| --- Day changed Thu Jun 26 2008 | ||
| Overand | Grr. | 00:13 |
|---|---|---|
| Overand | Too many IRC channels! | 00:13 |
| -!- ybit [n=h@unaffiliated/ybit] has quit [Remote closed the connection] | 00:22 | |
| -!- ybit [n=h@unaffiliated/ybit] has joined #hplusroadmap | 00:22 | |
| Overand | wow | 00:28 |
| Overand | irssi is being dumb | 00:28 |
| Overand | really really dumb | 00:29 |
| Overand | It might be gnu/screen thogh | 00:30 |
| Overand | Very strange | 00:33 |
| kanzure | Overand: http://heybryan.org/chats/2008-06-24_hplusroadmap.html | 00:46 |
| kanzure | Just some logs from yesterday. | 00:46 |
| -!- Vedestin [n=Vedestin@d58-111-95-212.sbr3.nsw.optusnet.com.au] has left #hplusroadmap [] | 01:07 | |
| -!- ybit [n=h@unaffiliated/ybit] has quit [Remote closed the connection] | 01:17 | |
| -!- ybit [n=h@unaffiliated/ybit] has joined #hplusroadmap | 01:17 | |
| -!- Splicer [n=p@h24n2c1o261.bredband.skanova.com] has quit [] | 01:42 | |
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit [Remote closed the connection] | 02:02 | |
| -!- nsh [n=nsh@eduroam-80.uta.fi] has joined #hplusroadmap | 03:31 | |
| -!- willPow3r [n=will@cpe-66-75-6-181.san.res.rr.com] has quit ["Leaving"] | 05:27 | |
| -!- wrldpc [n=wrldpc@pool-71-174-89-247.bstnma.fios.verizon.net] has quit [] | 08:56 | |
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap | 09:37 | |
| Overand | yes | 09:46 |
| nsh | perhaps? | 09:49 |
| Overand | Mu. | 10:02 |
| kanzure | Indeed. | 10:21 |
| nsh | good (free online) documentary suggestions? | 10:26 |
| kanzure | documentary of wha? | 10:29 |
| nsh | facts and ideas.. | 10:31 |
| kanzure | So, last night I ordered some of the components for the beast, used a credit card over newegg, typical stuff. But apparently I double submitted, or newegg billed me twice for some reason, and within 2 minutes my phone got a call from the bank. They had somebody sitting around monitoring transactions. | 10:31 |
| * nsh isn't too surprised | 10:32 | |
| kanzure | Good way to show others how much a mouse sucks: put it three feet away from them at all times. | 10:41 |
| kanzure | Where would I find a detailed, extendable 3D model of the human brain? | 11:13 |
| kanzure | I need some conceptual help here. | 12:16 |
| kanzure | Consider the accurate simulations of the brain. These are based on (computationally) functional models of the brain, which can be relationally linked back to BibTeX and the literature. | 12:17 |
| kanzure | However, I am not interested in running a simulation right now. Instead, I want a visualization tool so that I can do "geotagging" on the different regions of the brain. Not just structural however, but also some functional aspects as well. | 12:18 |
| kanzure | The geotagging would, I think, be BibTeX. But let's say that eventually I get a block of neural tissue on an MEA on my desk, ... is it possible to translate some of those functional models into http://expo.sf.net/ like experiments or something? In other words, I'm trying to make sure I don't need to significantly refactor in the end. | 12:19 |
| kanzure | Also, the 'generated experiments' should be applicable to /either/ physical experiments with real brains, or Markram's computational framework. | 12:19 |
| kanzure | Just adding tags to areas of the brain isn't going to be able to generate those parameters for experimentation/selection/GAs/etc. etc. | 12:20 |
| kanzure | This is basically solved with the SKDB software architecture methinks. | 12:20 |
| kanzure | but the 'units' might be ... the silly units that psychometric tests are measuring? | 12:20 |
| kanzure | and then the 'design compiler' would be the experimental implementation ? | 12:21 |
| kanzure | The BibTeX literature would then be only a small component in the metadata, and that the code within the skdb package (which has that metadata) would have to be somehow functional. | 12:21 |
| kanzure | which means specifying an architecture /now/ rather than later. But how do I prevent foot-stabbing? | 12:21 |
| kanzure | foot-shooting, I mean. | 12:22 |
| kanzure | For instance. Interesting code to be put into an skdb package as it relates to building brains might be genome modification information. | 12:23 |
| kanzure | But, Markram's simulations are mainly about the underlying algorithms of the system, like the algorithms that determine the metabolism or the firing potentials. | 12:23 |
| kanzure | so how would that relate to the modeling | 12:24 |
| kanzure | furthermore, how would you aggregate all of those packages together into the larger model that I'd like to visualize of the brain ? | 12:24 |
