=jm@p57B9D8ED.dip.t-dialin.net] has joined #hplusroadmap | 12:23 | |
| -!- jm [n=jm@p57B9CF9D.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] | 12:40 | |
| -!- Phreedom [n=freedom@ip-194-50-167-184.mir.dn.ua] has quit [Read error: 110 (Connection timed out)] | 12:54 | |
| -!- jm|earth [n=jm@p57B9D8ED.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] | 13:45 | |
| kanzure | Hm. | 13:48 |
| kanzure | " You can't possibly get a good technology going without an enormous number of failures. It's a universal rule. If you look at bicycles, there were thousands of weird models built and tried before they found the one that really worked. | 14:01 |
| kanzure | You could never design a bicycle theoretically. Even now, after we've been building them for 100 years, it's very difficult to understand just why a bicycle works - it's even difficult to formulate it as a mathematical problem. But just by trial and error, we found out how to do it, and the error was essential.[5]” | 14:01 |
| kanzure | Where are all of the prototypes for the transistor? | 14:02 |
| -!- ybit [n=h@unaffiliated/ybit] has quit [Remote closed the connection] | 14:06 | |
| -!- ybit [n=h@unaffiliated/ybit] has joined #hplusroadmap | 14:06 | |
| -!- jm [n=jm@p57B9FCF3.dip.t-dialin.net] has joined #hplusroadmap | 14:47 | |
| kanzure | http://www.crunchyroll.com/media-3409/Metal-Armor-Dragonar-Episode-1.html hm, I like the opening | 14:48 |
| * fenn mumbles something about heinlein | 14:57 | |
| kanzure | the server? | 15:19 |
| fenn | the lunar mass launcher blowing up geological landmarks on earth | 15:19 |
| kanzure | in the video? | 15:20 |
| fenn | it's straight out of 'the moon is a harsh mistress' | 15:20 |
| kanzure | heh | 15:20 |
| fenn | but then they went on some kind of gundam tangent and i dont really see the connection.. | 15:20 |
| kanzure | "Devised shortly after the release of Mobile Suit Gundam ZZ, Dragonar was intended to be both a "starter" series to get new fans into mecha anime, and a potential successor to the Gundam franchise. " | 15:21 |
| fenn | yeah the gundam storyline and science got really complex | 15:23 |
| fenn | minkowski particles and all that | 15:23 |
| fenn | er, minovsky | 15:24 |
| kanzure | right, so my understanding of gundam is incredibly limited | 15:25 |
| kanzure | but I do remember it looking like the actual shows spanned a decade or so? | 15:25 |
| kanzure | http://en.wikipedia.org/wiki/Gundam#Animated.2Flive-action_series_and_movies | 15:26 |
| kanzure | 1979 to 2008, ftw | 15:26 |
| fenn | yes but gundam ZZ is probably the best | 15:26 |
| fenn | new stuff gets into sappy gay-boy romance stuff | 15:26 |
| fenn | and i'm just like.. wait a sec, isnt this a war story? | 15:27 |
| kanzure | war on romance | 15:27 |
| kanzure | rawr | 15:27 |
| kanzure | so somehow I've been swooped into this whole "save the world" group thing here | 15:35 |
| fenn | oops, Z not ZZ | 15:36 |
| kanzure | 85? | 15:36 |
| kanzure | http://en.wikipedia.org/wiki/Mobile_Suit_Zeta_Gundam | 15:36 |
| fenn | right | 15:37 |
| kanzure | first time I saw Gundam was whichever one had "Flying in the Sky" which might have implied G Gundam | 15:39 |
| kanzure | oh, fenn, it turns out that Blue Brain might actually be just mostly statistical in nature, not an actual molecularly relevant simulation yet and such | 16:21 |
| kanzure | I thought it was more than statistics when I watched Henry's talk, but Steve pointed out that there's technically little evidence suggesting there's anything but the waveforms mentioned around 36min running the system | 16:22 |
| kanzure | erm, I might be wrong on the time | 16:22 |
| kanzure | fenn: oh, here's an idea. let's just convert our pdf's into mpeg | 16:28 |
| kanzure | there's too much processing going on to just normally read the pdfs, so why not convert them into mpeg ? the mpeg players run well enough | 16:28 |
| kanzure | heh, there's a http://fold.it/ team for CASP8 | 16:35 |
| kanzure | cheaters :) | 16:35 |
| kanzure | http://heybryan.org/mediawiki/index.php/Low-hanging_fruit_for_human-level_artificial_intelligence_augmentation | 16:46 |
