• Log in with Facebook Log in with Twitter Log In with Google      Sign In    
  • Create Account
  LongeCity
              Advocacy & Research for Unlimited Lifespans

Photo
- - - - -

Chat For Jan 19th 2003


  • Please log in to reply
4 replies to this topic

#1 Bruce Klein

  • Guardian Founder
  • 8,794 posts
  • 242
  • Location:United States

Posted 20 January 2003 - 04:50 AM


<BJKlein> Official Chat Begins Now...
<Gordon> no, they're all downloaded, just not displaying correctly :-(
<OcsRazor> Back on
<BJKlein> someone crank up the campfire...
<OcsRazor> I have a pretty serious laser here that should do the job
<MRAmes> Question #1... Has anyone here *already* got improved parts?
<BJKlein> Sun Jan. 19th - Topic: "Better Than Original Parts" ;)
* Gordon starts throwing people in, chokes on the smoke
<BJKlein> let me post the intro... just a sec::
<MRAmes> oh okay
<OcsRazor> Improved No, but good replacements...
<BJKlein> RE: http://www.imminst.o...t...;f=63&t=695
<OcsRazor> One artificial heart patient sill alive and kicking
<Gordon> MRA: well, I've got one improvement part, you know, for the ladies ;-)
<BJKlein> The current state and future promise of replacement human body parts.
<BJKlein> The current state and future promise of replacement human body parts. It is very likely that the ability of engineers to design biological tissues, bio-artificial hybrids, and better-than-biological synthetic replacments for human tissues will surpass the design capabilities of mother nature within our lifetimes.
<BJKlein> Discussion topics will include
<BJKlein> What is the current state of the technology in tissue engineering, organ replacements, and silico-neural interfaces?
<BJKlein> What is the impact of nanotechnology on these
<BJKlein> advances?
<BJKlein> When can we expect these advance to be made available
<BJKlein> to the public?
<BJKlein> How will these technologies affect human lifespan and
<BJKlein> quality of life?
<BJKlein> What are the ethical implications of better-than-human enhancement?
<BJKlein> Bring your questions, comments, and hot bioengineering
<BJKlein> tech news.
<BJKlein> (Credit: Peter Passaro)
<BJKlein> aka - OcsRazor
<BJKlein> ok thanks.. now we can fire away :)
<OcsRazor> OK gang, just to let you know who I am...
<OcsRazor> grad student working on neural implants...
<MRAmes> Ocs: Did you implant any in yourself yet?
<OcsRazor> used to be a science analyst for venture capitalists on aging research
<OcsRazor> MR i'd like to test it on someone else first ;^)
<MRAmes> Ocs: ETA on that?
<BJKlein> FYI: ImmInst's bio on Passaro: http://www.imminst.o...l.php?siteid=66
<OcsRazor> Two people already have implants in their cortex...
<OcsRazor> http://www.wired.com....09/vision.html
<MRAmes> Hmm... tell more...
<OcsRazor> crude, but functioning
<MRAmes> I read the article already... was it for real?
<MRAmes> Never can tell with Wired, eh?
<OcsRazor> yep, but not very good science
<OcsRazor> the real deal is still in cats and monkeys
<MRAmes> Very interesting.
<OcsRazor> My lab works in vitro, disassociated mouse brains in culture
<OcsRazor> brains on a chip ;^0
<MRAmes> Last year I was the director of a singing chorus in Ottawa, Canada.
<BJKlein> how difficult is it to know who's really at the front in this race to create BCI
<MRAmes> Of *blind* singers.
<BJKlein> is Japan doing this...
<MRAmes> I wondered if I should tell any of them about this yet... I held off.
<BJKlein> or other techno savy conutries... how far along are they.
<OcsRazor> one good lab in Japan, Jimbo lab (no kidding)
<OcsRazor> One good lab in Germany
<OcsRazor> Best labs are here though, much better funding
<MRAmes> Some of the chorus members would be very interested... as you may imagine. But I didn't want to get thier hopes up if the tech was too far away.
<OcsRazor> a german company makes all our equipment though...
<OcsRazor> http://www.multichannelsystems.com
<MRAmes> Ocs. Are the chips you are using still surface-mounted?
<OcsRazor> yep, but...
<OcsRazor> there is an implantable type also commercially available
<OcsRazor> we are focusing on outside the animal technologies
<MRAmes> I only heard of the 'tongue' jolter... who manufactures your chips?