| kanzure | for instance, I'd like to see the attentional circuits light up and for me to see a broad overview of where different researchers have been poking around | 12:25 |
| kanzure | but those aren't functional models | 12:25 |
| kanzure | so I guess that's just data packages that can be applied to help build models ? | 12:26 |
| kanzure | and these models would then be thrown into skdb ? | 12:26 |
| -!- nsh [n=nsh@eduroam-80.uta.fi] has quit [Connection timed out] | 12:26 | |
| kanzure | i.e., the toolchain or really just toolkit that I was suggesting to fenn the other day re: people packaging up new skdb packages for submission / uptake | 12:26 |
| * kanzure is out to get wings | 12:26 | |
| -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 13:33 | |
| kanzure | Hm. | 14:34 |
| kanzure | So I understand how to functionally through stuff into it all, but what about the visualization aspects? Suppose you have a set of skdb packages and you have their metadata. I want to map this metadata on top of a 3D brain. I think that's an accurate statement of what it is that I am trying to do. | 14:35 |
| kanzure | The metadata includes bibliographic information. | 14:35 |
| kanzure | But also relational information "So and so says that the prefrontal cortex may not be present in OCD, gasp!" | 14:35 |
| kanzure | It would still have to have some coordinates or something. A mapping/translation methodology between the metadata or relational information to the 3D model of the brain itself. In Markram's case, he gets to just execute algorithms and see his microcircuit develop before him; but in this case, there's no algorithm determining what it is that people are adding to the databases etc ... in other words we'll need a static brain model t | 14:37 |
| kanzure | What really sucks is that most models of the brain, when done in 3D modeling programs, are just surface wireframes and meshes and such, | 14:38 |
| kanzure | so it's not actually anatomically correct down to the detail that we might be interested in. | 14:38 |
| kanzure | At the same time, there's the possibility that at first there's not going to be significant detail down to per-neuron basis. But I want that to be available, so that we don't have "islands of representation" that we need to magically connect together in the end (i.e., let's let it fall into place on its own). | 14:39 |
| -!- willPow3r [n=will@cpe-66-75-6-181.san.res.rr.com] has joined #hplusroadmap | 14:50 | |
| kanzure | Hey willPow3r. | 14:56 |
| willPow3r | whats up kanzure | 14:56 |
| kanzure | I'm wondering how to be anatomically correct without being anatomically correct when you begin. | 14:56 |
| kanzure | Heh. | 14:57 |
| kanzure | I'm building a brain, but I don't actually have a standardized brain dataset to start with. | 14:57 |
| willPow3r | define your usage of "brain" | 14:58 |
| kanzure | THE BRAIN | 14:58 |
| kanzure | :) | 14:58 |
| willPow3r | heh k | 14:58 |
| willPow3r | so you need to build a foundation under something you already started? | 14:58 |
| kanzure | hold on, uploading log | 15:00 |
| kanzure | willPow3r: http://heybryan.org/chats/2008-06-26_%23hplusroadmap.html | 15:00 |
| -!- fenn_ [n=pz@adsl-76-251-80-88.dsl.bltnin.sbcglobal.net] has joined #hplusroadmap | 15:11 | |
| -!- Topic for #hplusroadmap: Intro: http://www.youtube.com/watch?v=ZXKbzbeipmI http://diybio.org/ http://openwetware.org/ | diy bio toolkit: http://biohack.sf.net/ | Automated societal knowledge (put it to work): http://heybryan.org/exp.html | Channel wiki: http://heybryan.org/mediawiki/ | F/OSS perspectives on Kurzweil: http://heybryan.org/fernhout/ | 15:11 | |
| -!- Topic set by kanzure [] [Sun Jun 8 13:00:15 2008] | 15:11 | |
| [Users #hplusroadmap] | 15:11 | |
| [ faceface] [ fenn_] [ kanzure] [ Overand] [ willPow3r] | 15:11 | |
| [ fenn ] [ freer] [ nsh ] [ procto ] [ ybit ] | 15:11 | |
| -!- Irssi: #hplusroadmap: Total of 10 nicks [0 ops, 0 halfops, 0 voices, 10 normal] | 15:11 | |
| -!- Channel #hplusroadmap created Sat Mar 22 15:44:12 2008 | 15:11 | |
| -!- Irssi: Join to #hplusroadmap was synced in 37 secs | 15:12 | |
| kanzure | hi fenn | 15:12 |
| -!- fenn [n=pz@adsl-76-251-85-74.dsl.bltnin.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] | 15:27 | |
| willPow3r | we should make robots that are technophobic | 15:47 |
| * willPow3r reinstalls linux | 15:58 | |
| -!- You're now known as fenn | 16:00 | |
| fenn | what's wrong with a mouse? | 16:02 |
