Conversation with #hplusroadmap at Sat 22 Mar 2008 02:47:43 PM CDT on kanzure@irc.freenode.net (irc)
(2008-03-22 14:47:43) mode (+ns ) by simmons.freenode.net
(2008-03-22 14:47:55) epitron [i=epi@bito.ponzo.net] entered the room.
(2008-03-22 14:48:00) epitron: hmm
(2008-03-22 14:48:04) kanzure: Let's see if I can get fenn in here.
(2008-03-22 14:48:06) epitron: this is the same as us chatting vya msg
(2008-03-22 14:48:09) epitron: via
(2008-03-22 14:48:10) epitron: ok :)
(2008-03-22 14:48:23) kanzure: He'll be around shortly, I guess.
(2008-03-22 14:48:25) epitron: i'm not that into the self replicator stuff though :)
(2008-03-22 14:48:41) kanzure: hehe
(2008-03-22 14:48:46) kanzure: well, you sort of have to be
(2008-03-22 14:48:53) kanzure: if you're interested in making *relevant* engineering changes to minds
(2008-03-22 14:49:16) kanzure: otherwise you're just going to mess up and your modifications will screw you over
(2008-03-22 14:49:17) epitron: that's a different scope than vonneumann probes though
(2008-03-22 14:49:26) epitron: that's a different kind of self modification
(2008-03-22 14:49:28) kanzure: what you need is mass-scale testing of millions of different changes and then seeing how they all affect
(2008-03-22 14:49:30) epitron: different constraints
(2008-03-22 14:49:38) kanzure: hm?
(2008-03-22 14:49:43) kanzure: You still need to be able to make the neurofarms
(2008-03-22 14:49:55) epitron: what's a neurofarm
(2008-03-22 14:49:58) kanzure: and there's no way in hell you're going to get enough people to work on such a super-massive project, I bet this would be a few km^2 in size
(2008-03-22 14:50:04) kanzure: I mentioned it to you last night, but
(2008-03-22 14:50:16) kanzure: the basic idea is that if you have a few million slices of neural tissue, you want to experiment on genetic modifications with them
(2008-03-22 14:50:20) epitron: you mean, a brain supercomputer cluster?
(2008-03-22 14:50:20) epitron: :)
(2008-03-22 14:50:23) kanzure: so that you can see how such changes would theoretically change your brain
(2008-03-22 14:50:24) kanzure: sort of
(2008-03-22 14:50:33) epitron: uhm
(2008-03-22 14:50:35) kanzure: except there's not necessarily any computers involved (although if we can add them in, that would be great)
(2008-03-22 14:51:20) kanzure: you can't just say "Hm, I want to delete this neurotransmitter receptor today." Not a good idea. You'd want to see what that would do. You have to have a sandbox of some sort, and it's best if this sandbox is not in your own head and not currently in use.
(2008-03-22 14:51:31) epitron: confused here.
(2008-03-22 14:51:35) kanzure: please explain
(2008-03-22 14:51:51) epitron: you're talking about solving a problem that i'm not aware of :)
(2008-03-22 14:51:56) epitron: please delineate problem space
(2008-03-22 14:52:09) kanzure: okay, within the domain of brain/mind modification, right?
(2008-03-22 14:52:17) epitron: higher level
(2008-03-22 14:52:26) epitron: what is the goal of all this vonneumann stuff
(2008-03-22 14:52:36) kanzure: entropy maximization, remember?
(2008-03-22 14:52:37) epitron: i am not tuned into the same vision here :)
(2008-03-22 14:52:43) epitron: minimization?
(2008-03-22 14:52:45) kanzure: nope
(2008-03-22 14:52:54) epitron: so you want maximum complexity
(2008-03-22 14:52:55) epitron: ?
(2008-03-22 14:52:56) kanzure: hm, let me see if I can connect it back to what we were talking about yesterday
(2008-03-22 14:53:02) epitron: don't do that
(2008-03-22 14:53:04) epitron: just tell me your vision :)
(2008-03-22 14:53:21) kanzure: "What do you want to do?"
(2008-03-22 14:53:21) kanzure: "... as much as I can."
(2008-03-22 14:53:28) kanzure: I wonder if that helps.
(2008-03-22 14:53:31) epitron: hahah
(2008-03-22 14:53:37) fenn [n=pz@adsl-76-251-86-248.dsl.bltnin.sbcglobal.net] entered the room.
(2008-03-22 14:53:44) epitron: so you want to do whatever is possible?
(2008-03-22 14:53:54) kanzure: that's one way to put it
(2008-03-22 14:54:00) epitron: how would you put it? :)
(2008-03-22 14:54:04) fenn: why, that's impossible!
(2008-03-22 14:54:04) epitron: (Besides "because i can")
(2008-03-22 14:54:21) kanzure: fenn: epitron just asked me about the theoretical basis of why I want replicators and self-modification; but I was figuring that we might as well aggregate each other into one channel for discussion
(2008-03-22 14:54:23) epitron: s/because/as much as/
(2008-03-22 14:55:37) fenn: i want replicators because they are extremely cost efficient (from the initiator's point of view)
(2008-03-22 14:55:52) fenn: even if it's a roundabout way of doing things
(2008-03-22 14:56:05) kanzure: I used to be able to derive von Neumann probes from first principles of philosophy, all the way back from entropy maximization and goal-orientation and productivity to "Why do anything at all?" and behavior etc.
(2008-03-22 14:56:07) epitron: what about externalities :)
(2008-03-22 14:56:26) fenn: epitron: externalities are ... external :)
(2008-03-22 14:56:27) kanzure: I seem to have lost the ability to pull it off the top of my head.
(2008-03-22 14:56:33) epitron: fenn: that's still a cost though
(2008-03-22 14:56:37) epitron: because you depend on the external world
(2008-03-22 14:56:48) fenn: then it's not an externality
(2008-03-22 14:56:53) epitron: sure it is :)
(2008-03-22 14:57:03) epitron: just because you don't account for it doesn't mean it can't bite you in the ass
(2008-03-22 14:57:15) fenn: now i'm confused
(2008-03-22 14:57:17) epitron: haha
(2008-03-22 14:57:18) kanzure: me too
(2008-03-22 14:57:22) epitron: ok sorry
(2008-03-22 14:57:24) epitron: let me explain
(2008-03-22 14:57:26) fenn: how do you predict the unpredictable?
(2008-03-22 14:57:38) epitron: replicators are cost effective in the sense that you turn one on, and then leave
(2008-03-22 14:57:44) epitron: replicators eat resources though
(2008-03-22 14:57:51) epitron: and you depend on resources, since you yourself are a replicator
(2008-03-22 14:58:03) epitron: you can imagine a universe where they grow like a cancer
(2008-03-22 14:58:03) fenn: not necessarily. imagine we have a clay/sand/sunlight replicator
(2008-03-22 14:58:07) epitron: and block out the sun
(2008-03-22 14:58:16) fenn: and we plop it into a barren desert wasteland
(2008-03-22 14:58:34) kanzure: epitron: You are a cancer for not allowing cancers. Reconcile *that*.
(2008-03-22 14:58:47) fenn: the replicator will prosper due to its artificial metabolism, where normal biology has no chance
(2008-03-22 14:58:47) epitron: only if i grow without bound
(2008-03-22 14:58:56) epitron: competition is how cancers are kept in check :)
(2008-03-22 14:59:33) fenn: if we ration the information available to the replicators (easiest way) then they can't get out of control
(2008-03-22 14:59:39) epitron: i mean, REAL cancer is just uncontrolled growth that's tricked the body into thinking that it's normal homeostatic growth
(2008-03-22 15:00:24) fenn: we dont understand "real" cancer
(2008-03-22 15:00:36) epitron: like, not metaphorical cancer
(2008-03-22 15:00:37) kanzure: Cancer is noise, waste, growth, and yet "what's one man's trash is another man's treasure," so how can you say that some process is truly cancerous and truly dangerous?
(2008-03-22 15:00:37) kanzure: That is why I simply cannot declare genocide against all of the spammers that email me: they too are meaningful and relevant, just that I do not want to spend the effort in that particular area.
(2008-03-22 15:00:38) epitron: an actual tumor
(2008-03-22 15:01:11) epitron: kanzure: ok, well, back to evolution...
(2008-03-22 15:01:14) kanzure: why?
(2008-03-22 15:01:17) kanzure: I thought we were past that last night? heh'
(2008-03-22 15:01:23) epitron: uncontrolled growth has no long-term survival benefits
(2008-03-22 15:01:32) fenn: yes it does
(2008-03-22 15:01:33) epitron: it will eventually consume all things it depends on
(2008-03-22 15:01:41) epitron: how so?
(2008-03-22 15:01:50) fenn: it maximizes the number of units that can survive
(2008-03-22 15:02:02) fenn: i'm not saying it's the best strategy..
(2008-03-22 15:02:03) epitron: but those units depend on resources. what happens when you run out?
(2008-03-22 15:02:06) kanzure: By definition, growth excludes meanings that say "growth but it's secretly not growing, and instead it's hurting itself"
(2008-03-22 15:02:11) fenn: you stop growing
(2008-03-22 15:02:18) epitron: more than that
(2008-03-22 15:02:19) epitron: you die! :)
(2008-03-22 15:02:25) fenn: um, no
(2008-03-22 15:02:32) epitron: you need resources for maintenance
(2008-03-22 15:02:35) epitron: thanks to thermodynamics
(2008-03-22 15:03:02) fenn: http://en.wikipedia.org/wiki/Growth_curve
(2008-03-22 15:03:12) epitron: kanzure: i was talking about things that are secretly not growing?
(2008-03-22 15:04:00) epitron: fenn: bacteria are homeostatic
(2008-03-22 15:04:01) kanzure: epitron: You were proposing positive growth, but then saying "Oh, but it consumes everything and this hurts the overall system" and is that really growth, then? In other words, when you are not ready for certain growth, I think that expansion may be damaging or something, but this is another topic for another time.
(2008-03-22 15:04:02) fenn: the plateau will gradually fall back a little bit, and then settle into a long-term phase which can last years
(2008-03-22 15:04:15) epitron: they are designed to not overshoot their environment's capacity to support them
(2008-03-22 15:04:29) kanzure: designed?j
(2008-03-22 15:04:29) fenn: depending on the species, it can last much longer (some bacteria have been revived from salt mines from hundreds of millions of years in the past)
(2008-03-22 15:04:34) epitron: structured
(2008-03-22 15:04:43) epitron: they contain codes to make them function that way :)
(2008-03-22 15:04:50) epitron: i use the word design to mean "formed"
(2008-03-22 15:05:09) epitron: design == billions of years of natural selection :)
(2008-03-22 15:06:27) fenn: epitron: please explain to us your long-term survival strategy
(2008-03-22 15:07:09) epitron: well, i suppose ultimately everything has to die... the game is to draw it out for as long as possible
(2008-03-22 15:07:14) epitron: and to make it interesting
(2008-03-22 15:07:15) epitron: :)
(2008-03-22 15:07:18) epitron: the art of living
(2008-03-22 15:07:27) kanzure: "Self-creation is the highest art."
(2008-03-22 15:07:42) Enki-2 [n=weechat@c-71-234-190-248.hsd1.ct.comcast.net] entered the room.
(2008-03-22 15:07:59) epitron: self-creation is quite an amazing feat
(2008-03-22 15:08:11) kanzure: Enki-2: fenn and I were working on self-replicators earlier this morning, but have been draw into a chat with epitron about cancerous growth and survival strategies.
(2008-03-22 15:08:25) Enki-2: ah
(2008-03-22 15:08:30) epitron: hi enki :)
(2008-03-22 15:08:55) epitron: i actually was just trying to understand why these guys are so hyper about self-replicators
(2008-03-22 15:08:59) epitron: and it somehow led into this
(2008-03-22 15:09:25) epitron: i mean, there's self replicators everywhere in nature... they're awesome...
(2008-03-22 15:09:36) epitron: but von neumann machines are like totally something else :)
(2008-03-22 15:09:43) fenn: why's that?
(2008-03-22 15:09:45) kanzure: there are a number of reasons for self-replicators, as I had explained earlier (and just because I like to repeat them): redundancy, testing for potential self-modifications, ...
(2008-03-22 15:09:56) epitron: well, describe your idea vonneumann replicator
(2008-03-22 15:09:59) epitron: ideal
(2008-03-22 15:10:13) kanzure: The basic idea of a von Neumann probe is to have a space-probe that is able to navigate the galaxy and use self-replication (see RepRap and bio). The probe would contain hundreds of thousands of digital genomes (sequenced DNA), DNA synthesizers and sequencers, bacteria, embryos, stem cells, copies of the Internet Archive and a significant portion of the WWW in general, plus
(2008-03-22 15:10:20) kanzure: the immediate means and tools to copy all of the information and create a material embodiment, kind of like running an unzip utility on top of the thousands of exabytes predicted to be inexistence today. This would probably include many people, societies, even entire civilizations if we can collect enough data and begin to 'debug' civilization.
(2008-03-22 15:10:27) kanzure: The system might end up using an ion drive and a hydrogen collector, with on-board nucleosynthesis to create the biomolecules necessary for life, plus ways to attach to asteroids and begin replicating and copying the data and biomaterials.
(2008-03-22 15:10:36) kanzure: Self-replication is, in general, an interesting problem. See the Center for Bits and Atoms, RepRap, some MNT websites, etc. In debates and discussions in wta-talk and extropy-chat, there has been many a suggestion that mastery of self-replication will lead to post-scarcity economics or acceleration in some way-shape-or-form.
(2008-03-22 15:10:48) kanzure: The three methods mentioned on this page are (1) atom holography (generally: precision control of matter/energy), (2) molecular nanotechnology (MNT), and (3) bacterial bootstrapping (with the open source synbio/biohacking community, biobricks, the synbio mailing list, Endy/Weiss/Tom Knight/etc.), though RepRap might get to it first with the 3D fabricators. An often unexplored and unmentioned method is semiconductor manufacturing but a good first step would be a si-fab that can make a fab on a chip.
(2008-03-22 15:10:48) ***epitron knits his brows
(2008-03-22 15:10:49) fenn: kanzure is a spam-apologist!
(2008-03-22 15:10:53) kanzure: heh'
(2008-03-22 15:11:07) epitron: kanzure also doesn't like simplicity :)
(2008-03-22 15:11:11) kanzure: sure I do
(2008-03-22 15:11:19) epitron: hahaha
(2008-03-22 15:11:20) epitron: come on
(2008-03-22 15:11:36) kanzure: if you have a way to make engineered self-replication simple, please do tell
(2008-03-22 15:11:52) epitron: if you think that pasting that whole wiki page again was the simplest way to get your point across..
(2008-03-22 15:11:58) kanzure: that wasn't on the wiki
(2008-03-22 15:12:07) kanzure: or maybe it was, I forget
(2008-03-22 15:12:08) epitron: i've seen that before
(2008-03-22 15:12:12) kanzure: it was on /projects/atoms/
(2008-03-22 15:12:17) kanzure: anyway, that's a pretty good ideal von Neumann probe
(2008-03-22 15:12:17) Enki-2: epitron: he wants to colonize the universe with non-biological life to make way for biological life, from what i understand
(2008-03-22 15:12:17) epitron: whatever!
(2008-03-22 15:12:20) epitron: :)
(2008-03-22 15:12:22) kanzure: Enki-2: nah
(2008-03-22 15:12:29) epitron: hah!
(2008-03-22 15:12:32) kanzure: Enki-2: it's just a process of entropy maximization
(2008-03-22 15:12:34) epitron: see even your followers don't share your vision
(2008-03-22 15:12:36) kanzure: i.e., doing as much as possible
(2008-03-22 15:12:38) epitron: :)
(2008-03-22 15:12:41) kanzure: but Enki-2 is not a follower
(2008-03-22 15:12:49) epitron: oh ok :)
(2008-03-22 15:13:01) kanzure: do I even have followers?