| fenn | "artificial intelligence augmentation" is confusing | 17:02 |
| kanzure | ai's augmentation to itself | 17:05 |
| kanzure | and yes, it is. | 17:05 |
| kanzure | http://heybryan.org/docs/Metal%20Armor%20Dragonar.avi | 17:20 |
| kanzure | help, I'm stuck in a loop | 17:33 |
| fenn | priority interrupt! | 17:34 |
| -!- wrldpc [n=wrldpc@pool-71-174-89-247.bstnma.fios.verizon.net] has joined #hplusroadmap | 17:35 | |
| -!- wrldpc [n=wrldpc@pool-71-174-89-247.bstnma.fios.verizon.net] has quit [Client Quit] | 17:36 | |
| fenn | why is anime the only scifi on television | 17:38 |
| fenn | i mean plausible scifi | 17:39 |
| fenn | not that silly star wars crap :) | 17:39 |
| kanzure | the stuff on tv is significantly worse than star wars | 17:41 |
| kanzure | dorm doesn't have anime-active channels | 17:41 |
| fenn | just as well, unless you like sailor moon | 17:42 |
| kanzure | the closest thing is like "Toon Disney" or "FOX: Family Guy" and so on. :( | 17:42 |
| kanzure | no, not even Sailor Moon | 17:42 |
| * kanzure admits to having a collection of Sailor Moon music for some reason | 17:42 | |
| * fenn admits to reading a vampire romance right now | 17:42 | |
| -!- kanzure_ [i=bbishop@66.112.232.157] has quit [Read error: 60 (Operation timed out)] | 17:43 | |
| kanzure | oh shit | 17:44 |
| kanzure | where'd my server go | 17:44 |
| kanzure | shit shit shit shit | 17:44 |
| kanzure | uhhh | 17:46 |
| fenn | you broke it! | 17:46 |
| kanzure | why didn't I have dyndns installed again? | 17:46 |
| fenn | because your update script was in /usr/bin instead of /usr/local/bin? | 17:47 |
| kanzure | because I didn't have one | 17:47 |
| kanzure | hurray | 17:48 |
| kanzure | it's back | 17:48 |
| kanzure | best stroke of luck ever | 17:48 |
| fenn | did it really just expire in the last hour? | 17:48 |
| -!- kanzure_ [i=bbishop@66.112.232.157] has joined #hplusroadmap | 17:48 | |
| fenn | same IP | 17:49 |
| -!- Endos [n=Endos@194.71.118.70.cfl.res.rr.com] has joined #hplusroadmap | 17:49 | |
| Endos | hey | 17:49 |
| fenn | oi | 17:49 |
| kanzure | Endos is someone who should have been in here for a while now but I neglected to mention the channel to him | 17:50 |
| kanzure | that sort of thing | 17:50 |
| Endos | well now I know | 17:50 |
| Endos | it should automatically join when I come on trillian now | 17:50 |
| kanzure | fenn: so I was trying to upload the mp3 http://heybryan.org/music2/%5bFamison8bit%20-%20by%20Momo-i%5d%2023%20-%20Moonlight%20Densetsu%20%5bInstrumental%5d.mp3 | 17:51 |
| kanzure | but anyway | 17:51 |
| kanzure | it's not that great | 17:51 |
| kanzure | just since you mentioned SM and I realized I had stuff | 17:51 |
| Endos | brb | 17:52 |
| kanzure | now if only I could get signifier to /connect and /join | 17:52 |
| Endos | back | 18:00 |
| kanzure | Endos: I'm playing around at http://heybryan.org/mediawiki/index.php/Low-hanging_fruit_for_human-level_artificial_intelligence_augmentation at the moment | 18:04 |
| kanzure | so not much is really happening in here | 18:04 |
| kanzure | and nothing much has been happening all the last week or so because I've been out and about | 18:04 |
| kanzure | but I think I left off with http://heybryan.org/fractal.html | 18:04 |
| kanzure | oh, something I just added to the page | 18:07 |
| kanzure | 'irrelevant); Proposal: mutation passing interface. Probably a language or specificiation of some sort based off of the constraints markup language that I mention in egan.html and maybe humancortex.html and definitely in the conversations from #hplusroadmap.' | 18:07 |
| kanzure | "mutation passing interface' | 18:07 |
| kanzure | MPI | 18:07 |
| kanzure | any laughs? come now .. | 18:07 |
| kanzure | I was proposing a constraints markup language for research to proceed, so passing around mutations between GAs or other applications of some sort, in some formalized manner, surely couldn't hurt | 18:07 |
| kanzure | I think the problem is that if it's truly a good mutation you're not going to be able to capture it | 18:08 |