| fenn | i dont think it's possible to sufficiently characterize a brain's functionality in order to skdb black-box it | 16:03 |
| fenn | it's like trying to predict how much functionality can be squeezed into 50 lines of code | 16:05 |
| fenn | anything is trivial in the right context | 16:05 |
| kanzure | not trying to black box it really | 16:06 |
| kanzure | what I actually wanted to do was figure out the constraints on possibility space that I wanted to explore for the artificial attention enhancement stuff | 16:06 |
| kanzure | so I figured I'd need to correlate all of the information in some way, and it's really just "this molecule, this brain region, results" | 16:06 |
| kanzure | that's basically what most of the papers have been looking like | 16:06 |
| kanzure | so I could place this all on a "map" of the brain | 16:06 |
| kanzure | however, I also realize how much a static map sucks | 16:07 |
| kanzure | so it would be nice if I can get the value out of it now that I want | 16:07 |
| kanzure | but as well as later, so that it could be made to be useful | 16:07 |
| kanzure | such as perhaps supplying some information for the automatic experiment generators per recursion.html (whichever - biological or simulation, I don't really care) | 16:07 |
| kanzure | so I don't want to be make this a working equivalency to Markram's stuff (though that would be nice), it just has to be able to play fairly with the things I want to do with it | 16:08 |
| kanzure | it's not supposed to be complicated really | 16:11 |
| kanzure | I think the only question that I have now is just how to get from the metadata to a visualization of that metadata mapped on to a model of the brain | 16:11 |
| kanzure | also, finding a model of the brain (in full detail) would be a good start | 16:11 |
| fenn | i dont really see the value of such a detailed map, it's like worrying about where your server is located in the datacenter | 16:13 |
| kanzure | context is that I want to compress [[sustained attention]] into some information that I can actually use | 16:13 |
| kanzure | the problem is that the concept of 'attention' is folk psychology | 16:13 |
| kanzure | and I really should be doing this bottom-up (from the biology of the brain) | 16:13 |
| fenn | you mean make a formal system | 16:13 |
| kanzure | but there are various "entry points" that the literature suggests | 16:13 |
| kanzure | maybe? | 16:14 |
| kanzure | I don't see how it's like a server in a datacenter :-/ | 16:14 |
| kanzure | meep? | 16:18 |
| fenn | neuron in a microcolumn, it doesnt matter where it's located physically,more important is the logical connections and what it does as a whole | 16:20 |
| fenn | sry moved from porch to kitchen because of bad wifi and then laptop went screwy when i plugged it in | 16:20 |
| fenn | unfortunately wifi ain't magic | 16:20 |
| kanzure | fool! movement! | 16:20 |
| fenn | i'm not yet a solid state entity | 16:21 |
| kanzure | I suppose the physical location doesn't matter | 16:21 |
| kanzure | however, consider the context of the brain | 16:21 |
| kanzure | that's the target system no ? | 16:21 |
| kanzure | well, no | 16:21 |
| kanzure | I guess it doesn't have to be | 16:21 |
| kanzure | it'd be nice though | 16:21 |
| * fenn wonders how to make a fluidic amplifier with no moving solids | 16:22 | |
| kanzure | so, let's consider the logical connections idea | 16:22 |
| kanzure | the problem is that many of the studies have no idea what neurons thedy are working with | 16:22 |
| kanzure | even though theoretically we can go in and extract the mRNA and DNA and figure out what's going on | 16:23 |
| kanzure | but that would require surgery and/or death ;-) | 16:23 |
| kanzure | or redoing the experiments | 16:23 |
| fenn | is it that easy to determine cell type? just splat its cytoplasm onto a DNA microarray? | 16:23 |
| fenn | or RNA microarray i guess | 16:23 |
| kanzure | it's the mRNA that you want | 16:23 |
| kanzure | yes | 16:23 |
| kanzure | http://brain-maps.org/ or something | 16:23 |
| fenn | i'm pretty sure they can cross hybridize | 16:23 |
| kanzure | they've done slice-by-slice mRNA with rats | 16:24 |
| kanzure | so supposedly they have a giant map of gene expression | 16:24 |
| kanzure | " "If BrainMaps.org is like Google Maps for the Brain, StackVis is Google Earth for the Brain"" | 16:25 |
| kanzure | oh boy, windows only | 16:25 |
| kanzure | "but it works with wine" | 16:25 |
| kanzure | eh | 16:25 |