(2008-03-22 15:13:20) ***epitron peers around
(2008-03-22 15:13:23) epitron: i guess not
(2008-03-22 15:13:24) epitron: :)
(2008-03-22 15:13:29) ***fenn blinks
(2008-03-22 15:13:47) fenn: i thought we were discussing out plans to take over the world
(2008-03-22 15:13:50) kanzure: oh
(2008-03-22 15:13:52) kanzure: my mistake
(2008-03-22 15:14:02) epitron: so kanzure, saying "entropy maximization" is so vague and abstract. what's an example of that
(2008-03-22 15:14:21) epitron: a fractal heat sink?
(2008-03-22 15:14:32) kanzure: epitron: An example of that would be gradient dissassembly. "Nature abhors a gradient". Organisms feed off of a gradient and energy supply, and then convert this energy into entropy, into other forms for other processes to use, etc.
(2008-03-22 15:14:43) Enki-2: the universe is entropy maximixation
(2008-03-22 15:14:54) epitron: Enki-2: see, i don't agree with that
(2008-03-22 15:14:57) kanzure: how's that?
(2008-03-22 15:15:03) kanzure: At any moment, there's maximum entropy production
(2008-03-22 15:15:09) kanzure: How the hell could the universe be any more real than it is at the moment?
(2008-03-22 15:15:16) epitron: i think it uses entropy as an intermediary state before collapsing back into non-entropy
(2008-03-22 15:15:22) fenn: stars are still full of fusionable elements.. they're so slow it takes billions of years to burn it all
(2008-03-22 15:15:28) Enki-2: kanzure: the universe is nothing but an excuse to maximize the entropy of imaginary energy
(2008-03-22 15:15:35) fenn: "we can do it better!" stupid stars!
(2008-03-22 15:15:46) kanzure: fenn: we can't do it better at the moment since it's all so inaccessible ;)
(2008-03-22 15:15:50) Enki-2: kanzure: the big bang was a start to entropy; everything after it is furthering entropy
(2008-03-22 15:15:53) fenn: shh
(2008-03-22 15:15:54) kanzure: fenn: theoretically, yes, but only once we go *do it*
(2008-03-22 15:16:11) epitron: Enki-2: right, but what happens as the universe evolves?
(2008-03-22 15:16:19) epitron: look at gravity
(2008-03-22 15:16:23) epitron: it's pulling stuff together
(2008-03-22 15:16:29) kanzure: Salthe would like to have a few words on this
(2008-03-22 15:16:30) kanzure: hold on
(2008-03-22 15:16:34) epitron: on the large scale, you get accumulation of black holes
(2008-03-22 15:16:37) Enki-2: epitron: the universe is simply a model of entropy in action
(2008-03-22 15:16:50) Enki-2: epitron: it's a very big, very good model, which is why we are in it
(2008-03-22 15:16:52) epitron: black holes are like de-enetropy :)
(2008-03-22 15:16:53) kanzure: Salthe says: In the material world, gradients, to the extent that they are steep, tend to generate, or attract and associate with, dissipative structures that will degrade them.
(2008-03-22 15:16:59) Enki-2: epitron: we are mechanisms of entropy
(2008-03-22 15:17:11) epitron: yeah, but you think we're the grand purpose of the universe?
(2008-03-22 15:17:16) Enki-2: epitron: no
(2008-03-22 15:17:17) epitron: what if we're just a little fluctuation
(2008-03-22 15:17:20) kanzure: diagram - http://www.nbi.dk/~natphil/salthe/levels.gif
(2008-03-22 15:17:26) kanzure: wait, what?
(2008-03-22 15:17:28) Enki-2: epitron: the universe has no purpose, because the universe does not exist
(2008-03-22 15:17:38) kanzure: Grand purpose of the universe?
(2008-03-22 15:17:39) epitron: and yet!
(2008-03-22 15:17:42) epitron: here we are. :)
(2008-03-22 15:17:46) epitron: existing, against all logic.
(2008-03-22 15:17:48) Enki-2: epitron: no. here we aren't
(2008-03-22 15:17:54) epitron: but at the same time, are.
(2008-03-22 15:17:55) epitron: :D
(2008-03-22 15:17:56) Enki-2: epitron: we don't exist. neither do you.
(2008-03-22 15:17:58) kanzure: So, do you mean to say that, "How should we know whether or not to use materials, since we do not 'own' them and they are not ours?", epitron?
(2008-03-22 15:18:04) Enki-2: epitron: we are a potenitality
(2008-03-22 15:18:11) epitron: kanzure: no no no not at all
(2008-03-22 15:18:16) epitron: i'm just trying to understand your purpose
(2008-03-22 15:18:16) Enki-2: epitron: a virtual potentialty, in fact
(2008-03-22 15:18:20) kanzure: epitron: because you're saying how do we know we are meaningfully relevant in the universe
(2008-03-22 15:18:27) epitron: guys
(2008-03-22 15:18:30) epitron: you're gang raping me
(2008-03-22 15:18:36) epitron: gimme a chance to answer these allegations! :)
(2008-03-22 15:18:39) Enki-2: orly?
(2008-03-22 15:18:45) kanzure: haha
(2008-03-22 15:18:47) ***Enki-2 grabs the dirty stick and does a better job
(2008-03-22 15:19:13) epitron: kanzure: i'm raising the question... i know that humans are very stupid primitive beings. :)
(2008-03-22 15:19:24) epitron: there are many dimensions to this universe that we don't fathom
(2008-03-22 15:20:03) epitron: unleashing a horde of terraforming machines to populate the galaxy may not be fruitful
(2008-03-22 15:20:17) epitron: and i'm still not totally with you on maximizing entropy. i think that any stable system has to be balanced.
(2008-03-22 15:20:36) fenn: fwiw i think terraforming is silly
(2008-03-22 15:20:48) fenn: just build a space habitat
(2008-03-22 15:21:06) epitron: hmmm
(2008-03-22 15:21:09) epitron: or adapt to the environment
(2008-03-22 15:21:17) epitron: like organisms usually do :)
(2008-03-22 15:21:36) epitron: if you put a habitat down, it's kinda like a big self-contained organsim full of little organisms
(2008-03-22 15:21:38) fenn: species adapt, organisms themselves are very limited in their adaptability
(2008-03-22 15:21:48) kanzure: you seem to be saying that "since we don't know what to do (no divine inspiration, seeing how we are rather stupid), we shouldn't just go do anything"
(2008-03-22 15:21:59) epitron: kanzure: no, i'm asking what you want to do :)
(2008-03-22 15:22:10) epitron: and you have a really really abstract answer
(2008-03-22 15:22:12) fenn: we can call it "gaia 2" if you like
(2008-03-22 15:22:30) epitron: here's my philosophy -- try everything!
(2008-03-22 15:22:34) kanzure: I want to do as much as possible -- I want to accelerate. I want to hack my mind and make it (subjectively) better, so that I can do even more and better things.
(2008-03-22 15:22:41) ***fenn gallops in all directions at once
(2008-03-22 15:22:51) epitron: but be mindful of consequences
(2008-03-22 15:23:13) epitron: kanzure: did you ever see the first star trek movie?
(2008-03-22 15:23:36) epitron: horrible movie, but the idea was written by asimov.. and it was interesting
(2008-03-22 15:23:53) fenn: i see that humans are growing at an exponential rate, regardless whether we stopped to consider if it's a good idea or not. space habitats are the best way of alleviating the situation because we can build them faster than humans can naturally reproduce
(2008-03-22 15:24:06) epitron: fenn: or humans can die :)
(2008-03-22 15:24:09) epitron: or stop breeding
(2008-03-22 15:24:11) kanzure: fenn: You make a good point.
(2008-03-22 15:24:12) epitron: (as fast)
(2008-03-22 15:24:19) fenn: the easiest and fastest way of refining minerals to build space habitats is von neuman replicators
(2008-03-22 15:24:26) kanzure: fenn: It's kinda like viruses. You can't just wish them away. But we can do things to help.
(2008-03-22 15:24:39) epitron: hey how about this -- what if every couple only had one kid instead of two
(2008-03-22 15:24:49) epitron: you'd halve the population in a few generations
(2008-03-22 15:25:04) epitron: it would take less time than building space habitats
(2008-03-22 15:25:19) fenn: epitron: what if i dont want to have less children? what will you do about it?
(2008-03-22 15:25:27) epitron: *shrug* :)
(2008-03-22 15:25:30) epitron: i'm just saying
(2008-03-22 15:25:36) fenn: it's already happening in china
(2008-03-22 15:25:39) epitron: everyone needs to be on board to build space habitats...
(2008-03-22 15:25:44) kanzure: epitron: not true
(2008-03-22 15:25:48) epitron: everyone could ALTERNATIVELY be on board to stop breeding as much
(2008-03-22 15:25:55) kanzure: uhh
(2008-03-22 15:25:57) epitron: kanzure: ok, then whatever percentage could be on board :)
(2008-03-22 15:26:01) epitron: let's sayyy.. 30%
(2008-03-22 15:26:15) kanzure: epitron: If fenn or I made a self-replicator and launched it, we could build the god damn space habitat ourselves.
(2008-03-22 15:26:25) kanzure: doesn't require 30% of the human population
(2008-03-22 15:26:31) epitron: true
(2008-03-22 15:26:38) epitron: but how long would it take?
(2008-03-22 15:26:44) kanzure: it's self-replication, so it's exponential
(2008-03-22 15:26:44) Blak1 [n=Computer@d154-20-83-209.bchsia.telus.net] entered the room.
(2008-03-22 15:26:45) kanzure: so
(2008-03-22 15:26:47) kanzure: Hey Blake.
(2008-03-22 15:26:56) Blak1: Hello
(2008-03-22 15:26:59) kanzure: Blak1: use /nick Blah to change your nick.
(2008-03-22 15:26:59) epitron: exponential growth still takes a while to ramp up :)
(2008-03-22 15:27:08) kanzure: we'd have to do the calculations
(2008-03-22 15:27:12) kanzure: but exponential models are very easy
(2008-03-22 15:27:18) kanzure: they teach that in like second-year math in high school
(2008-03-22 15:28:14) Blak1: have you considered mutation?
(2008-03-22 15:28:33) kanzure: I think fenn has.
(2008-03-22 15:28:46) epitron: anyhow, i'm not saying that the only two ways forward are population control or populating the galaxy... and i'm not saying that one is good and the other is bad.. i'm just saying that there are many paths
(2008-03-22 15:29:00) kanzure: epitron: that's true, and ideally we can explore as many paths as possible
(2008-03-22 15:29:08) kanzure: And that's the idea.
(2008-03-22 15:29:09) epitron: right, within reason :)
(2008-03-22 15:29:21) kanzure: of course, don't want to explore blowing up the planet
(2008-03-22 15:29:28) kanzure: generally frowned upon
(2008-03-22 15:29:31) epitron: or explore so many paths at once that you crash the ecosystem
(2008-03-22 15:29:46) kanzure: you just reboot the ecosystem yourself though
(2008-03-22 15:29:50) kanzure: go out and mine the resources to make it happen
(2008-03-22 15:29:52) epitron: dude
(2008-03-22 15:29:53) epitron: :)
(2008-03-22 15:30:04) epitron: ecosystems are really complicated
(2008-03-22 15:30:17) epitron: it's not easy to just stamp a new one out
(2008-03-22 15:30:26) Blak1: Retaining control of self replicating machines is pretty much impossible
(2008-03-22 15:30:35) epitron: right, you don't want control
(2008-03-22 15:30:36) epitron: you want competition
(2008-03-22 15:30:42) fenn: i may be a wingnut. i don't believe the generally accepted model of evolution is accurate - biologists have not considered the possibility of information from extraterrestrial sources such as viruses and bacteria riding on comets, or quantum entanglement effects changing 'random' events
(2008-03-22 15:30:45) epitron: external control is inefficient
(2008-03-22 15:31:01) kanzure: fenn: I don't see what implications that has re: the discussion going on.
(2008-03-22 15:31:03) epitron: fenn: sure they have. i've read lots of interesting theories about that
(2008-03-22 15:31:13) epitron: panspermia
(2008-03-22 15:31:15) fenn: so, from that, i dont think pure random mutation will lead to anything useful
(2008-03-22 15:31:16) kanzure: so?
(2008-03-22 15:31:21) kanzure: oh?
(2008-03-22 15:31:22) epitron: haha
(2008-03-22 15:31:24) fenn: because it hasn't been demonstrated
(2008-03-22 15:31:29) epitron: obviously you've never created a genetic algorithm
(2008-03-22 15:31:30) kanzure: pure random mutation has not been demonstrated?
(2008-03-22 15:31:39) epitron: they work dude
(2008-03-22 15:31:48) epitron: our brains are just not good at integrating millions of tiny variables
(2008-03-22 15:31:55) Blak1: "i may be a wingnut. i don't believe the generally accepted model of evolution is accurate" Bwah?
(2008-03-22 15:31:59) epitron: we're good at really obvious in-your-face variables
(2008-03-22 15:32:15) kanzure: I may agree that you can't declare something to be "truly random", but that doesn't mean that mutation doesn't help
(2008-03-22 15:32:38) epitron: well, ok, we can integrate millions of tiny variables through our senses... but our CONSCIOUS mind .. the intellect... isn't good at it
(2008-03-22 15:32:52) kanzure: alright, so
(2008-03-22 15:32:58) epitron: evolution is just exploring solution space randomly
(2008-03-22 15:33:04) epitron: (semi-randomly)
(2008-03-22 15:33:05) kanzure: I think you then see why I am interested in self-replication
(2008-03-22 15:33:12) epitron: you roll the dice enough times you're going to get all 1's
(2008-03-22 15:33:13) kanzure: right, semi-randomly wrt the current configuration state
(2008-03-22 15:33:39) epitron: kanzure: what, intelligence-guided evolution?
(2008-03-22 15:33:46) kanzure: what?
(2008-03-22 15:33:57) epitron: 16:29 <@kanzure> I think you then see why I am interested in self-replication
(2008-03-22 15:33:59) kanzure: ah,
(2008-03-22 15:34:00) epitron: i was guessing :)
(2008-03-22 15:34:02) kanzure: I just meant that comment in general
(2008-03-22 15:34:18) epitron: i don't see though :)
(2008-03-22 15:34:21) epitron: why are you interested?
(2008-03-22 15:34:24) kanzure: intelligence-directed evolution is interesting and useful, but I was really just refering to our previous discussion on entropy maximization, gradient utilization, etc.
(2008-03-22 15:34:39) epitron: i'm still not with you on EM and HU
(2008-03-22 15:34:40) epitron: GU
(2008-03-22 15:35:04) Blak1: Mutation and evolution can be problematic for the machines- you'll get artifacts that propagate and change them so that they no longer suit your purposes- it can't really help us, because we have total control over their structure initially.
(2008-03-22 15:35:09) fenn: i was poking at Salthe on the google, and he makes a clear distinction between gradient utilization and 'useful work'
(2008-03-22 15:35:32) epitron: Blak1: right, and that's the problem of creating life -- you have to give up contrl :)
(2008-03-22 15:35:33) Blak1: About the standard model of evolution, whether the mutations are truly random or not is not of consequence
(2008-03-22 15:35:55) kanzure: epitron: EM is like saying. Hm. Let me put it to you this way. Reality is always at its best: never will its fundamental laws be violated. Therefore, it's always at peak efficiency, **but** we can influence it to do even better/more, since "reality creates itself" and so on.
(2008-03-22 15:36:05) Blak1: Alien life forms on commets are also not an issue, because they too evolved
(2008-03-22 15:36:12) kanzure: yep
(2008-03-22 15:36:21) fenn: Blak1: mutation is a problem of a digital information system, and there are many many ways of dealing with it. the problem right now seems to be getting a digital information system to self-replicate in the first place (right down to standards of mass and distance)
(2008-03-22 15:36:33) epitron: kanzure: maybe i don't know what you mean by entropy then :)
(2008-03-22 15:36:41) kanzure: epitron: Oh. That's a good point.