| kanzure | these were the 'tools on the frontier' for dumping content into skdb methinks | 18:08 |
| kanzure | processing *then* dumping of course | 18:08 |
| kanzure | processing/packaging/etc | 18:08 |
| fenn | mutation ~== patch? | 18:09 |
| kanzure | why, yes | 18:10 |
| kanzure | ok, so that simplifies things a lot | 18:11 |
| -!- Netsplit zelazny.freenode.net <-> irc.freenode.net quits: ybit, nsh | 18:11 | |
| kanzure | hm, netsplit? | 18:11 |
| fenn | yep | 18:11 |
| fenn | so, a patch is somewhat different from a module in that the thing being patched has to have some common ancestry | 18:12 |
| -!- Netsplit over, joins: ybit, nsh | 18:13 | |
| kanzure | oh, that's true | 18:17 |
| kanzure | so there is certainly a difference worth pointing out | 18:18 |
| kanzure | I wonder what sort of framework we should have going | 18:18 |
| kanzure | I keep mentioning 'frameworks' for brains and such, either building them, compiling them from components found over the net and such, but I haven't addressed extensibility issues | 18:18 |
| kanzure | since there's no code basis of it anyway .. extendability seems always kinda tricky | 18:18 |
| fenn | i dont get "some laws of systematics" at all | 18:18 |
| kanzure | systemantics | 18:18 |
| kanzure | it's not directly applicable I guess | 18:19 |
| kanzure | but what don't you get specifically? some of the rules seem obvious to me | 18:19 |
| fenn | some of them seem wrong or just butt-headed | 18:19 |
| kanzure | oh? | 18:20 |
| kanzure | oh, wait, I remember something like this when I looked at them too, a few of them did seem like they were just angsty | 18:20 |
| fenn | well, they should take their own advice and just cut it down to "keep it simple" | 18:22 |
| kanzure | I was hoping I could use some of it as inspiration to ways to describe this so-called "low hanging fruit to ai self-augmentation" | 18:23 |
| fenn | the total amount of human energy in the universe is constant - wtf is this supposed to mean? | 18:23 |
| fenn | (law 3) | 18:24 |
| fenn | "Complicated systems seldom exceed five percent efficiency." wrong | 18:24 |
| kanzure | did you see the bicycle quote yesterday from Dyson? | 18:25 |
| Endos | lol - fenn i'm pretty sure that whoever wrote that just didn't want to quote the bible "Vanity, all is vanity... there is nothing new under the sun" | 18:25 |
| kanzure | basically he was saying that there were a ton of patches to bikes until something useful was constructed | 18:25 |
| kanzure | so I'm wondering how we could trace these patches ... | 18:25 |
| kanzure | hm, I'm sure somebody has written about the history of bicycles | 18:25 |
| kanzure | but I mean the patches for any system in general, like, a random boiler system in the factory down the street | 18:25 |
| fenn | kanzure: yeah, and the transistor probably went through a lot of failed design cycles before it was commercialized | 18:26 |
| fenn | but i dont think it's necessarily a law of nature, but more of a good rule of thumb | 18:26 |
| fenn | for example the first nuclear bomb worked | 18:27 |
| kanzure | isn't that failure-history kinda useful? | 18:27 |
| kanzure | eh, the first one? | 18:27 |
| fenn | yep | 18:27 |
| kanzure | what about the testing and experimentation they did? | 18:27 |
| Endos | (Just so there are no misconception, I consider myself atheist only because being agnostic would require putting too much effort into thinking about a god) | 18:27 |
| kanzure | they only declared it a nuclear bomb when they were ready to test the final design and putting together all of the components | 18:27 |
| fenn | trinity was the first bomb and it worked just as the theory predicted | 18:27 |
| kanzure | I'm pretty sure they had compartmentalized debugging before that | 18:27 |
| fenn | and the first nuclear pile before that | 18:28 |
| Endos | was trinity an implosion device or was it the tube design? I can't remember | 18:28 |
| kanzure | hm, what was that infinityplus story about augmented humans | 18:28 |