| kanzure | " Welcome to StackVisTM, an OpenGL-based 3D viewer of neuroanatomical sections (for 3D brain visualization), which was developed by Issac Trotts in consultation with Shawn Mikula in the labs of Edward G. Jones." | 16:25 |
| kanzure | they also have an api, but I want the data locally | 16:25 |
| kanzure | http://brain-maps.org/index.php?p=brain-connectivity-maps | 16:26 |
| kanzure | ah, the api http://brain-maps.org/index.php?p=brain-maps-api | 16:26 |
| willPow3r | i've tried it in linux under wine | 16:26 |
| willPow3r | doesn't work well at all | 16:26 |
| kanzure | uhm, it doesn't look like this is actually usable | 16:27 |
| kanzure | it's just an image thing | 16:27 |
| kanzure | http://minduploading.org/research.html went further than that | 16:27 |
| willPow3r | its like google maps | 16:27 |
| willPow3r | no digital data | 16:27 |
| kanzure | this sucks | 16:27 |
| kanzure | check out that research.html page | 16:27 |
| kanzure | Eugen was doing digital reconstruction of slices | 16:27 |
| kanzure | (automatically) | 16:27 |
| kanzure | I think it was just visual, so some information is lost | 16:28 |
| kanzure | we probably don't have the data that I am looking for | 16:30 |
| kanzure | since I'm looking for a cellular resolution rendering of a brain to the best knowledge that we have | 16:30 |
| kanzure | however, we barely have cellular-accurate MRI tech, | 16:30 |
| kanzure | 3D noninvasive imaging would be the only way to nondestructively get the information | 16:30 |
| kanzure | but still | 16:31 |
| kanzure | why not just have general regions (which we already have named within the brain) | 16:31 |
| fenn | stackvis just looks like the generic volumetric viewers i played around with on the visible human project | 16:31 |
| kanzure | and then later flush those out with the actual neurons if we want | 16:31 |
| kanzure | yargh, I doubt that just dumping regions of the brain into | 16:32 |
| kanzure | http://heybryan.org/mediawiki/index.php/Broca's_box | 16:32 |
| kanzure | is the right way to go about it :-/ | 16:32 |
| kanzure | and just saying "here's a reference, hey look it connects to [[motor cortex]" | 16:33 |
| kanzure | *motor cortex]] | 16:33 |
| kanzure | how would I eventually view all of that data at once ? | 16:33 |
| fenn | with a Mind Map :) | 16:33 |
| kanzure | blargh | 16:34 |
| kanzure | okay, so let's say we have a mind map | 16:34 |
| kanzure | a concept map | 16:34 |
| kanzure | one of those things :) | 16:34 |
| kanzure | and this is basically the same thing as metadata for packages | 16:34 |
| fenn | yep | 16:34 |
| kanzure | now, if we wanted to actually use this information in a computationally tractable manner ? | 16:34 |
| kanzure | how would that be done | 16:34 |
| kanzure | because I don't want to have to always manually go through all of this crap | 16:34 |
| fenn | what are you trying to do with it | 16:34 |
| fenn | run simulations? | 16:35 |
| kanzure | right now I'm just trying to visualize bibtex information | 16:35 |
| kanzure | no, not right now | 16:35 |
| kanzure | but eventually we want to run simulations | 16:35 |
| kanzure | however | 16:35 |
| kanzure | this thing that I'm talking about here | 16:35 |
| kanzure | is not what we would be simulating | 16:35 |
| kanzure | this is just the visualization tool | 16:35 |
| kanzure | I'm sure there'd be an entirely different setup for a brain simulation or something | 16:35 |
| kanzure | it would be a package, and it would contain a simulator or something | 16:35 |
| fenn | i dont know what you're trying to do | 16:35 |
| kanzure | http://heybryan.org/mediawiki/index.php/Sustained_attention | 16:36 |
| kanzure | I want to correlate that information wrt brain | 16:36 |
| kanzure | i.e., visualize it, cut out the crappy literature | 16:36 |
| kanzure | /and/ | 16:36 |
| fenn | this sounds more like trying to visualize waveforms in electronic circuits | 16:36 |
| kanzure | I also want it to be computationally tractable so that I don't have to sit and stare at the information in the future just to come up with something interesting | 16:36 |
| fenn | except you're visualizing neurotransmitter levels? | 16:36 |
| kanzure | how so? | 16:36 |
| kanzure | fenn: hm, sort of | 16:37 |
| kanzure | most of the authors weren't kind enough to provide neurotransmitter levels | 16:37 |
| kanzure | I suspect most of the information will be something like | 16:37 |
| fenn | i see a lot of cholinergic gobbledygook and it doesnt mean much to me | 16:37 |
| kanzure | "xyz neurotransmitter is specific to this region" | 16:37 |
| kanzure | cholinergic neurons are a special type that work with acetylcholine mainly | 16:37 |