(2008-03-22 15:36:47) kanzure: epitron: Some call 'entropy' stuff like 'disorder' and other bullshit.
(2008-03-22 15:36:50) kanzure: ENTROPY IS NOT DISORDER
(2008-03-22 15:36:53) kanzure: (there's a website with that title)
(2008-03-22 15:36:54) epitron: ah
(2008-03-22 15:36:57) epitron: see that's important :)
(2008-03-22 15:36:58) kanzure: In this case, entropy is just used energy.
(2008-03-22 15:37:00) kanzure: yes, very
(2008-03-22 15:37:02) kanzure: I apologize.
(2008-03-22 15:37:04) Blak1: Digital information helps prevent mutation, that's an end in itself
(2008-03-22 15:37:15) epitron: so you're saying that it is important to use every drop of energy in the universe
(2008-03-22 15:37:20) fenn: Blak1: agreed
(2008-03-22 15:37:45) epitron: Blak1: digital information is fragile too though
(2008-03-22 15:37:49) epitron: it's dependant on the storage medium
(2008-03-22 15:37:55) Blak1: I was not aware that this was a strictly digital process
(2008-03-22 15:38:03) Blak1: analog information is much more versatile
(2008-03-22 15:38:11) fenn: analog information degrades quickly
(2008-03-22 15:38:15) epitron: yeah, but fuzzier :)
(2008-03-22 15:38:40) fenn: for instance, your plastic goo squirter makes another plastic goo squirter, but the nozzle is messed up, so the next goo squirter is even uglier than the first
(2008-03-22 15:39:01) Blak1: yep
(2008-03-22 15:39:04) fenn: before long, the system is too degraded to function
(2008-03-22 15:39:19) Blak1: that was my initial question about mutation
(2008-03-22 15:39:26) Blak1: so you combat it with digital technology
(2008-03-22 15:39:27) kanzure: epitron: Right, but with specialization and differentiation, to have lots of variety and to make reality increasingly more 'real'. Better us than the redshift.
(2008-03-22 15:39:32) Blak1: some sort of checksum done?
(2008-03-22 15:39:39) kanzure: note that DNA replication is digital technology
(2008-03-22 15:39:54) Blak1: can they diagnose their offspring and repair each other?
(2008-03-22 15:40:00) kanzure: but DNA also means amorphous fabrication, while I want controllable fabrication
(2008-03-22 15:40:08) kanzure: not their offspring, but there are DNA repair mechanisms
(2008-03-22 15:40:08) epitron: kanzure: well, think about this though... what if the universe has a use for all that free stuff... what if using up ALL of it is actually going to doom the universe
(2008-03-22 15:40:22) fenn: Blak1: in terms of measurement and fabrication, you use calibration to a reference (preferably a "digital" reference such as a reproducible wavelength of light)
(2008-03-22 15:40:24) epitron: kanzure: i mean, matter is tight stable energy containment units...
(2008-03-22 15:40:35) Blak1: Cells don't check each other very thoroughly though
(2008-03-22 15:40:51) kanzure: epitron: Doom the universe as we know it today; but isn't that what is required to create? "Yes, I can create; but to create I must first uncreate."
(2008-03-22 15:40:52) epitron: kanzure: if you start messing with matter, you convert it into smaller bits, and it loses energy
(2008-03-22 15:41:05) kanzure: Blak1: That's true.
(2008-03-22 15:41:08) epitron: kanzure: right, but you're being hyperbolic
(2008-03-22 15:41:17) epitron: kanzure: it's possible to use some of the resources without using all :)
(2008-03-22 15:41:21) epitron: it's possible to be moderate
(2008-03-22 15:41:28) kanzure: moderate with respect to what?
(2008-03-22 15:41:35) epitron: ok her'es an example
(2008-03-22 15:41:51) epitron: it's good to explore all possibilities so that you find the best one
(2008-03-22 15:42:03) epitron: but once you've found the best one, do you need to populate the entire universe with that thing?
(2008-03-22 15:42:13) epitron: why not just have enough of them that they won't ever be destroyed
(2008-03-22 15:42:15) kanzure: fenn: yeah, please continue on with your callibration ideas
(2008-03-22 15:42:15) epitron: or die off
(2008-03-22 15:42:59) kanzure: epitron: That's wasteful, though, to populate the entire universe with that 'one' thing. That's the grey goo scenario.
(2008-03-22 15:43:06) epitron: right
(2008-03-22 15:43:08) Blak1: they'll all be destroyed and die off eventually
(2008-03-22 15:43:12) epitron: but if you look at it in a different way...
(2008-03-22 15:43:20) fenn: ok, you can make a plastic goo squirter with a 50% error rate. if it doesn't match the reference standard, throw out the defective unit and make another one. repeat until you have a goo squirter that's within specification
(2008-03-22 15:43:36) Blak1: grey good scenario isn't possible, heh'
(2008-03-22 15:43:37) kanzure: There is already grey goo (sort of) -- bacteria.
(2008-03-22 15:43:50) kanzure: (that was not in response to Blake.)
(2008-03-22 15:44:05) epitron: what if populating the universe with beings that maximize entropy and decompose all matter into free energy IS populating it with one 'thing', that thing being 'things that maximize entropy'
(2008-03-22 15:44:40) epitron: what if nature wants to save some of that energy :)
(2008-03-22 15:45:00) epitron: if you convert it all into radiation, it won't get sucked into black holes
(2008-03-22 15:45:07) epitron: what if it being sucked into black holes is important?
(2008-03-22 15:46:28) epitron: have you ever seen simulations of large-scale dark matter structure in the universe?
(2008-03-22 15:46:32) kanzure: and what if it is not? There's a 50/50 chance that black holes are important in that sense.
(2008-03-22 15:46:32) fenn: current human technology is just beginning to approach a 'digital' mass reference standard, in the form of counting individual atoms and also weighing them
(2008-03-22 15:47:02) epitron: haha
(2008-03-22 15:47:05) epitron: how do you know it's 50/50
(2008-03-22 15:47:07) kanzure: If it is important to feed black holes energy, and you don't, you didn't toughen up. If it is not important to feed black holes, and you do, you lose.
(2008-03-22 15:47:09) fenn: 50/50 isn't the same as unknowable
(2008-03-22 15:47:18) fenn: and unknowable isn't unknown
(2008-03-22 15:47:18) epitron: it is if he doesn't know that it's 50/50! :)
(2008-03-22 15:47:27) epitron: ok ok
(2008-03-22 15:47:31) epitron: unknown then
(2008-03-22 15:47:31) epitron: ;)
(2008-03-22 15:47:57) kanzure: black holes can feed off of CMB, IIRC.
(2008-03-22 15:47:57) epitron: kanzure: black holes don't need to be FED though
(2008-03-22 15:48:01) epitron: they eat voraciously
(2008-03-22 15:48:06) epitron: but you can starve them
(2008-03-22 15:48:10) epitron: by stealing their food
(2008-03-22 15:48:32) epitron: yeah, but radiation is much less likely to go into a black hole than particles
(2008-03-22 15:48:40) kanzure: radiation is particles.
(2008-03-22 15:48:41) kanzure: heh'
(2008-03-22 15:48:43) epitron: the more mass you have the more likely you are to hit a black hole
(2008-03-22 15:48:45) epitron: radiation has no mass
(2008-03-22 15:48:49) epitron: it's pure energy
(2008-03-22 15:48:49) fenn: hey here's an idea - camp out by black holes and utilize the x-rays coming out of them (free energy)
(2008-03-22 15:49:10) epitron: haha
(2008-03-22 15:49:14) epitron: that's a good idea
(2008-03-22 15:49:37) epitron: i mean, reasonably, we probably CAN'T eat the entire universe
(2008-03-22 15:49:49) epitron: the limit on the speed of light puts the brakes on that
(2008-03-22 15:49:54) kanzure: yep
(2008-03-22 15:50:01) epitron: the best we can do is probably corrupt a little chunk
(2008-03-22 15:50:07) kanzure: I want to get back to work, fenn, where were we?
(2008-03-22 15:50:12) epitron: but still.. it would be sad if we got owned by our own creations
(2008-03-22 15:50:18) kanzure: it would be sad, yes
(2008-03-22 15:50:19) epitron: ok i'll get back to work too :)
(2008-03-22 15:50:22) epitron: ttyl
(2008-03-22 15:50:25) kanzure: epitron: I was hoping you'd stay
(2008-03-22 15:50:29) kanzure: that's why I invited you in here heh'
(2008-03-22 15:50:32) epitron: i will
(2008-03-22 15:50:37) epitron: but i don't care about vonneumann machines
(2008-03-22 15:50:40) epitron: you guys can do that stuff
(2008-03-22 15:50:44) kanzure: even though you want self-modification?
(2008-03-22 15:50:51) kanzure: how are you going to figure out what modifications to make?
(2008-03-22 15:50:51) fenn: lets see.. i was bitcing about wiki's not being capable of showing multiple interconnections very well
(2008-03-22 15:50:52) epitron: again, different scope :)
(2008-03-22 15:51:05) kanzure: epitron: how are you going to do self-modification without killing yourself?
(2008-03-22 15:51:07) epitron: i'll let you guys figure out some high level principles of self modification :D
(2008-03-22 15:51:25) epitron: you seem interested
(2008-03-22 15:51:27) kanzure: you really trust us?
(2008-03-22 15:51:28) epitron: i'm interested in keeping my job
(2008-03-22 15:51:30) epitron: so i will do my work
(2008-03-22 15:51:38) epitron: i trust myself to review your work :)
(2008-03-22 15:51:45) kanzure: I see.
(2008-03-22 15:51:49) epitron: and empirical evidence is the most convincing to me
(2008-03-22 15:54:13) kanzure: So, Blak1, I think I've linked you to the website enough times so that you know what's going on already
(2008-03-22 15:54:22) kanzure: basically we are still stuck trying to figure out how to make a self-replicator
(2008-03-22 15:54:45) kanzure: Using biology as a stepping stone isn't a good idea, we've decided, for various reasons [that I'd be happy to explain]
(2008-03-22 15:55:31) kanzure: Although synthetic-ecologies might be useful: harvest the bacteria that are mining some element, then scoop it up and process the materials (getting rid of the cellular gunk)
(2008-03-22 15:56:21) Blak1: There's really not so much a difference
(2008-03-22 15:56:37) Blak1: between using very complex machines and "biological" organisms
(2008-03-22 15:56:56) Blak1: they really merge at a point
(2008-03-22 15:57:00) kanzure: http://www.kk.org/outofcontrol/ch8-e.html some notes on Bio2
(2008-03-22 15:57:04) kanzure: Somewhat.
(2008-03-22 15:57:09) kanzure: But biological organisms are amorphous fabricators
(2008-03-22 15:57:22) kanzure: You can't program "Make me a head: here's the points that you have to make" for cells
(2008-03-22 15:57:24) Blak1: they don't have to be
(2008-03-22 15:57:30) kanzure: you have to do a linear -> parallel program translation
(2008-03-22 15:57:38) kanzure: and that parallel program has to be "biochemically valid"
(2008-03-22 15:57:57) Blak1: So I can't tell a bunch of peasants to make me a head?
(2008-03-22 15:58:03) kanzure: peasants are not cells
(2008-03-22 15:58:08) Blak1: Humans can construct buildings
(2008-03-22 15:58:11) Blak1: I beg to differ
(2008-03-22 15:58:19) kanzure: Humans seem to have some brain function on 'em.
(2008-03-22 15:58:31) kanzure: And are really wasteful for mass production, it seems.
(2008-03-22 15:58:31) Blak1: so would a mechanical cell
(2008-03-22 15:58:37) kanzure: huh?
(2008-03-22 15:58:45) kanzure: a mechanical cell would not follow the same instructions as a group of humans
(2008-03-22 15:58:46) epitron: http://www.mpa-garching.mpg.de/galform/data_vis/lcdm_color2_highres_divx.avi
(2008-03-22 15:58:50) kanzure: a population of cells would be programmed by DNA
(2008-03-22 15:58:58) Blak1: the mechanical cell would have to organize, etc.
(2008-03-22 15:59:04) kanzure: a population of humans would be trained in the English language and so on
(2008-03-22 15:59:23) Blak1: I'm not saying use humans
(2008-03-22 15:59:25) kanzure: And has very constrained output from their 100 billion cells and 100 trillion connections.
(2008-03-22 15:59:26) kanzure: hm?
(2008-03-22 15:59:29) Blak1: we can createrinly imporve on that
(2008-03-22 15:59:37) Blak1: but there's no reason to depart from the biological resources we have
(2008-03-22 15:59:40) Enki-2: a population of bots would use other languages that we actually KNOW, and can use
(2008-03-22 15:59:56) Enki-2: DNA is slow, as is english and human memory
(2008-03-22 16:00:03) kanzure: Whereas DNA in cells "just so happens to work"
(2008-03-22 16:00:15) kanzure: whereas programming a computer *works* because we have the definition all the way down
(2008-03-22 16:00:32) kanzure: but if you have an idea for how to get solid-state fabrication from cells, Blake, I am listening
(2008-03-22 16:00:43) epitron: http://visservices.sdsc.edu/projects/enzo/movies/dark/viz3-dark_hl_s97.mov
(2008-03-22 16:00:50) Enki-2: the difference is that current computers are entirely von-neumann; they lack the parallelization and coordination at the most base of levels
(2008-03-22 16:00:56) kanzure: right
(2008-03-22 16:01:05) epitron: (large scale dark matter evolution)
(2008-03-22 16:01:08) Blak1: neurons are still ,more powerful processors than any computer
(2008-03-22 16:01:14) Enki-2: Blak1: not really
(2008-03-22 16:01:18) Blak1: and they can be configured into appropriate parts
(2008-03-22 16:01:32) kanzure: If you have a way to configure neurons to run programs that we specify, that would be interesting and useful
(2008-03-22 16:01:33) Blak1: it's analog processing, but it's all we really need
(2008-03-22 16:01:50) Blak1: Biological logic gates are possible
(2008-03-22 16:01:50) Enki-2: Blak1: a single neuron is essentially just a capacitor with the ability to trigger (in certain circumstances) changes to its own charge and capitance
(2008-03-22 16:01:55) kanzure: as in, I have a source code file in C++, if you have a way to make a physical neural net that expresses that program -- heh, do tell.
(2008-03-22 16:02:01) kanzure: Blak1: yes, biological logic gates are possible
(2008-03-22 16:02:15) fenn: kanzure: i think winfree already showed how to get 'solid state fabrication' from cells
(2008-03-22 16:02:23) kanzure: fenn: really?
(2008-03-22 16:02:29) kanzure: fenn: with the origami stuff?
(2008-03-22 16:02:30) fenn: with the DNA-CNT hybrid tile self assembly
(2008-03-22 16:02:48) kanzure: that was cellular automata though
(2008-03-22 16:02:56) kanzure: it was an algorithm with those emergent properties
(2008-03-22 16:03:04) fenn: you program the cellular automata to do more complex things
(2008-03-22 16:03:19) fenn: i dont see why you would need anything more complex (linear assembly)
(2008-03-22 16:03:25) fenn: at least as far as computers go
(2008-03-22 16:03:28) Blak1: I think it's rediculous to reinvent the wheel. If we make machines adequately complex, they will acquire the same attributes of biological components
(2008-03-22 16:03:31) kanzure: it is my understanding that you can't analytically generate a cellular automata rule to come up with what you want, but rather you have to search through known CA rules to find something that gives you waht you want
(2008-03-22 16:03:35) Blak1: biomechanical is the way to go
(2008-03-22 16:03:47) Blak1: wherein you need mechanical parts
(2008-03-22 16:04:18) Blak1: and micro-controllers make of organs that can check each other for consistency
(2008-03-22 16:04:34) kanzure: but think about the problems involved
(2008-03-22 16:04:36) kanzure: suppose you have a soup of cells
(2008-03-22 16:04:41) kanzure: and you want to program them to make something, like a doll
(2008-03-22 16:04:47) kanzure: none of them know the information about where they are with respect to all of the others
(2008-03-22 16:04:53) Blak1: it wouldn't be a soup of cells
(2008-03-22 16:04:57) kanzure: hm?