<BJKlein> The Honda Asimov seems like a huge step forward for Japan.. it seems to be much better than anything we can come up with.
<OcsRazor> http://www.multichannelsystems.com/
<BJKlein> correction Honda"Asimo"
<dreamingamoeba> just saw the commercial on tv
<OcsRazor> in what way better?
<BJKlein> http://world.honda.com/robot/
<BJKlein> portable power supply.. cordination... faster computing..
<BJKlein> all the robotic thingies one would need for a transhumanized person
<OcsRazor> They are good at hardware, but our labs are much better at software
<OcsRazor> for example...
<MRAmes> BJ: Well, not quite... but definitely progress.
<OcsRazor> I just saw a talk by a carnegie Mellon group...
<OcsRazor> that used Sony Aibos to create a team that can play soccer
<BJKlein> yeh.. i've seen the Discover Channel story on that..
<BJKlein> they plan to beat a human team
<OcsRazor> http://www.cmu.edu/home/news/robo.html
<dreamingamoeba> heh, soccer
<OcsRazor> the robotics already exist to create robust artificial limbs
<MRAmes> BJ: Isn't that a human *controlled* team of robots that they are trying to beat?
<BJKlein> what would you think is the greatest motivator for your group OcsRazor? helping humans? etc?
<OcsRazor> you are correct MR
<BJKlein> actually, they were talking about eventually beating "humans"
<OcsRazor> with Aibos?
<dreamingamoeba> the rational altruist :D
<MRAmes> Not gonna happen.
<BJKlein> yeh... if I remember right
<BJKlein> long term of course
<MRAmes> Oh.. well thats a given.
<BJKlein> but, point is.. that was the stated long term objective
<OcsRazor> On motivation, I can't speak for anyone else in my group, for
<OcsRazor> me the motivation started about 2 years ago
<OcsRazor> after working in longevity for a while I realized...
<BJKlein> to eventually take a group of robotic bots networked .. to beat a human team.. that would definately take a Honda Asimo robotic team
<BJKlein> on steroids ;)
<dreamingamoeba> welcome
<OcsRazor> Hallo
<OcsRazor> so....
<BJKlein> yes.. we're listening intently
<BJKlein> i'm cursious at to your motivation
<OcsRazor> biotech isn't going to cut it for extremely extended lifespans
<MRAmes> I would hope that the motivator for creating implantable chips is profit.
<OcsRazor> that too ;^)
<OcsRazor> enhancing intelligence is one of my primary motivators
<MRAmes> If the tech is proven, then I'd ante-up for a cranial jack... hell! I'd moonlight to raise the cash.
<OcsRazor> although, I very interested in artificial limb and sensory organ implants in the short term
<caliban> Sorry to intrude into the discussion as a newcomer but I am curious: what do you mean by "working in longevity" @OcsRazor?
<OcsRazor> I was a science analyst for venture capital for a while
<OcsRazor> specialized in aging biotech
<BJKlein> Welcome Back caliban
<MRAmes> Ocs: Any insight you can give us on the most promising areas in that?
<MRAmes> Near term.
<MRAmes> 2-5 years.
<OcsRazor> in neuroengineering...
<MRAmes> Logevity
<MRAmes> n
<OcsRazor> Longevity, I would say best bet is probably at the Buck Inst.
<MRAmes> What are they working on?
<OcsRazor> Some serious anti-ox drugs in development
<caliban> @BJ thanks! J @OcsRazor: What an intriguing job! Why did it end?
<OcsRazor> http://www.buckinstitute.org
<OcsRazor> No more $ ;^)
<OcsRazor> and some serious mismanagement
<MRAmes> :/
<caliban> As these things go... still a very interessting position. Are there many people working in this area? I suppose you have a science background?
<OcsRazor> Not many jobs anymore...
<OcsRazor> I have a BS in microbiology and an MS in Molecular Neuro
<OcsRazor> working on the PhD
<MRAmes> Dissertation topic chosen yet?
<caliban> Indeed. May I ask where and what topic?
<OcsRazor> GaTech...
<OcsRazor> Analysis of signals from Multielectrode Array Neural Recordings
<OcsRazor> http://www.imminst.o...t...
<OcsRazor> that was the post I put on my work
<OcsRazor> I'm part of a multi-displinary NeuroEngineering Lab
<MRAmes> Hmm... interesting. What are you views about non-invasive interfaces of the sort Kurzwiel has talked about?
<Eliezer> OcsRazor: Are you familiar with William Calvin's hypothesis for signal organization in neural microcolumns?