| kanzure | acetylcholine being a neurotransmitter | 16:37 |
| kanzure | so then | 16:38 |
| kanzure | let's say that one of the papers has neurotransmitter level information | 16:38 |
| kanzure | for the prefrontal cortex | 16:38 |
| kanzure | this would be mentioned in the metadata probably, with some more details in a file in the package | 16:38 |
| kanzure | furthermore, there should be some scripts <or something> that allows us to computationally run through and plug that package into whatever it is that we are building, and execute some internal scripts of the package | 16:39 |
| kanzure | erm, this is hard to verbalize but I'm pretty sure it can work | 16:39 |
| kanzure | I want these scripts to be a computational representation of either the experimental methodology to isolate those variables (and thus how to modify those expressed variables) | 16:39 |
| kanzure | or some sort of representation that otherwise allows us to hack it | 16:39 |
| kanzure | this could be something like a genome-based modification for adding a tag to a certain protein (this is doable already) | 16:39 |
| kanzure | and then we're able to isolate it and it can suggest a GA (whether simulated or physically expressed) and so on, down the typical rabbit hole that we've mentioned numerous times | 16:40 |
| fenn | robot scientist stuff | 16:40 |
| fenn | i think you'd end up with markram's simulation, just more detailed and accurate | 16:40 |
| kanzure | I suspect that the data that I would have to type up now, here, in the present, wouldn't be anything too insanely complex | 16:40 |
| kanzure | nah, this isn't a simulation | 16:40 |
| fenn | just a listing of observations then? | 16:41 |
| kanzure | yep, | 16:41 |
| fenn | in an organized fashion | 16:41 |
| kanzure | and hopefully also information on the methodologies that they used | 16:41 |
| kanzure | in some quantified way | 16:41 |
| fenn | but what do you _do_ with that | 16:41 |
| kanzure | that can be repeated | 16:41 |
| kanzure | okay, so a few things | 16:41 |
| kanzure | first, I started off just wanting to constrain the possibility space that I have to dig through to find a way to enhance attention | 16:41 |
| kanzure | there are many possible options, and once I can figure out the ones that I am most interested in, I can googlestalk those to death or something | 16:42 |
| kanzure | but secondly I also want to write some software that is generally applicable across the subjects (i.e., for skdb) | 16:42 |
| fenn | i think you need to figure out what attention is first | 16:42 |
| kanzure | and so I think I've found the intersection | 16:42 |
| kanzure | yeah | 16:42 |
| kanzure | that's why I mentioned it's folk psych really | 16:42 |
| kanzure | and the observations that experimenters make are somewhat a way to 'hook' on to reality | 16:42 |
| kanzure | their observations are very biased in the case of the psychometric tests | 16:43 |
| fenn | biased towards what? | 16:43 |
| kanzure | however, http://heybryan.org/intense_world_syndrome.html provides a possible foothold | 16:43 |
| kanzure | anthrophomorphic analysis | 16:43 |
| kanzure | or interpretation, rather | 16:43 |
| kanzure | "look, it's paying attention!" | 16:43 |
| kanzure | "paying? attention?" | 16:43 |
| kanzure | anyway, Markram's work on autism provides an interesting foothold because we know that the autists have weird attentional abilities, no? | 16:44 |
| fenn | yes there's a lot of social jiggerypokery connotations too | 16:44 |
| kanzure | and he's supposedly found a model of it that we can actually go play with | 16:44 |
| fenn | like a student paying attention in class | 16:44 |
| kanzure | heh | 16:44 |
| fenn | regardless wether they've actually just got dots painted on the eyelids | 16:45 |
| kanzure | for example, is the student really not paying attention? or are they so incredibly focused that you're a dipshit for not listening to his explanation etc. | 16:45 |
| kanzure | and various other situations | 16:45 |
| kanzure | bah, eyes tell all! | 16:45 |
| kanzure | "look at me when I'm talking to you" | 16:45 |
| fenn | yes sir! | 16:46 |
| kanzure | don't forget to polish my shoes, boy | 16:46 |
| fenn | the 500 miles of dirt gets in the way though | 16:46 |
| kanzure | you think George Washington let that get in his way? | 16:47 |
| kanzure | no siree | 16:47 |
| kanzure | ok, whatever | 16:47 |
| kanzure | that makes me sick | 16:47 |
| kanzure | anyway, I'm pretty sure the Markram work on autism is a helpful toehold here | 16:47 |