(2008-03-22 16:04:58) kanzure: what would it be?
(2008-03-22 16:04:59) Blak1: it would be a series of organs
(2008-03-22 16:05:06) kanzure: to do what?
(2008-03-22 16:05:08) fenn: Blak1: biological self replicators (cells) have a lot of unusual material and environmental requirements, in the grand scheme of things
(2008-03-22 16:05:30) kanzure: fenn: the unusual requirements are irrelevant; I am more worried about the fundamental programming difficulties
(2008-03-22 16:05:41) Blak1: Not necessarily.
(2008-03-22 16:05:51) Blak1: One could make organs that can do anything a machine can
(2008-03-22 16:06:13) fenn: not so
(2008-03-22 16:06:15) kanzure: how's that?
(2008-03-22 16:06:25) fenn: show me a biological hard vacuum
(2008-03-22 16:06:36) kanzure: I have a file overhere in XML format that details a way to make an engine
(2008-03-22 16:06:38) Enki-2: :P
(2008-03-22 16:06:59) Enki-2: fenn: show me a mechanical hard vacuum :P
(2008-03-22 16:07:07) Enki-2: though i have to admit
(2008-03-22 16:07:27) fenn: Enki-2: it's a rather arcane technical subject, unfortunately
(2008-03-22 16:07:30) Enki-2: something like, say, a gun or a rocket propulion system would be tough in biological parts
(2008-03-22 16:07:41) kanzure: biological parts generally can't be made into those forms though
(2008-03-22 16:07:45) fenn: there are biological rockets, like bombardier beetles
(2008-03-22 16:07:47) kanzure: what's the program to make a rocket engine
(2008-03-22 16:07:54) kanzure: you can't just specify the way that the cells should end up
(2008-03-22 16:07:57) kanzure: it's nothing like casting
(2008-03-22 16:08:04) kanzure: it's a parallel amorphous program that they have to follow
(2008-03-22 16:08:17) kanzure: and this requires very tricky computer science for converting our linear representations into those parallel programs
(2008-03-22 16:08:25) fenn: a better requirement is 'a biological organism that can get to earth orbit'
(2008-03-22 16:08:43) fenn: i can picture it, but it's huge and sprawling and shoots tiny seeds
(2008-03-22 16:09:01) kanzure: I am sure there are some proteins or something that can make diamond CNTs or something, this would make the inside of the engine or exhaust exit
(2008-03-22 16:09:15) Enki-2: fenn: interstellar ones?
(2008-03-22 16:09:17) fenn: you'd have to support that sprawling launch-bug with a large ecosystem
(2008-03-22 16:09:24) Enki-2: :P
(2008-03-22 16:09:27) kanzure: fenn: that's an easy problem
(2008-03-22 16:09:35) fenn: Enki-2: once you get to orbit there is the potential for further in-space evolution
(2008-03-22 16:09:38) kanzure: why are you ignoring the computation aspects of making the program to have them make the engine in the first place
(2008-03-22 16:09:45) kanzure: you have to "evolve" the engine basically
(2008-03-22 16:09:57) fenn: yeah, that's not gonna happen
(2008-03-22 16:10:17) kanzure: not without very large supercomputers that can only be built from self-replication
(2008-03-22 16:10:18) kanzure: oh wait ;)
(2008-03-22 16:10:24) fenn: there's no evolutionary advantage until you overcome a huge barrier
(2008-03-22 16:10:48) fenn: but we do have coconut trees
(2008-03-22 16:10:56) kanzure: hm?
(2008-03-22 16:10:56) fenn: maybe it's possible
(2008-03-22 16:11:09) fenn: the ocean's a big barrier, islands, analogy n stuff
(2008-03-22 16:11:13) kanzure: it might be possible to evolve the programs to make the rocket engines from cells
(2008-03-22 16:11:35) kanzure: but, in general, wouldn't it be much more productive to have a mechanical self-replicator?
(2008-03-22 16:11:46) kanzure: I mean, you need to do *massive* simulations to make that rocket launcher from cells
(2008-03-22 16:11:49) kanzure: very massive GAs.
(2008-03-22 16:12:29) fenn: we're still in the reverse-engineering phase with biology
(2008-03-22 16:12:31) kanzure: and then the program to make a space habitat in orbit, like fenn was talking about? heh'
(2008-03-22 16:12:46) kanzure: fenn: not quite, there are others who seem to be going in the forward direction
(2008-03-22 16:12:50) fenn: once the principles are well understood you dont have to do so much simulation
(2008-03-22 16:12:53) kanzure: (Andy seems to be one of them. Hurray for evolutionary engineering.)
(2008-03-22 16:15:33) kanzure: alright, so it seems we agree at the moment about the predicament with biology
(2008-03-22 16:15:36) kanzure: don't know if Blake agrees
(2008-03-22 16:15:52) kanzure: but this prediacement seems fixable with 'skdb', like we were talking about, fenn
(2008-03-22 16:15:58) fenn: anyway, i'll lump general purpose genetic engineering in with artificial intelligence as a buzzword we'll never reach because of rising expectations
(2008-03-22 16:16:09) kanzure: http://heybryan.org/mediawiki/index.php/Skdb
(2008-03-22 16:16:32) fenn: btw i came up with a name for a skdb package manager: 'autogenix'
(2008-03-22 16:16:45) kanzure: heh, but it doesn't necessarily autogenerate
(2008-03-22 16:16:50) kanzure: but it's a good name
(2008-03-22 16:17:02) Enki-2: fenn: aren't astronauts like cocoanuts, in terms of the analogy?
(2008-03-22 16:17:10) Enki-2: and humanity being the tree
(2008-03-22 16:17:18) Enki-2: and earth the original island
(2008-03-22 16:17:19) fenn: not yet, there arent any space colonies
(2008-03-22 16:17:21) kanzure: aww, how metaphorically romantic
(2008-03-22 16:17:23) Enki-2: true
(2008-03-22 16:17:53) fenn: an astronaut is like a cell from a coconut tree
(2008-03-22 16:17:54) Enki-2: but the cocoanut shell, though produced by the cocoanut tree, is not part of the cocoanut's lifecycle
(2008-03-22 16:17:58) kanzure: so, I wonder if we can make autogenix/skdb and the self-replicator projects conveniently converge
(2008-03-22 16:18:21) kanzure: Enki-2: I suggest you go check out that skdb link. http://heybryan.org/mediawiki/index.php/Skdb
(2008-03-22 16:18:32) Enki-2: just as the shuttle is not part of a human, yet is still made and used by humans
(2008-03-22 16:18:47) Enki-2: kanzure: i can do that, yes; once i can load firefox agian
(2008-03-22 16:18:55) Enki-2: sublims = high load
(2008-03-22 16:18:55) kanzure: hehe
(2008-03-22 16:19:02) kanzure: really?
(2008-03-22 16:20:06) Enki-2: yeah
(2008-03-22 16:20:15) Enki-2: now that i'm keeping them on much of the time
(2008-03-22 16:20:31) Enki-2: too much load to run firefox in the normal way
(2008-03-22 16:20:43) kanzure: so, for convergence of skdb and replication -- we can have an 'abstract' project for mechanical self-replication with a specification saying that the materials it uses must be able to self-replicate themselves from their forms
(2008-03-22 16:20:49) Enki-2: much of that is just in terms of the data fetching though
(2008-03-22 16:21:15) Enki-2: elias is working on something to centralize data grabbing and "broadcast" subliminal messages to clients
(2008-03-22 16:21:20) Enki-2: syncronously
(2008-03-22 16:21:42) Enki-2: i'm secretly afraid he's going to cause a giant orgy
(2008-03-22 16:21:54) kanzure: hm?
(2008-03-22 16:22:03) Enki-2: er
(2008-03-22 16:22:12) Enki-2: side affect of syncrosublimming
(2008-03-22 16:22:18) kanzure: ah
(2008-03-22 16:22:36) Enki-2: it causes those who sublimmed syncronously to become romantically/sexually attracted
(2008-03-22 16:22:45) Enki-2: due to similar primings
(2008-03-22 16:22:54) kanzure: fenn: You think we could possibly use autogenix to automatically "fill in the blanks" (the outsourcing-boxes) for a self-replicator?
(2008-03-22 16:23:07) kanzure: once we sufficiently fill the database
(2008-03-22 16:23:18) Enki-2: i kind of feel bad using nicole as a guinea pig for that
(2008-03-22 16:23:38) Enki-2: since when i chanced upon the effect initially, i lost my virginity
(2008-03-22 16:23:54) kanzure: O.o
(2008-03-22 16:24:53) Enki-2: er
(2008-03-22 16:25:08) Enki-2: nicole was used in a further experiment
(2008-03-22 16:25:17) Enki-2: testing variations of the initial circumstance
(2008-03-22 16:25:51) Enki-2: i still am skeptical that her feelings (which are on-and-off) are caused only by that
(2008-03-22 16:26:21) Enki-2: since we had chemistry before that
(2008-03-22 16:26:44) Enki-2: becky and i did not; in fact, before the experiment, becky and i found each other rather unattractive
(2008-03-22 16:27:02) fenn: scary stuff
(2008-03-22 16:27:08) kanzure: fenn: answer :-p
(2008-03-22 16:27:14) kanzure: Enki-2: So you just set them up with xsublims?
(2008-03-22 16:27:32) fenn: i meant the hypnotism stuff
(2008-03-22 16:27:36) kanzure: fenn: right
(2008-03-22 16:27:48) kanzure: fenn: You think we could possibly use autogenix to automatically "fill in the blanks" (the outsourcing-boxes) for a self-replicator? once we sufficiently fill the database
(2008-03-22 16:28:00) kanzure: we can't "search for software" though
(2008-03-22 16:28:01) Enki-2: kanzure: yeah, basically
(2008-03-22 16:28:06) kanzure: so it'd be hard
(2008-03-22 16:28:11) Enki-2: fenn: it's not hypnosis really
(2008-03-22 16:28:13) fenn: there are many databases of components, with engineering drawings, performance specifications, etc
(2008-03-22 16:28:21) kanzure: we'd have to break it down into enough components so that others can conveniently work on the components that we just-so-happen to require
(2008-03-22 16:28:31) fenn: it would be feasible to import each database by hand as a bootstrapping source
(2008-03-22 16:28:32) Enki-2: fenn: it's just that having the same concepts primed will cause a feeling of being on the same wavelength
(2008-03-22 16:28:41) kanzure: fenn: by hand? yikes, but if that's what it takes
(2008-03-22 16:28:41) Enki-2: fenn: increased syncronity..., etc
(2008-03-22 16:28:49) fenn: oops
(2008-03-22 16:28:54) fenn: i mean, write the code to import the database by hand
(2008-03-22 16:29:03) kanzure: I want to see if we can do it in theory, so why not have a database of 'fake parts' and then see if we can sort through the fake parts in trying to make a proejct work
(2008-03-22 16:29:04) fenn: write the code by hand :)
(2008-03-22 16:29:19) kanzure: so what would a part specification be
(2008-03-22 16:29:28) kanzure: it would be something like the software specs in APT, no doubt
(2008-03-22 16:29:47) kanzure: http://en.wikipedia.org/wiki/Aptitude_(program)
(2008-03-22 16:30:03) fenn: yes that's the whole point of specifying requirements and capabilities
(2008-03-22 16:30:13) kanzure: http://en.wikipedia.org/wiki/Advanced_Packaging_Tool
(2008-03-22 16:30:26) kanzure: you can't specify emergent properties of biobricks though
(2008-03-22 16:30:26) kanzure: oh
(2008-03-22 16:30:28) fenn: for automatic resolution of unintended opportunities
(2008-03-22 16:30:31) kanzure: I guess we avoid that problem with mechanical parts
(2008-03-22 16:30:33) kanzure: there's no emergence :)
(2008-03-22 16:30:34) kanzure: awesome!
(2008-03-22 16:30:46) fenn: 'emergence' is a buzzword :P
(2008-03-22 16:30:51) kanzure: nah, think about it
(2008-03-22 16:30:55) kanzure: oh
(2008-03-22 16:30:57) kanzure: I have a good example
(2008-03-22 16:30:58) fenn: please elaborate
(2008-03-22 16:31:06) kanzure: DNA logic gates require RNA taggers to key into them and activate certain gates
(2008-03-22 16:31:15) kanzure: you don't want your logic gates all flipping each other simultaneously, right?
(2008-03-22 16:31:30) kanzure: so, emergence is where we realize that there's some novel function that comes up when we didn't get RNA sequences that don't interact with each other
(2008-03-22 16:31:38) kanzure: in some way that we didn't think to check for. That's a pretty good example.
(2008-03-22 16:31:39) fenn: non-deterministic structure?
(2008-03-22 16:31:51) kanzure: more like "we didn't think about trying to predict that"
(2008-03-22 16:31:57) kanzure: or "we couldn't have known"
(2008-03-22 16:32:10) kanzure: but in mechanical parts I think we avoid this problem
(2008-03-22 16:32:21) kanzure: we just say "this *will* work in this way if you put it in a circuit in this manner"
(2008-03-22 16:32:31) fenn: uh, automated innovation?
(2008-03-22 16:32:36) kanzure: hm?
(2008-03-22 16:32:49) kanzure: please elaborate
(2008-03-22 16:32:54) fenn: when you start talking about DNA logic gates it gets hard to follow
(2008-03-22 16:33:02) fenn: there are so many interpretations of what you're saying..
(2008-03-22 16:33:06) kanzure: hrm
(2008-03-22 16:33:12) kanzure: okay, let's talk about proteins then
(2008-03-22 16:33:22) fenn: no, i want a mechanical example :)
(2008-03-22 16:33:27) fenn: lego bricks? :)
(2008-03-22 16:33:28) kanzure: but there isn't, and that's the beauty of it
(2008-03-22 16:33:36) kanzure: if you put a lego down, it will not interact in some weird way
(2008-03-22 16:33:40) kanzure: you know how it's going to wokr.
(2008-03-22 16:33:41) kanzure: *work
(2008-03-22 16:33:44) kanzure: not so with biochem
(2008-03-22 16:33:46) fenn: um, cats and dogs
(2008-03-22 16:33:55) kanzure: cats and dogs are biochem-mechanical
(2008-03-22 16:33:57) fenn: a room full of cats and dogs
(2008-03-22 16:34:02) kanzure: that's bio though
(2008-03-22 16:34:09) fenn: with buttered toast on some of them
(2008-03-22 16:34:14) kanzure: heh
(2008-03-22 16:34:16) kanzure: what is your point?
(2008-03-22 16:34:29) Blak1: A biological hard vacuum would not be difficult to create, the shells of many mollusks are sufficiently strong to encase a vacuum
(2008-03-22 16:34:34) fenn: it's bio but it can be visualized and communicated much more easily
(2008-03-22 16:34:42) Blak1: biological pumps aren't impossible either
(2008-03-22 16:35:02) fenn: Blak1: how do you get the precision necessary for a good seal?