| kanzure | Stephen Baxter I think. | 18:28 |
| kanzure | Ted Chiang. | 18:30 |
| kanzure | http://www.infinityplus.co.uk/stories/under.htm | 18:30 |
| fenn | i liked ted chiang's "story of your life" | 18:31 |
| kanzure | http://www.amazon.com/Stories-Your-Life-Others-Chiang/dp/0765304198 | 18:32 |
| fenn | http://fennetic.net/pub/irc/Ted Chiang - Story of Your Life.pdf | 18:34 |
| fenn | hmm i think that's the right one | 18:34 |
| kanzure | 158k ? | 18:34 |
| fenn | its not too long | 18:37 |
| kanzure | oh, while randomly browsing fenn's /pub/irc/ I saw a mention of konq again I remembered signifier's suggestion to check out the gtk browser 'kazehakase' whch seems to be another c-compiled browser but it's at least something I missed when going through the ol' list of browsers to hope don't suck | 18:38 |
| kanzure | http://kazehakase.sourceforge.jp/ | 18:38 |
| kanzure | http://kazehakase.sourceforge.jp/screenshots/ss20040828.png | 18:39 |
| fenn | is "Understand" about the two people taking some kind of alzheimers medication that makes them super intelligent? | 18:39 |
| * kanzure still wants a perl-based browser | 18:39 | |
| kanzure | fenn: yeah, are you asking to see if you remember? | 18:39 |
| fenn | right | 18:39 |
| kanzure | or if you can skip it? it's worth not skipping | 18:39 |
| kanzure | ah | 18:39 |
| kanzure | okaty | 18:39 |
| fenn | it's pretty depressing | 18:39 |
| fenn | can't we all just get along? | 18:40 |
| kanzure | what? | 18:40 |
| kanzure | oh | 18:40 |
| kanzure | right | 18:40 |
| kanzure | two months of an advantage producing that much of a difference :) hehe | 18:40 |
| kanzure | that's kinda what these guys (that I'm living with now) are fearful about though, right? | 18:40 |
| kanzure | "well, if bad ai is first then game over" etc. | 18:40 |
| kanzure | I tried to come up with a way to verbalize my objections to the 'we are all doomed' enthusiasts, and I came up with something somewhat crappy but somewhat not crappy, let me upload | 18:43 |
| kanzure | http://heybryan.org/eli.html was sent to someone across the room (since I fail at verbalizing this for some reason -- though less so in text?) so that's why it has strange context | 18:43 |
| kanzure | I know that I'm kinda an island in here about those disagreements | 18:43 |
| kanzure | esp. since I think fenn has some disagreements with my disagreements :-) (right?) | 18:44 |
| kanzure | meep | 18:47 |
| kanzure | maybe it would be best to just say "if ufai versus fai is such a big deal and a wrong move is so big an issue, let's go find another way" but I don't know if that directly maps to the stuff I wrote in eli.html | 18:48 |
| fenn | sounds like you are having fun | 18:55 |
| kanzure | hm? | 19:02 |
| kanzure | you totally glazed over my eli.html argument | 19:03 |
| kanzure | perhaps intentionally? | 19:03 |
| * fenn reads it for the third time | 19:06 | |
| fenn | are you saying that improving the brain is more likely to yield a friendly intelligence because it derives from a supposedly friendly source? (human mind) | 19:08 |
| fenn | "the main problem with trying to stop the emergence of any technology is that it's an external 'improvement' rather than an internal improvement in your own designs for the extrinsic risks and UFAI" | 19:09 |
| kanzure | are you saying -> no. | 19:09 |
| kanzure | for the moment I think ""if ufai versus fai is such a big deal and a wrong move is so big an issue, let's go find another way [of doing what?]"" is a better way of saying it | 19:10 |
| kanzure | the /what/ I think might be summarized by http://heybryan.org/fractal.html but people can still say "ok, sure, but ufai will still kill you" | 19:11 |
| kanzure | the step required to do fai/ufai is kinda important, if you fail you might get ufai (or no ai, but whatever); maybe there's another way to make sure you don't die and continue on into this 'tech future' and so on, without making such a big gamble, while still surviving even if (u)fai does turn up. | 19:13 |