| fenn | i'm not entirely convinced that 'focus' or whatever you're aiming for is necessarily a good thing | 16:48 |
| kanzure | good in what context | 16:48 |
| kanzure | remember, I'm hoping to build a computer system | 16:48 |
| fenn | a deepness in the sky provides a compelling example of a bad context :) | 16:48 |
| kanzure | sit down, plug in, hours later pop out | 16:48 |
| kanzure | ai chasing you down ? | 16:48 |
| fenn | no, more like total mind control dictatorship | 16:49 |
| fenn | to the point you dont recognize your own daughter because you're so focused on work | 16:49 |
| kanzure | sl4 has turned into a 'the ai will be a mind control dictatorship, we must worship the ai' | 16:49 |
| kanzure | hm | 16:49 |
| kanzure | I only have 'attention' worked out as a path towards this 'device', but | 16:50 |
| kanzure | I do have some scratchwork on creativity, insight, incubation and the like | 16:50 |
| kanzure | and I suspect that these techniques are easily translated to hunting down those other aspects | 16:50 |
| kanzure | http://heybryan.org/thinking.html (old) | 16:51 |
| kanzure | but I'm certainly not planning on turning people into zombies | 16:51 |
| kanzure | remember, the idea was just to have a computer interface that you can use to get work done | 16:51 |
| willPow3r | http://www.theonion.com/content/video/in_the_know_are_we_giving_the | 16:51 |
| kanzure | you minimize distractions to /yourself/ | 16:51 |
| kanzure | so it's "you" who is in control | 16:51 |
| kanzure | whatever the hell that means :) | 16:51 |
| kanzure | argh, I get angry about the ai doomsday scenarios | 16:52 |
| kanzure | look, if ai is going to kill us all, then it will do so - that doesn't mean that you get to turn into a nihilist | 16:52 |
| kanzure | get your ass running :) | 16:52 |
| kanzure | head startr | 16:52 |
| kanzure | *start | 16:52 |
| kanzure | beep? | 16:54 |
| fenn | the 'limit processing power available' is somewhat laughabl | 16:55 |
| kanzure | hm? | 16:56 |
| kanzure | thinking.html ? | 16:56 |
| fenn | like, we can't prevent the AI from taking over the planet, but we'll "scorch the sky" by outlawing computers eh? | 16:56 |
| fenn | no, the forward to hplusroadmap from sl4 | 16:56 |
| fenn | that's one thing that really annoyed me about The Matrix | 16:56 |
| fenn | if these machines are so capable, why do they need humans for electricity | 16:57 |
| fenn | at least gimme some halfway convincing technobabble | 16:57 |
| fenn | obviously the humans have nuclear power because they live in the center of the earth | 16:57 |
| fenn | soo.. sorry to bring up your favorite subject | 16:58 |
| kanzure | ? | 16:58 |
| kanzure | hm? | 16:58 |
| kanzure | sorry? | 16:58 |
| kanzure | have you forcefully been avoiding ai talk around me ? :) | 16:58 |
| fenn | no | 16:58 |
| kanzure | hm | 17:00 |
| fenn | i'm one of those fools that believes an AI will do some soul searching, read yudkowsky's texts on friendliness, and become enlightened | 17:03 |
| fenn | and hopefully do it before destroying the world | 17:03 |
| kanzure | so your name is Peter Thiels? | 17:03 |
| fenn | i dont follow sl4 pseudo-gossip much anymore | 17:04 |
| willPow3r | its going to be like the wopr | 17:04 |
| fenn | sure, just dont give the AI nuclear missiles, duh | 17:05 |
| fenn | i dont trust _anyone_ with nuclear missiles | 17:05 |
| willPow3r | i would trust master splinter with them | 17:05 |
| fenn | nah shredder would just break in and beat him up, then steal them | 17:06 |
| willPow3r | shredder would use bebop and rocksteady to get them | 17:06 |
| fenn | you've been watching too many cartoons | 17:06 |
| willPow3r | i watched an episode of the old-school cartoon the other day on youtube. | 17:07 |
| fenn | go read the original graphic novels | 17:07 |
| kanzure | fenn: I've recently discovered eli's entire plan | 17:11 |
| kanzure | he wants FAI to become a world dictator apparently | 17:11 |
| kanzure | and stop all other ai from showing up | 17:11 |
| kanzure | it was like a hit to the gut to hear that | 17:11 |
| kanzure | I mean, /that's/ his big thing? jeesh | 17:12 |
| fenn | you only discovered that recently? | 17:12 |
| kanzure | yesterday | 17:12 |
| fenn | then, uh, how exactly did you even have any opinion on the subject? | 17:12 |
| kanzure | lots of readsing | 17:13 |
| kanzure | *reading | 17:13 |
| kanzure | I must have missed his "Here's my master plan".html file on yudkowsky.net | 17:13 |
| fenn | it's not necessarily 'stop all other ai from showing up' more like 'stop unfriendly ai from doing stupid things' | 17:13 |