(2008-03-22 16:35:07) kanzure: fenn: ah, well, in the case of cats and dogs you have billions of neurons all interacting with each other and there's all sorts of mutations that can arise in their behavior since they are just doing molecular diffusion and so on, lots of statistics to account for
(2008-03-22 16:35:09) Blak1: a biological organ could manufacture parts out of iron and calcites
(2008-03-22 16:35:14) Blak1: and assemble them
(2008-03-22 16:35:15) kanzure: but in mechanical situations you don't have that problem really
(2008-03-22 16:35:19) kanzure: a lego is a lego
(2008-03-22 16:35:33) Blak1: you use a living glue
(2008-03-22 16:35:36) fenn: kanzure: ok, no. simulation is out of the scope of autogenix (although a simulator might call autogenix as a subroutine)
(2008-03-22 16:36:06) kanzure: fenn: 'erm. I don't think you understand the distinction that I was making. I was not saying that a simulation would be required. Sigh. Hold on.
(2008-03-22 16:36:10) fenn: Blak1: you need a good seal for the pumping action
(2008-03-22 16:36:22) kanzure: fenn: first, emergent properties will not be in autogenix since mechanical parts are generally well-defined and don't do any of that emergence stuff
(2008-03-22 16:36:37) Blak1: There's no reason something biologically constructed could not be precise
(2008-03-22 16:36:42) fenn: Blak1: a human could assemble a machine tool that could turn the iron and calcites.. but that's not what you're saying is it?
(2008-03-22 16:37:22) Blak1: I'm talking about building them up molecule by molecule
(2008-03-22 16:37:26) kanzure: fenn: secondly, simulation will be possible, in the sense of CFD simulations for spaceships and fluid flows. A part in skdb for a space ship might have various parameters like size, up to some extent, so that it's a template or partially variabilized, and if you want to
(2008-03-22 16:37:26) kanzure: do a mission simulation, then go do it. You will get back your numbers and you'll know if the parts work together sufficiently or not.
(2008-03-22 16:37:36) Blak1: mollusks build things out of mineral bricks of a few molecules
(2008-03-22 16:37:37) fenn: Blak1: there are good reasons, it's due to the multiple repeated analog steps - cell replication and mineral secretion
(2008-03-22 16:37:42) Blak1: with organic glue between
(2008-03-22 16:38:12) fenn: mollusks dont assemble molecules one by one, they lay down a grid of protein that approximates calcite's crystal structure
(2008-03-22 16:38:14) kanzure: Blak1: their building process has been *evolved* and finding a biochemical correlate to our linear programs to make things -- like a rocket engine out of biomolecules or even just out of cells -- will require either accelerated evolution with awesome selections or a giant simulation.
(2008-03-22 16:38:16) Blak1: many organic processes are very precise
(2008-03-22 16:38:31) Blak1: look at a baboon's arse
(2008-03-22 16:38:37) Blak1: it's blue; why?
(2008-03-22 16:38:54) Blak1: Answer that, and you'll see what kind of precision can be done by biological organisms
(2008-03-22 16:39:28) kanzure: their precision is *coincidental* -- a byproduct of natural evolution; you can't just hack wetware with a few simple commands heh'
(2008-03-22 16:39:59) fenn: it's blue because biology can create exact structures on the atomic scale
(2008-03-22 16:40:07) fenn: (chemistry)
(2008-03-22 16:40:28) Blak1: there are insects that have very exact small structures
(2008-03-22 16:40:35) Blak1: why can't it create a seal?
(2008-03-22 16:40:37) kanzure: but this is all because of biological mutation and testing otu different chemistries
(2008-03-22 16:40:39) kanzure: *out
(2008-03-22 16:40:50) kanzure: and the ones that happened to test it out in such a manner resulted in the blue ass
(2008-03-22 16:41:01) Blak1: It happens on it's own due to the nature of the molecules
(2008-03-22 16:41:01) kanzure: not because it was aiming for blue, necessarily
(2008-03-22 16:41:06) Blak1: to form coherent lattices
(2008-03-22 16:41:08) kanzure: and we are *aiming* for stuff
(2008-03-22 16:41:25) Blak1: we can aim for things, structure largely sorts itself out
(2008-03-22 16:41:35) kanzure: huh?
(2008-03-22 16:41:45) kanzure: then tell me the sequence to have an organism produce a rocket engine please :)
(2008-03-22 16:41:51) Blak1: it's silly to reinvent the wheel, since you get many of the same problems with larger machines
(2008-03-22 16:41:53) fenn: kanzure: now i'm picturing a recursive set of function calls between the package manager and the simulator..
(2008-03-22 16:42:05) Blak1: I have to go
(2008-03-22 16:42:09) Blak1: later
(2008-03-22 16:42:34) kanzure: fenn: maybe, maybe not. In some packages I expect there might be a rule for generating the part that meets specs, and in the cases where not this obviously will not be a feature
(2008-03-22 16:42:39) fenn: 'i need a pump' simulator gets a pump design from autogenix, loads it onto the simulated rocket 'pump mass distribution is wrong'
(2008-03-22 16:42:51) kanzure: hrm, would it be able to debug so easily?
(2008-03-22 16:42:57) fenn: no
(2008-03-22 16:43:14) kanzure: the pump package would provide its own simulator component
(2008-03-22 16:43:25) kanzure: and produce some sort of standardized output (whatever rocket designers programmed)
(2008-03-22 16:43:31) fenn: more likely a unit test for the rocket functionality would fail due to a simulated instability of some sort
(2008-03-22 16:43:43) kanzure: yeah, it's all unit testing
(2008-03-22 16:43:49) kanzure: and I think it's a legit idea to let packages specify their own unit tests
(2008-03-22 16:43:54) fenn: yes
(2008-03-22 16:44:08) fenn: instead of just dumb data
(2008-03-22 16:44:10) kanzure: then, in your overall project, run all of the unit tests together
(2008-03-22 16:44:27) kanzure: heh, with various debug modes like "assume this part works perfectly" ;)
(2008-03-22 16:44:35) Blak1 left the room.
(2008-03-22 16:45:42) fenn: people really underestimate the difficulties in metrology
(2008-03-22 16:46:15) kanzure: where's this coming from?
(2008-03-22 16:46:15) fenn: it's a whole set of bootstrapping in its own right
(2008-03-22 16:46:27) fenn: trying to go from a blue babboon ass to a vacuum pump
(2008-03-22 16:46:41) kanzure: hehe :)
(2008-03-22 16:46:46) kanzure: it's evolved, not engineered
(2008-03-22 16:47:33) fenn: kanzure: actually, you can hack wetware with a few simple commands :\
(2008-03-22 16:47:43) kanzure: not to the extent that Blake was wanting to
(2008-03-22 16:47:55) kanzure: you can't just give it an overall map and say "make this engine" as far as I can tell.
(2008-03-22 16:47:59) fenn: ah, wetware == brain software, traditionally
(2008-03-22 16:48:03) kanzure: oh
(2008-03-22 16:48:04) kanzure: sorry
(2008-03-22 16:48:10) kanzure: didn't know
(2008-03-22 16:48:20) kanzure: http://openwetware.org/ does synthetic bio stuff, not synthetic brain stuff
(2008-03-22 16:48:37) kanzure: (they are big into the biobricks stuff)
(2008-03-22 16:48:54) fenn: analogy to MIT open courseware
(2008-03-22 16:49:03) fenn: or whatever that linguistic construct is called
(2008-03-22 16:49:03) kanzure: ah, good eye
(2008-03-22 16:49:13) kanzure: didn't think of that
(2008-03-22 16:49:25) fenn: rhyme :)
(2008-03-22 16:49:56) kanzure: or like one of those "GNU is Not Unix" recursive acronyms / play on words sort of things
(2008-03-22 16:50:32) kanzure: hm
(2008-03-22 16:50:52) kanzure: so it looks like we need to first characterize a few components that we would want in a database
(2008-03-22 16:50:57) kanzure: like a pencil, and a spoon, and a fork
(2008-03-22 16:51:03) kanzure: and then figure out what we want to know about them in the database
(2008-03-22 16:51:14) kanzure: and then try to set up projects that involve all three of them to see what features we want
(2008-03-22 16:51:54) kanzure: although I am having trouble visualizing how this will lead to the parts required for self-replication
(2008-03-22 16:52:37) fenn: well, originally the idea was for cleaning up openfarmtech
(2008-03-22 16:52:50) kanzure: yeah, but I think we've both had this sort of idea before
(2008-03-22 16:52:57) fenn: so, an openfarmtech village could self-replicate (wonk)
(2008-03-22 16:52:58) kanzure: oh, does openfarmtech have all of their parts designed?
(2008-03-22 16:53:03) fenn: no
(2008-03-22 16:53:09) kanzure: bleh
(2008-03-22 16:53:16) fenn: they dont design everything either, so the analogy with debian
(2008-03-22 16:54:08) kanzure: I see.
(2008-03-22 16:54:22) fenn: they're more about functionality and raw materials usage
(2008-03-22 16:54:49) fenn: rather than buying components in a catalog that werent designed to work together in the first place
(2008-03-22 16:55:43) fenn: kanzure: self replication is a recursive dependency tree.. might not fit into the model i'm providing
(2008-03-22 16:56:20) kanzure: it seems to be more than a recursive dependency tree, since the tree just loops back to itself in some weird incestuous way.
(2008-03-22 16:57:10) fenn: if you dont need, say, aluminum, you can delete all the parts of the tree that only provide aluminum
(2008-03-22 16:57:32) kanzure: right, but some of those parts might have been helping to make other parts in the overall system
(2008-03-22 16:57:40) fenn: ^^ only
(2008-03-22 16:57:45) kanzure: ah
(2008-03-22 16:58:05) kanzure: but in that case, you *do* need aluminum, so you have to provide aluminum
(2008-03-22 16:58:06) kanzure: erm
(2008-03-22 16:58:13) kanzure: we need to work on the terminology to discuss these systems
(2008-03-22 16:58:20) fenn: ok
(2008-03-22 16:58:58) fenn: presumably there is some standard math terminology for discussing trees
(2008-03-22 16:59:33) kanzure: for example, the distinction made above is that there's a "the self-replicator already exists; it has aluminum subcomponents; we don't need aluminum for self-replication; so take out all parts that require aluminum; refactor the entire design to eliminate reliance on aluminum subsystems or subsystems made out of aluminum"
(2008-03-22 16:59:42) kanzure: as opposed to "just yank the subsystems that process aluminum"
(2008-03-22 17:00:28) fenn: hmm
(2008-03-22 17:00:36) kanzure: so: mineral-processing and minieral-reliant
(2008-03-22 17:00:43) kanzure: the mineral-processing subsystems take the raw minerals and process it
(2008-03-22 17:00:59) kanzure: while the mineral-relying subsystems are those that rely on other minerals for their *construction*
(2008-03-22 17:01:02) fenn: more like: 'the self replicator exists; it has aluminum components; we need aluminum for replication becase that's the design we have; refactor the design to eliminate reliance on alu'
(2008-03-22 17:01:07) kanzure: we can have another term for mineral-active-relying for "relying on it for active functionality"
(2008-03-22 17:01:14) kanzure: right
(2008-03-22 17:02:42) fenn: 'replicates using aluminum'?
(2008-03-22 17:02:56) kanzure: that's all-encompassing
(2008-03-22 17:03:06) fenn: meaning the design we're talking about uses alu somewhere in its replication process
(2008-03-22 17:03:19) kanzure: okay
(2008-03-22 17:03:27) kanzure: so Al-dependent replication?
(2008-03-22 17:03:45) fenn: functionality 'replication' depends on aluminum?
(2008-03-22 17:03:54) kanzure: right
(2008-03-22 17:04:55) fenn: so functionality becomes an optional package associated with the main design package?
(2008-03-22 17:05:17) fenn: i mean, if you modify the functionality, you might end up changing the entire tree
(2008-03-22 17:05:27) fenn: so the concept of packages doesnt work so well there
(2008-03-22 17:06:15) kanzure: hm, let me process that in a sec
(2008-03-22 17:06:25) kanzure: I just added this to the wiki: Note that the "sand-processing subsystem" need not be made out of sand. The "sand-processing subsystem" needs to be made out of X mineral where there exists in the design "X-processing subsystem" which, again,
(2008-03-22 17:06:26) kanzure: is not necessarily made out of X, but it has to be made out of something like "X" that eventually makes it full circle back to sand.
(2008-03-22 17:06:49) fenn: ... if you modify the functionality of leaf packages, it doesnt necessarily change the requirements of branch packages, but you can rebuild branch packages so as to have different requirements and capabilities
(2008-03-22 17:07:30) kanzure: please define leaf/branch/trunk in that context
(2008-03-22 17:07:36) kanzure: since branches, well, branch
(2008-03-22 17:08:02) fenn: in the context of a self replicator, the trunk is what is being replicated, and it's a capability of one of the leaves
(2008-03-22 17:08:15) fenn: our tree might have only one leaf
(2008-03-22 17:08:35) fenn: directed cyclic graph
(2008-03-22 17:08:39) kanzure: hm
(2008-03-22 17:08:48) kanzure: I am familiar with directed cyclic graphs
(2008-03-22 17:08:52) kanzure: are the vertices the functionality?
(2008-03-22 17:08:56) kanzure: or are the edges the functionality?
(2008-03-22 17:08:56) fenn: i like 'tree' though :)
(2008-03-22 17:09:40) fenn: uh, vertices are packages, edges are functionality being used
(2008-03-22 17:09:48) kanzure: interesting
(2008-03-22 17:10:14) kanzure: so what if we just have a giant graph
(2008-03-22 17:10:17) kanzure: where we through in tons of resources
(2008-03-22 17:10:22) kanzure: like different functions and different capabilities
(2008-03-22 17:10:30) kanzure: and then we have a computer program search for internal directed cyclic graphs
(2008-03-22 17:10:36) fenn: and fish around with a stick until we pull out something that's a ring
(2008-03-22 17:10:39) kanzure: right
(2008-03-22 17:10:45) kanzure: kind of inefficient
(2008-03-22 17:10:45) fenn: that's the general idea
(2008-03-22 17:10:58) kanzure: but so far it seems like we can't do this from first principles :)
(2008-03-22 17:11:08) kanzure: so that's always a backup method I guess.
(2008-03-22 17:11:11) fenn: we can use intelligence to only add packages we think are useful
(2008-03-22 17:11:15) kanzure: ah.
(2008-03-22 17:11:41) kanzure: but we'd have to express all possible functionality
(2008-03-22 17:11:47) kanzure: it's very hard to detail *all possible uses* of a mineral
(2008-03-22 17:11:52) fenn: indeed
(2008-03-22 17:11:58) fenn: that's where the simulator comes in
(2008-03-22 17:12:42) fenn: the simulator can enumerate different states the mineral might assume
(2008-03-22 17:12:49) kanzure: yikes, from basic chemistry?
(2008-03-22 17:12:52) fenn: (i dont know if such a thing is actually possible)
(2008-03-22 17:13:05) kanzure: this would be bad though
(2008-03-22 17:13:16) kanzure: we'd have to do O(n^(2)) testing
(2008-03-22 17:13:23) kanzure: to see what different reactions we can get with all of the different components
(2008-03-22 17:13:46) kanzure: although in organic chemistry we have a list of basic reactions
(2008-03-22 17:13:53) kanzure: plus a general way of getting the environment required for those basic reactions
(2008-03-22 17:14:00) kanzure: maybe this isn't so bad?
(2008-03-22 17:14:09) ***fenn shrugs
(2008-03-22 17:14:22) fenn: i dont think its impossible to figure out a self replicator by hand
(2008-03-22 17:14:34) fenn: freitas did it, right? :)
(2008-03-22 17:14:41) fenn: it was just unfeasibly large
(2008-03-22 17:15:03) kanzure: hrm
(2008-03-22 17:15:08) fenn: er, the self-replicating interstellar probe
(2008-03-22 17:15:10) kanzure: maybe that's where we should start? just reduce his design?
(2008-03-22 17:15:17) kanzure: because obviously he did not do any refactoring
(2008-03-22 17:16:52) fenn: he doesn't go into enough detail (because it would be too expensive/time consuming)
(2008-03-22 17:17:12) kanzure: so how do we know his design would work, then?