| kanzure | basically I like to s/ai/natural intelligence/ and just picture Hitler or even myself ... I mean, if you really wanted to blow the planet up, how hard is it to make a nuclear weapon, eh? | 19:13 |
| fenn | i think the issue is that nobody knows where the hard takeoff line is | 19:13 |
| fenn | so a ufai could just materialize out of some more or less unrelated projects | 19:14 |
| fenn | remember charles stross' lobsters | 19:15 |
| kanzure | "but then you just have a giant jupiter brain monitoring everything and becoming creative enough to stop those emergent/materializing situations" | 19:15 |
| kanzure | eh, and all the while we didn't know if that jupiter brain wasn't a case itself too | 19:15 |
| kanzure | erm | 19:15 |
| kanzure | but it's important to note that I'm not so much interested in "anti authority, oh teh noes a dictatorship" | 19:15 |
| kanzure | (it's an important point, but I suspect it's not my central) | 19:16 |
| fenn | i think of friendliness as a sort of golden ratio | 19:16 |
| fenn | the ai should be able to figure it out eventually if we start close enough | 19:17 |
| fenn | but in the end it's only a few extra decimal places | 19:17 |
| kanzure | to me it just looks like a race for the local light sphere or something ... screw ai, just s/ai/any superawesome blackswan/ ... have to be able to isolate the cancer | 19:18 |
| kanzure | grr, I'm not being coherent | 19:18 |
| kanzure | maybe I get to cite the semantic forest issues re: intelligence and how I'm still not certain if it's intelligence or just lots of patches leading up to something interesting | 19:19 |
| kanzure | hm, action potentials == patches? "the minimal information to update the perceptual models implemented in the brain" as Markram said | 19:23 |
| kanzure | perhaps not quite a direct mapping | 19:23 |
| kanzure | i.e., other neurophysiological stuff going on in there of course | 19:23 |
| ybit | hey fenn and kanzure | 19:49 |
| kanzure | you'll never take me alive | 19:50 |
| * kanzure hides | 19:50 | |
| ybit | haha | 19:50 |
| ybit | how's it going in ¿s.f.? | 19:50 |
| * ybit needs to read the above log, brb | 19:50 | |
| kanzure | we're all a bunch of aspies sitting around complaining about compiler errors all day | 19:50 |
| fenn | compiler? what's that? | 19:53 |
| fenn | you mean you're using one of those old dead languages? | 19:53 |
| kanzure | back in the old days, symbol lookups costed 15 cents, and a system call? a nickle. | 19:56 |
| kanzure | *cost | 19:56 |
| fenn | real men write their own evaluation loop | 19:57 |
| kanzure | hm, that's an obscure reference | 19:58 |
| kanzure | http://www.google.com/search?num=100&q=%22evaluation+loop%22+compiler | 19:58 |
| kanzure | one of the results is 'Design of a LISP-Based Microprocessor' heh' | 19:58 |
| kanzure | oh, looks like something in interpreters | 19:59 |
| kanzure | http://lambda-the-ultimate.org/node/965 wondering where this might have come from | 19:59 |
| fenn | it was more or less a lisp reference | 20:02 |
| fenn | i wish it applied to the rest of reality | 20:03 |
| ybit | aand i´m back, shortly | 20:04 |
| ybit | i think my brain exp.ed more damage reading the logs than not ;) | 20:05 |
| ybit | i must say this made me laugh: http://pastebin.ca/1087813 | 20:06 |
| fenn | heh that was a netsplit | 20:07 |
| ybit | ssshhh | 20:08 |
| ybit | this is similar to what you are doing there kanzure: http://www.aleph.se/andart/archives/2008/07/the_graphs_of_war.html | 20:09 |
| ybit | not quite, but it deals with doom, war that is | 20:10 |
| kanzure | hm | 20:13 |
| kanzure | http://correlatesofwar.org/ | 20:15 |
| kanzure | hm | 20:38 |
| kanzure | nobody should see references or the bibtex | 20:39 |
| kanzure | unless they absolutely have to | 20:39 |
| kanzure | otherwise the fetches should be mostly automatic etc. | 20:39 |
| -!- jm [n=jm@p57B9FCF3.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] | 20:53 | |
| fenn | "The world can’t end in 1997 or 2029 because it already ended in 1999 when the Day of Lavos occurred." | 22:12 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!