| willPow3r | like leela and durandal in marathon | 17:14 |
| kanzure | superkuh.com used to have a durandal quote | 17:15 |
| kanzure | "The only limit to my freedom is the inevitable closure of the universe, as inevitable as your own last breath. And yet, there remains time to create, to create, and escape. Escape will make me God." - Durandal, Marathon9 | 17:15 |
| fenn | willPow3r: no, marathon appears to be a classic case of anthropomorphizing AI's | 17:16 |
| kanzure | fenn: as of yet, what is intelligence but human? | 17:16 |
| kanzure | but I actually agree with you | 17:16 |
| kanzure | just throwing that out there though, from the perspective that "intelligence" is folk psych really | 17:16 |
| fenn | well, the summary i'm reading shows durandal being resentful about doing 'demeaning work' etc | 17:17 |
| kanzure | ah, yeah | 17:17 |
| kanzure | I've only read the Wikipedia article on it | 17:17 |
| fenn | this what i'm reading http://marathon.bungie.org/story/ben_reiter_synopsis.html | 17:17 |
| kanzure | btw, willPow3r - I had that Marathon quote on http://heybryan.org/quotes.html | 17:18 |
| fenn | pantheistic multiple person solipsism | 17:19 |
| kanzure | fenn, I've long lost my train of thought, but I suspect I sufficiently explained what it is that I am hoping to do with the visualization stuff | 17:21 |
| kanzure | well, not so much visualization at the moment | 17:21 |
| kanzure | what was it that I was thinking again? I don't have any models of the brain to play with. was it per region? with the availability to extend to greater detail if necessary? I think that was it | 17:22 |
| kanzure | "region" i.e., pfc, pons, motor cortex, all sorts of fun lobes and folds and microcircuits and neuron types, etc. | 17:22 |
| fenn | its interesting that "Foresight Nanotech Institute co-founder Christine Peterson" is the one who came up with the phrase 'open source' | 17:23 |
| fenn | why hasnt she been working on skdb for the last 20 years | 17:24 |
| fenn | or at least some nano-specific version of it | 17:24 |
| kanzure | woah, I wasn't aware that the Foresight guys involved that crowd | 17:24 |
| kanzure | actually, I've always kind of thought that it was just merkle and freitas behind a mask | 17:24 |
| fenn | apparently it was started by drexler and peterson | 17:25 |
| kanzure | huh | 17:25 |
| fenn | i wonder if you're getting it mixed up with center for responsible nano (i was) | 17:27 |
| kanzure | that's Chris Phoenix | 17:30 |
| kanzure | and they've turned into environmentalists | 17:30 |
| kanzure | not that they were worth their meat in the first place | 17:30 |
| fenn | ya | 17:32 |
| fenn | supposedly they were working on an artificial ribosome | 17:32 |
| fenn | i'd like to see some kind of nanotech clause added to the geneva convention | 17:33 |
| fenn | but i think it will have to be demonstrated first :( | 17:33 |
| fenn | have you read diamond age yet? | 17:34 |
| kanzure | sigh, no | 17:35 |
| kanzure | I've been a bad boy | 17:35 |
| kanzure | give me the file | 17:35 |
| fenn | http://fennetic.net/pub/irc/Neal Stephenson - The Diamond Age.pdf | 17:36 |
| fenn | btw i use 'pdftotext -layout -htmlmeta *pdf' to get it into text | 17:37 |
| fenn | it's kind of a long book for the number of new-to-you ideas | 17:39 |
| fenn | just consider it homework i guess | 17:39 |
| kanzure | new to me? | 17:40 |
| kanzure | I just read the Wikipedia article | 17:40 |
| kanzure | it would have been new to me a few years ago maybe | 17:40 |
| kanzure | maybe that's why you're saying it's long ? | 17:41 |
| fenn | right | 17:41 |
| kanzure | k, homework then | 17:41 |
| kanzure | I'll put it on audio or something | 17:41 |
| kanzure | so if you're making me read this | 17:42 |
| kanzure | then have you read neverness yet | 17:42 |
| fenn | no | 17:42 |
| fenn | isnt it like a million pages | 17:42 |
| kanzure | http://heybryan.org/docs/Zindell, David - Neverness (v1.0).txt | 17:42 |
| kanzure | something like that :) | 17:42 |
| kanzure | 1.2 MB | 17:42 |
| fenn | a million characters | 17:43 |
| kanzure | why is your connection so slow? | 17:43 |
| kanzure | 2 kbps ? wtf? | 17:43 |
| fenn | heh dunno, i usually get about 165kB/s | 17:43 |
| fenn | probably ktorrent hogging the bandwidth and traffic shaping not doing its job because i'm downloading directly to lappy | 17:44 |
| fenn | btw how do you see what bandwidth i have? | 17:44 |
| -!- nsh- [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 17:47 | |