(2008-03-22 17:17:17) fenn: well, we dont
(2008-03-22 17:17:19) kanzure: heh
(2008-03-22 17:17:25) kanzure: I think the way that we're going, we basically know if it will work or not
(2008-03-22 17:17:43) fenn: what if it throws a monkey wrench in its own gears
(2008-03-22 17:17:54) fenn: 'i refuse to work for stinking organic life forms!'
(2008-03-22 17:18:01) kanzure: he didn't even go down to the detail of gears
(2008-03-22 17:18:11) fenn: no, just system-level enumeration
(2008-03-22 17:18:34) fenn: i guess you could call them packages
(2008-03-22 17:18:43) kanzure: I was hoping this would be a good way to figure out tools and processes for a mineral -- http://en.wikipedia.org/w/index.php?title=Special:WhatLinksHere/Bauxite&limit=500&from=0
(2008-03-22 17:18:49) kanzure: but apparently not.
(2008-03-22 17:19:57) fenn: CRC handbook, amateur science websites, i'm sure nasa has something but they might not want to share
(2008-03-22 17:20:42) fenn: http://en.wikipedia.org/wiki/Bayer_process
(2008-03-22 17:21:33) kanzure: so how would we formalize all of the information into our database/simulation/graph scenario
(2008-03-22 17:22:03) fenn: well, first of all it has to be quantitative (wikipedia is not)
(2008-03-22 17:23:26) fenn: i.e. instead of 'electricity' we need 100 amp*hours/kg at 2.6 volts, 300C and 20 molar solution of cryolite
(2008-03-22 17:23:32) kanzure: hehe
(2008-03-22 17:23:40) kanzure: and not only that
(2008-03-22 17:23:43) kanzure: but we need the context for it too
(2008-03-22 17:23:50) kanzure: as in, we need the wires for being able to do 100 amps/hr
(2008-03-22 17:23:58) fenn: some variables can be deduced from others (specific power input, amount of cryolite required, etc)
(2008-03-22 17:24:08) fenn: right, that's part of the 'source code'
(2008-03-22 17:24:10) kanzure: in the chemical reaction scenario
(2008-03-22 17:24:17) kanzure: suppose we have to simulate all of the possible different reactions
(2008-03-22 17:24:22) fenn: but what i just listed was the package requirements
(2008-03-22 17:24:24) kanzure: each of these reactions require chambers and environments
(2008-03-22 17:24:31) kanzure: and these chambers and environments require material resources as well
(2008-03-22 17:24:40) fenn: er, sorry, the package run-time requirements
(2008-03-22 17:25:08) fenn: build-time requirements depends on much more, generally
(2008-03-22 17:25:21) kanzure: build-time is subset of run-time?
(2008-03-22 17:25:36) fenn: no, run-time is a subset of build-time though, for testing purposes
(2008-03-22 17:25:50) fenn: unless you are so confident that it will work you ship it out the door without trying it out
(2008-03-22 17:26:12) fenn: uh, i guess run-time is a subset of testing then
(2008-03-22 17:26:45) kanzure: run-time of the replicator would mean it's daily operations
(2008-03-22 17:26:51) kanzure: it's not constantly replicating (is it?)
(2008-03-22 17:27:03) fenn: hmm
(2008-03-22 17:27:06) fenn: maybe
(2008-03-22 17:27:17) fenn: if it has no other leaf nodes, yes
(2008-03-22 17:27:53) kanzure: it has to wait for resource accumulation, you realize
(2008-03-22 17:27:59) kanzure: so it can't constantly be replicating, it will hit run-time errors
(2008-03-22 17:28:01) fenn: yes, that's part of replication
(2008-03-22 17:28:16) kanzure: okay
(2008-03-22 17:28:34) fenn: it won't assemble the new unit until the parts are built
(2008-03-22 17:29:14) kanzure: so,
(2008-03-22 17:29:33) kanzure: (1) can skdb and autogenix do the package/functionality requirements we need to be able to specify?
(2008-03-22 17:29:49) kanzure: and (2) can we do a small proof of our 'fishing' method for some other design project?
(2008-03-22 17:30:16) kanzure: also, it looks like for each package and functionality we add, we have to include everything that it could possibly be used to do
(2008-03-22 17:30:25) kanzure: and mention its relationships to all existing functionality
(2008-03-22 17:30:33) fenn: fishing the ring out of the swamp?
(2008-03-22 17:30:36) kanzure: right
(2008-03-22 17:30:49) fenn: there arent any replicators
(2008-03-22 17:31:09) fenn: or do you mean just fake it, and prove the algorithm works?
(2008-03-22 17:31:20) kanzure: the algorithm for fishing?
(2008-03-22 17:31:21) fenn: s/prove/demonstrate/
(2008-03-22 17:31:27) fenn: yes
(2008-03-22 17:31:27) kanzure: yes, I suppose
(2008-03-22 17:31:34) kanzure: but with packages and functionality in there
(2008-03-22 17:31:39) kanzure: where *we* know a solution
(2008-03-22 17:31:42) kanzure: but we want the computer to find it
(2008-03-22 17:31:43) fenn: re:1, skdb/autogenix only exist in our minds right now ;)
(2008-03-22 17:32:10) kanzure: yes, but I mean to say that we get to do functional specs for skdb/autogenix and make sure the file formats could possibly define all of this information that we need
(2008-03-22 17:32:39) kanzure: I guess this is going to be ad-hoc really
(2008-03-22 17:32:44) kanzure: since there's all sorts of variables we will want to include
(2008-03-22 17:32:57) fenn: the design can change if it's needed
(2008-03-22 17:32:58) kanzure: perhaps we will make it abstract
(2008-03-22 17:33:07) fenn: we can do .agx version 1, version 2, etc
(2008-03-22 17:33:08) kanzure: in the sense that we can "include package-required-for-electricity-variables"
(2008-03-22 17:33:24) kanzure: and then other variables can be further subunits which will each add functionality to the overall project
(2008-03-22 17:34:20) epitron: http://kol.coldfront.net/thekolwiki/index.php/Time_sword
(2008-03-22 17:34:39) fenn: what variables? diesel for your diesel generator?
(2008-03-22 17:34:53) kanzure: fenn: as well as flow and other information about that fluid
(2008-03-22 17:34:58) kanzure: there's tons of information that CRC keeps on such stuff
(2008-03-22 17:35:05) kanzure: but we can't know all of these variables from first principles
(2008-03-22 17:35:11) fenn: true
(2008-03-22 17:35:13) kanzure: so when we want to include a new variable for a new functionality
(2008-03-22 17:35:20) kanzure: we should be able to just add a new 'library' addon thing
(2008-03-22 17:35:21) kanzure: or extension
(2008-03-22 17:35:31) kanzure: which would include subcomponents that would be called upon during simulation
(2008-03-22 17:35:39) kanzure: I think this is the 'virtual function' idea in C/C++
(2008-03-22 17:35:50) kanzure: or template factory
(2008-03-22 17:35:51) kanzure: not sure.
(2008-03-22 17:36:23) fenn: terminology update: streams are run-time capabilities/requirements, packages are build-time, packages can provide streams
(2008-03-22 17:37:00) kanzure: packages still vertices?
(2008-03-22 17:37:08) fenn: yes
(2008-03-22 17:37:49) kanzure: packages deliver streams to the new unit?
(2008-03-22 17:38:05) fenn: i guess i wasnt specific enough
(2008-03-22 17:38:15) kanzure: no, I think you were
(2008-03-22 17:38:19) fenn: streams are things like materials, energy, information
(2008-03-22 17:38:31) fenn: i havent figured out if they're vertices or edges yet
(2008-03-22 17:39:01) fenn: there's going to be a lot of streams that have no package providing them, like sunlight, air
(2008-03-22 17:40:05) fenn: the aluminum smelter relies at run-time on the clay processor to provide bauxite
(2008-03-22 17:40:18) fenn: smelter, processor = packages
(2008-03-22 17:40:26) fenn: clay, bauxite = streams
(2008-03-22 17:40:41) kanzure: streams sound kind of like vectors
(2008-03-22 17:41:10) fenn: then you can have a factory object that makes packages
(2008-03-22 17:41:21) fenn: a package that makes packages
(2008-03-22 17:41:37) kanzure: including itself?
(2008-03-22 17:41:44) fenn: let's not go there :)
(2008-03-22 17:41:44) kanzure: isn't that what we wanted in the first place?
(2008-03-22 17:41:54) fenn: yes of course
(2008-03-22 17:42:37) fenn: but there are lots of intermediate packages that make other packages (i guess)
(2008-03-22 17:43:00) fenn: maybe that's just poor refactoring on society's part
(2008-03-22 17:43:28) kanzure: that was my 'criss-crossing' example: where package X requires stream Y during build-time to reproduce package X in the new unit, but in normal operations, package Y would require stream X, or something
(2008-03-22 17:43:34) kanzure: erm
(2008-03-22 17:43:46) kanzure: criss-crossing:
(2008-03-22 17:43:47) kanzure: Note that the "sand-processing subsystem" need not be made out of sand. The "sand-processing subsystem" needs to be made out of X mineral where there exists in the design "X-processing subsystem" which, again, is not necessarily made out of X, but it has to be made out of something like "X" that eventually makes it full circle back to sand.
(2008-03-22 17:44:10) fenn: ah a sub-loop
(2008-03-22 17:44:17) kanzure: so, package X (which processes Y) needs to be made out of stream L where stream L is
(2008-03-22 17:44:24) kanzure: I am having trouble variabilizing it, because it's a hard statement to make
(2008-03-22 17:44:38) fenn: the loop is in stream-space, whereas the package tree is very linear and loop-free
(2008-03-22 17:45:10) fenn: maybe draw a diagram
(2008-03-22 17:45:18) epitron: TIME SWORD!
(2008-03-22 17:45:20) kanzure: the dependency loops are important, of course
(2008-03-22 17:45:27) kanzure: let's see
(2008-03-22 17:45:44) fenn: the subtle knife
(2008-03-22 17:47:05) kanzure: (($mineralX)-processing-subsystem) can be made out of $mineralY, where there exists in the overall design a (($mineralY)-processing-subsystem), which either must immediately be made out of ($mineralX) or some (($mineralI-don't-care)-processing-subsystem) which eventually is made out of ($mineralX)
(2008-03-22 17:47:08) kanzure: where:
(2008-03-22 17:47:18) kanzure: you would probably say the mineralX-processing-subsystem is a package, yes?
(2008-03-22 17:48:11) kanzure: and then streams would be tapped during build-time by a certain package to manufacture a new package, for the replicated entity to have.
(2008-03-22 17:50:19) fenn: i'm not convinced this is a problem we have to worry about
(2008-03-22 17:50:32) kanzure: how so?
(2008-03-22 17:51:17) fenn: do you think a loop will kill the dependency resolution code?
(2008-03-22 17:51:26) kanzure: no?
(2008-03-22 17:51:33) kanzure: it's not a problem
(2008-03-22 17:51:36) kanzure: we're just formalizing what we need
(2008-03-22 17:54:03) kanzure: right?
(2008-03-22 17:58:30) fenn: yah
(2008-03-22 18:02:05) fenn: its funny, usually i'm trying to cut apart dependency loops, but for self-replication they're quite useful
(2008-03-22 18:03:09) kanzure: the more the better (in terms of design)
(2008-03-22 18:03:14) kanzure: but the more the worse, in terms of implementation
(2008-03-22 18:12:50) fenn: grr do they really expect me to pay money for nasa publications?
(2008-03-22 18:14:47) kanzure: here's my attempt at defining a package:
(2008-03-22 18:15:26) kanzure: a package must (1) be made of a certain material, (2) has an input stream that gives it a material to make, (3) this input material is different than the material in #1, and (4) a package must make *some other package* with the input stream material.
(2008-03-22 18:16:16) fenn: package must make a package?
(2008-03-22 18:16:20) kanzure: yep
(2008-03-22 18:16:31) kanzure: must make *another* package
(2008-03-22 18:16:37) fenn: what about instruments
(2008-03-22 18:16:38) kanzure: and that other package might end up making the original package
(2008-03-22 18:16:53) kanzure: a package might be a fabricator, and that fabricator, by coincidence, can make extra stuff
(2008-03-22 18:16:56) kanzure: like an instrument
(2008-03-22 18:17:11) fenn: i mean sensors, like an oscilloscope or something
(2008-03-22 18:17:24) fenn: a spectrometer
(2008-03-22 18:17:43) fenn: what does the spectrometer make?
(2008-03-22 18:17:47) kanzure: right, it makes nothing
(2008-03-22 18:17:51) kanzure: so these are your leaves
(2008-03-22 18:17:54) kanzure: leafs
(2008-03-22 18:17:59) fenn: it makes information, which is necessary for replication
(2008-03-22 18:18:04) kanzure: interesting
(2008-03-22 18:18:21) fenn: i feel funny calling data a package
(2008-03-22 18:18:48) kanzure: instrument == package?
(2008-03-22 18:19:04) fenn: well sure, it's a collection of components that performs a function
(2008-03-22 18:20:19) fenn: and there are control instruments that dont produce 'information' exactly
(2008-03-22 18:20:32) fenn: the computer that's directing operations
(2008-03-22 18:20:41) kanzure: arrgh
(2008-03-22 18:20:47) kanzure: I just wrote down "Reverse Method of Making a Fabricator"
(2008-03-22 18:20:51) kanzure: (1) start with a fabricator
(2008-03-22 18:20:54) kanzure: 'erm, replicator
(2008-03-22 18:20:58) kanzure: (2) Break it down into its components.
(2008-03-22 18:21:03) fenn: well gee with #1 we're already done :)
(2008-03-22 18:21:04) kanzure: (3) Make a machine to do the reverse. Done.
(2008-03-22 18:21:06) kanzure: heh'
(2008-03-22 18:21:08) kanzure: yeah :(
(2008-03-22 18:21:28) kanzure: man, relying on biology really helps though
(2008-03-22 18:21:33) kanzure: it gives you something to start with
(2008-03-22 18:21:42) kanzure: because if you want to be able to create enough for a human to live and grow
(2008-03-22 18:21:51) kanzure: then you can just require your materials for self-replication rely on that same stuff
(2008-03-22 18:21:58) fenn: human is too complex, start with a minimal bacterial cell
(2008-03-22 18:22:09) kanzure: okay, what's agar made out of?
(2008-03-22 18:22:29) kanzure: unbranched polysaccharide?
(2008-03-22 18:22:32) fenn: input streams: amino acids, sugar, nucleic acids, lipids, salts
(2008-03-22 18:22:45) kanzure: `Agar is a heterogeneous mixture of two classes of polysaccharide: agaropectin and agarose`
(2008-03-22 18:23:04) fenn: now we need proteins to import each of those streams into the cell
(2008-03-22 18:23:28) fenn: then there are assemblies of proteins that turn those streams into higher order molecules, like proteins and nucleic acids
(2008-03-22 18:23:36) kanzure: huh?
(2008-03-22 18:23:45) fenn: nucleic acid polymers :)
(2008-03-22 18:23:46) kanzure: we don't need those proteins, do we?
(2008-03-22 18:23:52) kanzure: just make agar for the bacteria and that's it
(2008-03-22 18:24:02) fenn: i'm talking about how the bacterium works
(2008-03-22 18:24:04) kanzure: hell, I am sure we can make bacteria that can eat moon rock
(2008-03-22 18:24:13) kanzure: that's hacking the bacterium
(2008-03-22 18:24:17) kanzure: not really necessary IMO
(2008-03-22 18:24:24) fenn: no, as a model for how replicators work in general
(2008-03-22 18:24:26) kanzure: ah
(2008-03-22 18:24:41) kanzure: okay, so then the packages that we would define
(2008-03-22 18:24:45) kanzure: would they be genes and protein expressions ?