<Eliezer> The Cerebral Code?
<OcsRazor> I'm not a huge fan of Kurzweil's in the neuro dept.
<OcsRazor> its not his hypothesis, and its not veyr likely
<OcsRazor> Calvin, I mean
<Eliezer> yeah... I also thought it unlikely
<OcsRazor> he borrowed that from another researcher
<OcsRazor> yes computation is modular but...
<Eliezer> but it did occur to me that if you could intercept neurons in microcolumns in his triangular array you could conceivably test his theory
<MRAmes> Ocs: Is it that the control/detection isn't fine enough with non-invasive, or do you have more fundamental doubts?
<OcsRazor> we would like to get finer resolution chips
<Eliezer> Ames: Personally, I'd worry that invasive has vastly more potential for improvement than noninvasive ever will, and it would be sad if noninvasive solutions temporarily outcompeted invasive ones - it'd get the field stuck at a local optimum
<OcsRazor> main problem is software though
<MRAmes> A worry yes. But I have the same worry from the opposite point of view.
<MRAmes> Ocs: no kidding.
<OcsRazor> noninvasive is too messy to deal with signal processing
<Eliezer> Ocs: Have you thought about doing two-way intercommunication between neurons in situ and neurons in a dish, to see if the neurons in a dish organize themselves in any interesting way that you could probe at leisure later?
<Eliezer> I.e., see if the dish neurons learn to do the signal processing
<Eliezer> then put them under a microscope to see what they're doing
<OcsRazor> yep, we are already thinking about dish-dish comm
<Eliezer> any thoughts on dish-brain?
<OcsRazor> dish-animal is a little further off
<OcsRazor> 2 groups have good in vivo stuff going though
<OcsRazor> Normann at Utah, and Nicolelis at Duke
<Eliezer> how bout snipping a section of animal brain, putting it in a dish, linking it to the animal brain, seeing if it rewires itself, then snooping the traffic and examining the structure of the dish network?
<OcsRazor> the in vivo groups already snoop traffic...
<OcsRazor> problem is... too much signal
<Eliezer> (This all reflecting my thought that neurons already know how to interpret neurons, and we should snoop as much of that information as possible because reinventing the decoding process from scratch is likely to be hideously hard.)
<Eliezer> what do you mean when you say "too much signal"?
<Eliezer> too much to be computationally feasible to process?
<MRAmes> Eliezer: I imagine doing the same thing non-invasively... and as Ocs pointed out, there is currently so much signal that the analysis of it is the bottleneck.
<Eliezer> significant signal hidden among a lot of insignificant signal?
<Eliezer> data that is just too complex to begin to parse because there's no incremental path that lets you start at simple things and work up to complex things?
<OcsRazor> right now 60 channels of data coming from 10,000 neurons
<OcsRazor> we are looking at broad patterns right now
<OcsRazor> neurons are essentially probabilistic
<OcsRazor> so we might not need to understand all of it
<OcsRazor> but no one has yet come up with a good model...
<OcsRazor> for this size network
<MRAmes> And even if they did, it would have to mapped to individual brains... very hard.
<BJKlein> are you guys using one processing languge to do the calculations?
<OcsRazor> not necessarily
<OcsRazor> Matlab mainly
<MRAmes> Ocs: Insights on mapping?
<Eliezer> that doesn't sound to me like "too much signal", it sounds like "we don't have any idea how to decode it"
<OcsRazor> el - both ;^)
<Eliezer> but in what sense is there "too much" signal?
<Eliezer> that's what I still don't understand
<OcsRazor> Right now it is human eyes doing the crunching
<OcsRazor> we are just beginning to apply software to try and find patterns
<BJKlein> Matlab right?
<OcsRazor> we need a good neural (software) net...
<OcsRazor> Matlab, sometimes Mathematica
<BJKlein> and is it functional enough? or do you have to go into the processing code
<Eliezer> what you really need, dear sir, is an AI
<OcsRazor> we right our own code
<MRAmes> You are attempting to deduce the patterns via emulation?
<BJKlein> wow
<Eliezer> and what I really need is a neural implant so I can build AI faster