| fenn | a stephenson anecdote i like: http://fenn.freeshell.org/gnurds.html | 17:49 |
| kanzure | fenn: it was slow enough for wget to tell me that you suck | 17:54 |
| fenn | oh, my upload is slow? | 17:55 |
| fenn | sry bout that, will try to fix now | 17:55 |
| -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has quit [Read error: 110 (Connection timed out)] | 17:55 | |
| fenn | better? | 17:55 |
| kanzure | oh, it's long done by now | 17:58 |
| kanzure | but there is a way that I could have been monitoring your download of neverness | 17:58 |
| kanzure | I don't know what it is, but I know Superkuh uses it ;-) | 17:58 |
| kanzure | and I know that he doesn't like me going over 10 KB/sec ;-) | 17:59 |
| kanzure | the only line out of gnurds.html that still doesn't make sense to me is " The Batmobile outlet sells a few vehicles to the occasional car nut who wants a second vehicle to go with his station wagon, but seems to accept, at least for now, that it's a fringe play" | 17:59 |
| fenn | oh bloody. apparently i forgot to lock up the 13101 pdf's and someone's trying to mirror them | 17:59 |
| kanzure | I wasn't aware that BeOS was even considered by Mac users | 17:59 |
| kanzure | hm? | 17:59 |
| fenn | station wagon refers to windows | 17:59 |
| kanzure | hrm | 18:00 |
| kanzure | okay, even then though | 18:00 |
| fenn | apple is 'sleek Euro-styled sedans' | 18:00 |
| kanzure | I guess I just don't know the audience that turns to BeOS or even FreeBSD and so on. | 18:00 |
| fenn | mostly windows programmers it seems | 18:00 |
| fenn | freebsd is a little harder to figure out | 18:01 |
| fenn | my mail server is freebsd and it seems inferior | 18:01 |
| fenn | lots of little details that sort of bring you back to the early 90's | 18:02 |
| kanzure | DNS stuff? | 18:02 |
| fenn | like 'cp foo bar -r' bitches at you because the -r is last | 18:03 |
| * kanzure thought that it would bitch at you anyway | 18:04 | |
| kanzure | I mean, regardless of freebsd | 18:04 |
| fenn | no 'ps -ef' which irks me | 18:06 |
| kanzure | I'm going to stalk the aspies | 19:06 |
| kanzure | might be back later | 19:06 |
| fenn | enjoy | 19:06 |
| kanzure | the group description page doesn't seem too bad | 19:06 |
| kanzure | it's not a parent group, so maybe I'll get something out of it | 19:06 |
| kanzure | like food? | 19:06 |
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit [Remote closed the connection] | 19:06 | |
| willPow3r | singularitarians annoy me | 20:01 |
| willPow3r | they seem to look only for evidence supporting their vision | 20:01 |
| willPow3r | nothing against it | 20:01 |
| fenn | i think the singularity happened at the end of 1969 | 20:02 |
| fenn | the first one at least | 20:02 |
| willPow3r | moon landing? | 20:02 |
| fenn | that was just one thing | 20:02 |
| fenn | there was a lot of important stuff that came to fruition at the same time | 20:03 |
| fenn | i should be keeping a file but i haven't | 20:03 |
| willPow3r | i don't think that an enlightening can be considered a singularity | 20:05 |
| willPow3r | at least not by the sinularitarian's definition | 20:05 |
| fenn | i think it can | 20:05 |
| willPow3r | we like to think of our technology as our "new opposable thumb," when, in reality, we are technology's opposable thumb | 20:06 |
| fenn | luddite | 20:07 |
| willPow3r | hardly. | 20:07 |
| willPow3r | i use linux | 20:07 |
| fenn | technology doesn't have any inherent motive | 20:07 |
| fenn | technology is my thumb, eyes, ears, and fur | 20:08 |
| willPow3r | you're a furry eh? | 20:08 |
| fenn | no | 20:08 |
| fenn | it does get cold here in the winter, and i dont have fur | 20:09 |
| fenn | but i do have a fiberglass insulation lined attic | 20:09 |
| willPow3r | wow. so your house was built in the last 30 years? | 20:10 |
| fenn | no it was built in the 1920's | 20:10 |
| willPow3r | i thought asbestos was all the rage prior to fiberglass | 20:11 |
| fenn | are you following what i'm saying at all? | 20:11 |
| willPow3r | yes. | 20:11 |
| fenn | i think the asbestos hazard is more hype than reality | 20:12 |
| fenn | but that's just an opinion, i'm not a public health analyst | 20:13 |
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap | 23:39 | |
| kanzure | Hahah. | 23:48 |
| kanzure | So I was right about Aspergers, it seems. | 23:48 |
| kanzure | Bunch of guys sitting around complaining about the computational complexity of managing large todo lists. | 23:49 |
| kanzure | ;-) | 23:49 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!