(2008-03-22 18:24:58) fenn: take te bacteria apart, abstract and generalize the functions and structure, put it back together using other technologies
(2008-03-22 18:25:08) kanzure: remember the Minimal Cell Project?
(2008-03-22 18:25:27) fenn: i've heard of it, not read much though
(2008-03-22 18:25:28) kanzure: http://heybryan.org/mediawiki/index.php/Minimal_cell_project
(2008-03-22 18:25:32) kanzure: lucky for you, I have a page
(2008-03-22 18:25:59) kanzure: we could try to model the whole damn cell, I guess
(2008-03-22 18:26:09) kanzure: and then figure out how to apply our terminology to it
(2008-03-22 18:26:12) fenn: yeah
(2008-03-22 18:28:33) kanzure: isn't that kind of cheating, shouldn't we be able to come up with the model ourselves?
(2008-03-22 18:28:39) kanzure: and why can't we use von Neumann's work on theoretical self-replicators?
(2008-03-22 18:28:48) kanzure: surely there must be somebody who has been exploring the mathematics of this topic
(2008-03-22 18:29:06) fenn: i was just trying to make sense of your 'reverse method'
(2008-03-22 18:30:01) kanzure: we're operating on many different levels here
(2008-03-22 18:30:04) kanzure: ah
(2008-03-22 18:30:09) kanzure: well, I see what you mean in that case
(2008-03-22 18:30:17) kanzure: the other reverse methods that I have been proposing are different of course
(2008-03-22 18:30:26) kanzure: such as starting with a fabricator arm that can move "completed components" to assemble the fabricator arm itself
(2008-03-22 18:30:36) kanzure: and then adding in the closed-form dependency loops to make all of the components on the shelf
(2008-03-22 18:30:51) kanzure: which is a reverse recursive definition method, but not quite the same thing as starting with a cell instead
(2008-03-22 18:31:05) fenn: that's the reprap way, but unfortunately plastic goo isnt good for shit
(2008-03-22 18:31:13) fenn: so you have to start with something that has potential
(2008-03-22 18:31:45) fenn: and legos, have you ever tried to manufacture something using only legos?
(2008-03-22 18:32:06) kanzure: nope
(2008-03-22 18:32:10) kanzure: but
(2008-03-22 18:32:27) kanzure: I wonder if it would be possible to make a lego factory out of legos
(2008-03-22 18:32:29) fenn: i saw a youtube where a 'factory' was assembling legos.. looked godawful inefficient
(2008-03-22 18:32:37) fenn: the factory was made of legos too
(2008-03-22 18:32:46) kanzure: it surely had some electronics or someting
(2008-03-22 18:32:47) kanzure: *something
(2008-03-22 18:32:56) fenn: yes there's an electronics brick
(2008-03-22 18:33:17) fenn: i guess you could use control cams like in a screw machine (not that legos lend themselves to that method)
(2008-03-22 18:33:45) kanzure: Google isn't being kind to me - not much info on synthesis of agar
(2008-03-22 18:33:50) kanzure: or agarose for that matter
(2008-03-22 18:33:56) fenn: agar is made of agarose :)
(2008-03-22 18:33:59) kanzure: but it occurs to me that there does exist such things as aerobic bacteria
(2008-03-22 18:34:03) fenn: its refined from seaweed, fuchus i think
(2008-03-22 18:34:06) kanzure: hrm
(2008-03-22 18:34:07) kanzure: that's not good
(2008-03-22 18:34:17) fenn: agarose is not a nutrient
(2008-03-22 18:34:18) kanzure: how does fuchus make it?
(2008-03-22 18:34:23) kanzure: ooh
(2008-03-22 18:34:24) fenn: it's used for bacteria because they can't digest it
(2008-03-22 18:34:28) kanzure: well, we can just have giant water tanks pointed at the sun
(2008-03-22 18:34:31) kanzure: oh?
(2008-03-22 18:34:35) fenn: so you get a nice smooth plate with bumps sitting on top
(2008-03-22 18:35:00) fenn: instead of having the bacteria forming branching colonies inside the agar
(2008-03-22 18:35:12) fenn: well, they do that too, but it doesnt break down at least
(2008-03-22 18:35:24) fenn: if the agar broke down it would turn into a pool of slime
(2008-03-22 18:37:21) fenn: lego factory: http://www.youtube.com/watch?v=GQ3AcPEPbH0
(2008-03-22 18:41:11) fenn: with some odometry that whole thing could be replaced by a car with a gripper
(2008-03-22 18:44:12) fenn: the minimal cell project has too much information processing in the form of molecules, it's like a factory made of cams and levers
(2008-03-22 18:44:35) fenn: so it's a bad model to follow for macro-scale replicators
(2008-03-22 18:44:40) kanzure: take a look at this
(2008-03-22 18:44:41) kanzure: http://heybryan.org/mediawiki/index.php/Self-replication#Sipper_excerpt
(2008-03-22 18:46:02) kanzure: http://en.wikipedia.org/wiki/Langton's_loops
(2008-03-22 18:46:07) fenn: sure but it relies on the cellular grid being just right
(2008-03-22 18:46:14) fenn: the rules of the universe, if you will
(2008-03-22 18:49:01) fenn: von neumann's design separates structure from information, which is desirable when dealing with a fuzzy non-digital universe
(2008-03-22 18:49:39) kanzure: http://ti.arc.nasa.gov/people/jlohn/bio.html <-- good guy to know
(2008-03-22 18:49:57) kanzure: "the rules of the universe, being just right" --> With any sufficient preparation ...
(2008-03-22 18:49:58) fenn: so now we have DNA and proteins, instead of an RNA-based ribozyme mishmash
(2008-03-22 18:51:21) fenn: neat stuff
(2008-03-22 18:52:12) kanzure: http://lslwww.epfl.ch/pages/embryonics/ interesting
(2008-03-22 18:54:07) fenn: evolution is good at circuit-hiding :\
(2008-03-22 18:54:41) kanzure: it doesn't look like anybody has actually focused on the mathematics of self-replication, away from cellular automata
(2008-03-22 18:54:49) kanzure: in fact, why the hell has cellular-automata been used in the first place
(2008-03-22 18:54:53) kanzure: this should have been graph based since the beginning
(2008-03-22 18:55:10) fenn: because cellular automata is a math-based representation of 'the real world'
(2008-03-22 18:55:32) kanzure: so if we had a cellular automata configuration file that was a self-replicator
(2008-03-22 18:55:36) fenn: also because, in math you can just say 'a = b' and you're done
(2008-03-22 18:55:37) kanzure: how would we use that to help us find a material implementation
(2008-03-22 18:56:08) fenn: well, remember they came up with the information tape before DNA was discovered
(2008-03-22 18:56:32) kanzure: so?
(2008-03-22 18:56:54) fenn: what if we hadn't discovered DNA?
(2008-03-22 18:57:24) kanzure: heh, we'd still be stuck on CAs.
(2008-03-22 18:57:38) fenn: i guess they had paper tape already
(2008-03-22 18:57:56) fenn: anyway, i dont know what the CA simulation is good for
(2008-03-22 18:58:09) kanzure: nothing as far as I can tell
(2008-03-22 18:58:28) kanzure: looks like we were going in the right direction with our definitions/graph-theory approach
(2008-03-22 18:59:11) kanzure: with the extendible simulator to generate posisble reactions and interactions between streams so that we can figure out if we have missed any optimizations
(2008-03-22 19:00:17) fenn: btw have you ever seen animations of the vonneumann replicator? it's really cool to watch
(2008-03-22 19:04:52) kanzure: nope
(2008-03-22 19:05:05) fenn: well too bad, you probably never will :P
(2008-03-22 19:07:12) kanzure: hm, this is interesting - 1. W.M.Stevens "NODES: An Environment for Simulating Kinematic Self-Replicating Machines" Proc. of the Ninth International Conference on the Simulation and Synthesis of Living Systems (ALIFE9) 39-44, 2004.
(2008-03-22 19:07:30) kanzure: the guy's web site - http://www.srm.org.uk/home.html
(2008-03-22 19:08:11) fenn: yep found it
(2008-03-22 19:10:41) fenn: looks quite biological
(2008-03-22 19:11:56) fenn: havent any of these guys heard of a tape reel? :)
(2008-03-22 19:13:42) kanzure: http://www.landesbioscience.com/books/special/id/912
(2008-03-22 19:13:49) kanzure: $150. I may just have to go beg Freitas.
(2008-03-22 19:13:56) kanzure: Or maybe Max can lend me his copy, if he has it.
(2008-03-22 19:16:29) fenn: it's on his website
(2008-03-22 19:16:47) fenn: http://www.molecularassembler.com/KSRM.htm
(2008-03-22 19:17:22) fenn: i need to read that too
(2008-03-22 19:17:29) kanzure: let's wiki it
(2008-03-22 19:17:42) fenn: wiki what? the whole thing? :)
(2008-03-22 19:18:38) fenn: gosh there are a lot of citations
(2008-03-22 19:20:21) kanzure: yep
(2008-03-22 19:29:32) kanzure: http://heybryan.org/mediawiki/index.php/KSRM
(2008-03-22 19:29:34) kanzure: my start to it
(2008-03-22 19:30:30) fenn: this book must be rather heavy
(2008-03-22 19:31:14) kanzure: Freitas is a peculiar fellow. I saw him on video once, he had no clue as to how to run a presentation. He was reading straight from a transcript, looking down, was not energetic, very monotonous tone. But if he's really all that he's cracked up to be ...
(2008-03-22 20:22:19) kanzure: fenn: http://heybryan.org/mediawiki/index.php/KSRM/3
(2008-03-22 20:22:22) kanzure: it's getting all messed up
(2008-03-22 20:22:35) kanzure: http://heybryan.org/projects/ksrm/output.txt has the entire wiki code for the book, sort of
(2008-03-22 20:22:39) kanzure: but it's all out of order I think
(2008-03-22 20:23:13) kanzure: oh wait, nevermind
(2008-03-22 20:41:06) kanzure: hm,
(2008-03-22 20:41:10) kanzure: Freitas cites Winfree.
(2008-03-22 20:42:19) kanzure: 1175. Brent A. Ridley, B. Nivi, Joseph M. Jacobson, “All-inorganic field effect transistors fabricated by printing,” Science 286(22 October 1999):746-749; http://www.media.mit.edu/molecular/Science10-99.pdf
(2008-03-22 20:42:37) kanzure: 1987. Kinneret Keren, Rotem S. Berman, Evgeny Buchstab, Uri Sivan, Erez Braun, “DNA-templated carbon nanotube field-effect transistor,” Science 302(21 November 2003):1380-1382, 1310 (comment). See also: “DNA used to create self-assembling nano transistor,” ISRAEL21c, 23 November 2003; http://www.israel21c.org/bin/en.jsp?enPage=BlankPage&enDisplay=view&enDispWhat=object&enDispWho=Articles%5el557&enZone=Technology&enVersion=0&
(2008-03-22 21:06:38) kanzure: fenn: it occurs to me that we can use humans as the 'black box' for as long as we want for self-replication, slowly porting over certain functions to the actual hardware, until eventually the machine replicates on its own
(2008-03-22 21:08:39) fenn: reading the printed FET paper.. i wonder if the nanocrystals could be aligned by an externally applied electric field
(2008-03-22 21:09:51) fenn: kanzure: it's more than just 'humans' it's industrial civilization
(2008-03-22 21:10:09) kanzure: as long as we can move all of the functionality over to the human, we're good
(2008-03-22 21:10:16) fenn: i dont know how to make aluminum from scratch :P
(2008-03-22 21:10:19) kanzure: what printed FET paper?
(2008-03-22 21:10:25) kanzure: are you reading my nanocrystals page?
(2008-03-22 21:10:37) fenn: all inorganic field effect transistors fabricated by printing
(2008-03-22 21:10:54) fenn: you just linked to it
(2008-03-22 21:12:36) kanzure: oh
(2008-03-22 21:12:38) kanzure: well
(2008-03-22 21:12:57) kanzure: more on semiconductor nanocrystals - http://heybryan.org/alternate_transistors.html
(2008-03-22 21:13:01) kanzure: i.e., printed FETs.
(2008-03-22 21:13:09) kanzure: but this isn't the DNA-based FETs.
(2008-03-22 21:14:12) fenn: i cant load the israel21c.org url
(2008-03-22 21:15:02) kanzure: http://web.archive.org/web/*/http://www.israel21c.org/bin/en.jsp?enPage=BlankPage&enDisplay=view&enDispWhat=object&enDispWho=Articles%5el557&enZone=Technology&enVersion=0&
(2008-03-22 21:15:12) fenn: did you ever read about qtc pills?
(2008-03-22 21:17:44) kanzure: "The DNA serves as a scaffold, a template that will determine where the carbon nanotubes will sit," Braun said.
(2008-03-22 21:17:45) kanzure: hmm
(2008-03-22 21:17:49) kanzure: nope
(2008-03-22 21:18:35) kanzure: looks like pressure-based switches?
(2008-03-22 21:18:41) fenn: yes
(2008-03-22 21:18:57) fenn: they're rubber-state switches
(2008-03-22 21:20:23) fenn: well, variable resistors
(2008-03-22 21:21:00) fenn: the fabrication sounds really low tech
(2008-03-22 21:21:16) fenn: mix nickel powder and adhesive
(2008-03-22 21:21:50) kanzure: woah
(2008-03-22 21:21:54) kanzure: that's definitely worth it
(2008-03-22 21:21:57) kanzure: why haven't I seen more of this
(2008-03-22 21:22:08) kanzure: searching for 'qtc pills' doesn't show up much but a single company's website
(2008-03-22 21:22:13) kanzure: can you get me some other links or something ?
(2008-03-22 21:22:20) fenn: it's fairly new, and there's only one company developing it (the inventor)
(2008-03-22 21:22:46) fenn: hell if i know why it isnt more widespread
(2008-03-22 21:38:01) fenn: popular article on the history of qtc http://www.sep.org.uk/catalyst/articles/catalyst_18_1_333.pdf
(2008-03-22 21:43:54) kanzure: hm
(2008-03-22 21:45:45) fenn: i wonder if you could make a composite material that was both a qtc-resistor and a piezo actuator
(2008-03-22 21:45:58) fenn: so a small electric field would change the conductivity
(2008-03-22 21:46:30) fenn: it would probably have to be anisotropic, like a fiber embedded in silicone rubber
(2008-03-22 21:46:31) kanzure: how does a piezo actuator work?
(2008-03-22 21:46:38) fenn: good question
(2008-03-22 21:46:49) kanzure: is it just a simple case of applying a current and making it move?
(2008-03-22 21:47:35) kanzure: I am pretty sure piezo actuator == piezo buzzer
(2008-03-22 21:47:51) kanzure: `http://www.sep.org.uk/catalyst/articles/catalyst_18_1_333.pdf`
(2008-03-22 21:47:58) kanzure: oops
(2008-03-22 21:47:59) kanzure: Inkjet printers: On some inkjet printers, particularly those made by Epson, piezoelectric crystals are used to control the flow of ink from the inkjet head to the paper.
(2008-03-22 21:48:11) fenn: i think piezoelectric materials deform due to voltage, but there is a certain amount of charge that must be transfered (as if it were a capacitor)
(2008-03-22 21:48:43) kanzure: Piezoelectric motors: piezoelectric elements apply a directional force to an axle, causing it to rotate. Due to the extremely small distances involved, the piezo motor is viewed as a high-precision replacement for the stepper motor.
(2008-03-22 21:48:44) fenn: yeah the piezo inkjets physically squirt the ink onto the page
(2008-03-22 21:48:54) kanzure: yes, piezoelectrics work just like that
(2008-03-22 21:49:03) fenn: ugh, no, ignore the piezoelectric motor snippet
(2008-03-22 21:49:04) kanzure: but I have not figured out how that applies to piezo steppers / actuators
(2008-03-22 21:49:07) kanzure: oh?
(2008-03-22 21:49:15) kanzure: because that doesn't make sense
(2008-03-22 21:49:18) kanzure: if you apply a voltage
(2008-03-22 21:49:21) kanzure: it will deform and push the axil
(2008-03-22 21:49:26) kanzure: but then when you stop the voltage
(2008-03-22 21:49:30) kanzure: it should just move back to where it started
(2008-03-22 21:49:31) kanzure: which makes no sense :)
(2008-03-22 21:49:39) fenn: you can cause rotational motion, but not _continuous_ rotation
(2008-03-22 21:49:46) fenn: unless you have some ratchet mechanism or something
(2008-03-22 21:49:55) kanzure: I doubt they are using ratchets
(2008-03-22 21:50:01) kanzure: because you're using twice the energy
(2008-03-22 21:50:06) kanzure: that you need.
(2008-03-22 21:50:47) kanzure: where's Superkuh when you need him? heh'
(2008-03-22 21:50:54) kanzure: he's a self-made expert in piezos last time I checked
(2008-03-22 21:51:01) kanzure: was designing a piezo-based particle accelerator
(2008-03-22 21:51:14) fenn: you could use a piezo actuator as the replacement for a piston in an engine (sorta)
(2008-03-22 21:51:22) fenn: ah i bet he was doing something with fusion
(2008-03-22 21:51:38) fenn: lithium deuteride crystals yes?
(2008-03-22 21:51:41) kanzure: so how would you cause rotational motion with piezos?
(2008-03-22 21:51:42) kanzure: let me chekc
(2008-03-22 21:52:08) kanzure: Ferroelectric/pyroelectric particle accelerators and research with homebrew methods.
(2008-03-22 21:52:17) kanzure: [context pressure ~=10^-3 Pa] Has anyone considered using pyroelectric crystal mediated acceleration of charged particles (spontaneous polarization of voltage oriented through the c-axis crystal faces (when heating the , +z surface has a negative potential and the -z surface has a positive potential, reverse for cooling) as the result of shifting ions in the crystal shape/lattice combined with funky dilute gas charge masking behavior, ref: main, http://arxiv.org/abs/physics/0404113, bipolar setup for higher energies, more detailed, even more detailed)?
(2008-03-22 21:52:23) kanzure: Particle (electrons or ions depending on the crystal face and heat/cool cycle) energies up in the 150keV+ range are very doable (or 200kev+ with dilute light gases and a bipolar setup). Plus with cylindrical crystals the beams are self focusing. A common choice for this is Lithium Tantalate, and making thin films and stacking this material is possible, but there are lots of tricks involved (and making crystal boules is a magnitude harder):
(2008-03-22 21:52:49) fenn: that's pyroelectric, not piezoelectric
(2008-03-22 21:53:04) kanzure: yeah .... oops.
(2008-03-22 21:53:12) kanzure: but
(2008-03-22 21:53:18) kanzure: he's the guy that introduced me to piezoelectricity
(2008-03-22 21:53:23) fenn: i see
(2008-03-22 21:54:18) fenn: well, anyway, there are piezoelectric polymers (kynar/polyvinylidene)
(2008-03-22 21:54:52) kanzure: hm, that's convenient
(2008-03-22 21:54:55) fenn: but the most promising are ceramic
(2008-03-22 21:54:58) kanzure: but how does the rotational motion work that you were mentioning?
(2008-03-22 21:55:18) fenn: a piston in a car engine goes up and down, so replace the piston with a piezo actuator
(2008-03-22 21:55:29) fenn: the crank turns reciprocating motion to rotation
(2008-03-22 21:55:36) fenn: there's more ways to do it
(2008-03-22 21:55:53) kanzure: huh
(2008-03-22 21:55:57) kanzure: so if you just do rapid voltage pulses
(2008-03-22 21:55:58) fenn: like a clock wheel, hit each tooth with a hammer as it goes by
(2008-03-22 21:56:06) kanzure: you'd convert this into rotational motion with a crank
(2008-03-22 21:56:21) fenn: i saw a 'wiggle motor' somewhere
(2008-03-22 21:56:59) fenn: it's a piezo rod that wiggles its way around, using a gearing principle similar to harmonic drives to provide a decent output torque
(2008-03-22 21:59:44) fenn: it was about 5mm wide 20mm long
(2008-03-22 21:59:56) fenn: cylinder
(2008-03-22 22:00:29) kanzure: not much turning up on Google.
(2008-03-22 22:02:14) fenn: http://www.newscaletech.com/squiggle_overview.html
(2008-03-22 22:05:21) kanzure: that's really convenient
(2008-03-22 22:05:50) kanzure: I wonder how we could make it.
(2008-03-22 22:06:04) kanzure: It looks like just a piezo + voltage + ... I don't know what.
(2008-03-22 22:06:17) kanzure: either a rachet or a crank
(2008-03-22 22:07:28) kanzure: from Superkuh's bookmarks - http://www.pollin.de/shop/shop.php?cf=detail.php&pg=NQ==&a=MDA5OTQ3OTk= giant piezo crystals
(2008-03-22 22:07:37) kanzure: http://www.chem.pacificu.edu/Johnson/JohnsonResearch/STM/PIEZO.HTM piezo tube
(2008-03-22 22:07:52) kanzure: http://www.colomar.com/Shavano/mini_piezo_tweeter.html DIY mini piezo tweeter
(2008-03-22 22:08:27) kanzure: Hm, in my trash I apparently have a Google search for pyroelectrical transistors - http://www.google.com/search?client=opera&rls=en&q=pyroelectric+transistor&sourceid=opera&ie=utf-8&oe=utf-8
(2008-03-22 22:09:25) fenn: wow that's cheap (pollin.de)
(2008-03-22 22:10:35) kanzure: maybe we should stock up ;)
(2008-03-22 22:12:53) fenn: the piezo tube looks like what the squiggle motor uses
(2008-03-22 22:13:35) fenn: heh, go back on that page it says 'inchworm motor housing' for STM
(2008-03-22 22:15:08) kanzure: the piezo tube is for motion in x/y/z directions, not up/down, although ... if we could find a piezo that could change shape by say, 5 mm or something, that would be great
(2008-03-22 22:15:15) kanzure: or I guess it doesn't have to be too much
(2008-03-22 22:15:20) kanzure: it could just be a rubber band stretching system
(2008-03-22 22:15:25) fenn: oh i guess 'inchworm motor' is the stick-slip thing i keep reading about
(2008-03-22 22:15:42) kanzure: http://en.wikipedia.org/wiki/Inchworm_motor
(2008-03-22 22:16:02) fenn: why 5mm? what do you plan to do with it?
(2008-03-22 22:16:13) kanzure: well
(2008-03-22 22:16:19) kanzure: a 1 m piezo would be great
(2008-03-22 22:16:23) kanzure: as a piston or something :)
(2008-03-22 22:16:39) kanzure: http://www.physikinstrumente.com/en/news/fullnews.php?newsid=123&Wiki_iw I hate this website, but it's about the principles of the inchworm motor
(2008-03-22 22:16:50) kanzure: physikinstrumente spams all over the place
(2008-03-22 22:16:57) kanzure: tons of domain squatting
(2008-03-22 22:17:47) kanzure: if this piezo motor can scale from 1 Newton to up to thousands of Newtons that'd be great
(2008-03-22 22:18:49) fenn: a motor might have force and speed but not at the same time
(2008-03-22 22:19:18) fenn: but why does it matter? what are you going to use it for?
(2008-03-22 22:19:35) kanzure: huh?
(2008-03-22 22:19:41) kanzure: I was assuming you had a few ideas
(2008-03-22 22:19:45) kanzure: for using it in a replicator
(2008-03-22 22:20:16) fenn: well, i never considered piezo's to be useful for producing mechanical power
(2008-03-22 22:20:24) fenn: that might be a mistake on my part
(2008-03-22 22:20:54) kanzure: it looks fairly convenient
(2008-03-22 22:21:17) fenn: did you notice the efficiency chart? looks like they are much less efficient when size goes up
(2008-03-22 22:21:51) kanzure: hrm
(2008-03-22 22:22:04) kanzure: qtc-resistor + piezo would be really useful though
(2008-03-22 22:22:07) fenn: on the right side of here http://www.newscaletech.com/doc_downloads/SQUIGGLE_Overview_2-18-08.pdf
(2008-03-22 22:22:08) kanzure: combines a lot of parts
(2008-03-22 22:22:57) fenn: yes, i was thinking mainly piezos for squeezing qtc's to control a more conventional motor
(2008-03-22 22:23:16) fenn: like, a pancake motor or switched reluctance
(2008-03-22 22:24:01) fenn: voltage is easy to get from a tiny computer chip
(2008-03-22 22:24:28) fenn: anyway this all needs some empirical data
(2008-03-22 22:24:48) fenn: i've never used either qtc's or piezo's
(2008-03-22 22:27:45) kanzure: the qtc research group should be responsive
(2008-03-22 22:27:57) kanzure: http://www.dur.ac.uk/psm.group/qtc.html
(2008-03-22 22:30:04) fenn: i'd like to know if there are other metals besides nickel that form nano-spikes
(2008-03-22 22:31:09) kanzure: I need to take a break from this stuff and go figure out a circuit for Andy ... he wants me to come up with something that has an emergent property
(2008-03-22 22:31:13) kanzure: but this is pretty hard
(2008-03-22 22:31:30) kanzure: I am thinking about telling him about my 'evolvable DNA logic circuits' idea, but I think that's sidestepping his request
(2008-03-22 22:32:01) fenn: something with an emergent property eh
(2008-03-22 22:32:07) kanzure: heh'
(2008-03-22 22:32:14) kanzure: he cited the ring oscillator as the "classic example"
(2008-03-22 22:32:15) fenn: cats and dogs and buttered toast!
(2008-03-22 22:32:18) kanzure: indeed
(2008-03-22 22:33:22) kanzure: the oscillator example is really simple and takes up a huge class of possibilities
(2008-03-22 22:33:25) kanzure: so I can't do anything that oscillates
(2008-03-22 22:33:35) kanzure: I was thinking about trying to implement sine
(2008-03-22 22:33:43) kanzure: but that's analog, whereas this is discrete-state
(2008-03-22 22:34:14) fenn: forget about emergence, come up with something that has 'a spiritual property'
(2008-03-22 22:34:45) fenn: they're both meaningless words
(2008-03-22 22:35:53) kanzure: then what should I implement
(2008-03-22 22:36:01) kanzure: calculations are mostly meaningless in this context
(2008-03-22 22:36:29) kanzure: although the prospects are interesting for large-scale computing really
(2008-03-22 22:36:35) kanzure: because if you have a few million gates for a single circuit
(2008-03-22 22:36:41) kanzure: and want to do simultaneous computation all at once
(2008-03-22 22:36:46) kanzure: then you have them all extremely localized
(2008-03-22 22:37:08) kanzure: if you have it acting in a centrifuge then it would be interesting to store the result of the computation eventually, but I don't see how
(2008-03-22 22:37:16) kanzure: maybe a flip flop system
(2008-03-22 22:37:32) kanzure: but there's no guarantee of saturation or propagation of all of the same states to all of the flip flops
(2008-03-22 22:37:32) kanzure: hrm
(2008-03-22 22:38:51) kanzure: maybe a massive cellular automata implementation
(2008-03-22 22:39:01) kanzure: where the logic gates are the cells
(2008-03-22 22:39:46) kanzure: each logic gate has four neighbors, and they are all executed in parallel, so what would happen?
(2008-03-22 22:39:50) fenn: *drool* http://sourceforge.net/projects/xandra
(2008-03-22 22:40:24) kanzure: awesome
(2008-03-22 22:42:29) fenn: how do you connect the neighbors? through winfree's toehold addressing mechanism?
(2008-03-22 22:48:22) kanzure: yes
(2008-03-22 22:48:25) kanzure: and in this scenario,
(2008-03-22 22:48:29) kanzure: you don't have to be too specific either
(2008-03-22 22:48:34) kanzure: since you can have four neighbors for each 'cell'
(2008-03-22 22:48:35) kanzure: heh
(2008-03-22 22:48:45) kanzure: so that means I don't have to be as strict when selecting the toeholds and the recognition sequences
(2008-03-22 22:49:02) kanzure: sneaky, isn't it?
(2008-03-22 22:50:18) fenn: screenshot from xandra showing a hexaglide actuator: http://imagebin.org/15164
(2008-03-22 22:51:33) fenn: i dont get it. the reason you have to be strict when selecting toehold sequences is to reduce crosstalk
(2008-03-22 22:51:36) kanzure: yes
(2008-03-22 22:51:40) kanzure: well
(2008-03-22 22:51:41) kanzure: not only that
(2008-03-22 22:51:42) fenn: crosstalk will bunch up your fabric
(2008-03-22 22:51:52) kanzure: but you also don't want one toe-hold to match for the wrong logic gate
(2008-03-22 22:51:55) kanzure: it's kind of like a lock-and-key method
(2008-03-22 22:52:02) kanzure: where the logic gate output is a key to the next one in the pathway
(2008-03-22 22:52:10) kanzure: so if your key unlocks the 4 neighboring cells ..
(2008-03-22 22:52:22) kanzure: where 'cells' are really just gates
(2008-03-22 22:54:21) fenn: 'cell' as in cellular automata right? not bacterial cell
(2008-03-22 22:54:52) fenn: blah my brain is hurting
(2008-03-22 22:56:24) kanzure: right
(2008-03-22 22:56:26) kanzure: heh
(2008-03-22 22:56:27) fenn: another xandra screenshot http://imagebin.org/15165
(2008-03-22 22:56:44) kanzure: so what does it simulate
(2008-03-22 22:56:54) kanzure: just a 3D visualization?
(2008-03-22 22:57:03) fenn: machine tool kinematics
(2008-03-22 22:57:08) kanzure: ooh
(2008-03-22 22:57:36) fenn: windows only right now or i'd be playing with it
(2008-03-22 22:58:00) kanzure: wine?
(2008-03-22 23:07:36) kanzure: fenn: you wouldn't happen to remember how to use the X11 screenshot feature to take a snapshot of a particular program?
(2008-03-22 23:08:53) fenn: i use ksnapshot
(2008-03-22 23:09:13) kanzure: I need to automate it. I think there was a program named xsd or something.
(2008-03-22 23:09:31) fenn: try `man import`
(2008-03-22 23:09:41) kanzure: okay
(2008-03-22 23:10:28) fenn: with import you have to click on the window..
(2008-03-22 23:12:34) fenn: xwd is what you're remembering
(2008-03-22 23:16:00) kanzure: yes
(2008-03-22 23:16:03) kanzure: thank you
(2008-03-22 23:16:10) kanzure: turns out xplanet ( http://xplanet.sf.net/ ) does image output itself
(2008-03-22 23:16:11) kanzure: so all is good
(2008-03-22 23:16:32) kanzure: a guy from the Moon Society wants me to write a quick script to show the earth from the moon at the given moment
(2008-03-22 23:16:41) kanzure: he said his web dev team has been at it for two years
(2008-03-22 23:16:47) kanzure: I laughed and set it up in 2 minutes :(
(2008-03-22 23:17:14) fenn: two years!
(2008-03-22 23:17:35) kanzure: script kiddies do better than two years ...
(2008-03-22 23:22:28) fenn: i feel like i've been doing JITT (just in time training) for a few years now
(2008-03-22 23:22:43) fenn: exhibiting symptoms of 'jitt-stick'
(2008-03-22 23:23:16) fenn: wikipedia overdose
(2008-03-22 23:27:54) kanzure: JITT indeed
(2008-03-22 23:28:02) kanzure: but what can be done about that
(2008-03-22 23:30:52) fenn: mobile computing for exercise (faster removal of brain waste products, increased oxygen)
(2008-03-22 23:31:26) fenn: might cause more harm than good though, you could walk into the street or something