<BJKlein> that's gotta be time consuming
<Eliezer> darned chicken-and-egg problem
<OcsRazor> Amen brother
<OcsRazor> MR trying to find the pattern
<BJKlein> so you're really a programmer 80% of the time
<BJKlein> ?
<OcsRazor> right nw I'm more of a mathematician
<OcsRazor> and a biologist
<Eliezer> hm... don't mean to sound depressing, Ocs, but without an AI, I can think of some scenarios where trying to sight-read the signal would *never* work
<Eliezer> have you made any progress yet?
<BJKlein> does your work output look like graphs most of the time? as in energy spikes.
<OcsRazor> we are beginning to understand what kind of math to use
<OcsRazor> voltage spikes
<OcsRazor> yes...
<BJKlein> real time?
<OcsRazor> yep, but we look at it after
<BJKlein> ohh yeh.. the videos.. do you guys watch those often?
<BJKlein> timelaps videos
<OcsRazor> we are coming up with all sorts of visualization tools
<BJKlein> speed up the process from days to minutes
<OcsRazor> we need to slow it down for human eyes
<MRAmes> It sounds a bit like you're trying to learn 'brain' language while knowing not the syntax *or* the semantics. Not much of the syntax anyway...
<MRAmes> Tough job.
<OcsRazor> good summary
<BJKlein> so you can actually see the neurons growing in vitro and "connect"
<OcsRazor> reverse engineering neural computation
<OcsRazor> yep
<OcsRazor> we have kick ass microscopy here too
<BJKlein> visual ?
<OcsRazor> we are starting to look at the physical network connectivity now
<BJKlein> sorry, light or micron
<OcsRazor> light, cells are large
<BJKlein> gotcha
<BJKlein> what do you feed the stuff
<OcsRazor> horse serum nad other stuff
<BJKlein> do you have any online photos?
<OcsRazor> hold on
* BJKlein is very happy
<OcsRazor> http://www.neuro.gat...ter/movies.html
* MRAmes reminds BJ to stoke the fire
* BJKlein brrrrr
<OcsRazor> firing up the 2-photon laser
* BJKlein produces interal body heat and radiates to other
<caliban> hm can I slip in a brief question about the topic since I must be of?
<OcsRazor> shoot
<OcsRazor> check out this too http://www.neuro.gat...tter/HSCCD.html
<MRAmes> Nice video... can you explain a bit about what we see on it?
<BJKlein> yeh. it seems the neurons are pulsating
<OcsRazor> which one?\
<MRAmes> The 2-photon one.
<Guest> nick/John_Ventureville
<BJKlein> add a space heh
<BJKlein> welcome John
<Guest> thanks!
<Guest> what did I miss?
<Guest> lol
<OcsRazor> the 2-photon is a an overall view of a network
<OcsRazor> its a fast scanning technique
<MRAmes> Are we seeing physical structure only?
<MRAmes> s/physical/visual/
<OcsRazor> pretty much, that stain is the whole neuron one
<MRAmes> okay
<BJKlein> we're talking with OcsRazor/aka Peter: http://www.imminst.o...l.php?siteid=66
<OcsRazor> the entire cell is filled with fluorophore
<BJKlein> to change your nick type: /nick John V.
<OcsRazor> 2-photon allows much longer recording times
<Guest> my dyslexia strikes again!
<Guest> thanks
<BJKlein> heh, np
<OcsRazor> we also are now able to do voltage sensitive recordign
<BJKlein> btw, john we need to add you to Gilgamesh
<John_Ventureville> Gilgamesh?
<MRAmes> Ocs: *that* would be interesting.
<BJKlein> if you get a chance, please submit a bio (link is on the frontpage) http://www.imminst.org/gilgamesh
<John_Ventureville> I will
<BJKlein> Ahh, Thank You.
<John_Ventureville> cool
<OcsRazor> we are planning on doing long term elec and visual recodings
<OcsRazor> I'm building an environment right now to house the cells
<MRAmes> What kind of frame rate can you get?
<OcsRazor> on volatge sensitive it is upwards of 1000/s
<BJKlein> do you see a change if you change the horse serum? like add Vitamin C or something?
<OcsRazor> yep the recipe has been finely tuned
<OcsRazor> neurons are very sensitive
<BJKlein> that's very interesting.. in terms of understanding the effects of chemicals and nutrients on lifespan...
<OcsRazor> I believe we have the record for long term neural culture
<MRAmes> Ocs: How robust are the cultures? How long do they last before dying?
<OcsRazor> we have one that is two years and still going
<BJKlein> wow
<OcsRazor> but...
<MRAmes> woooah..
<BJKlein> can you see the telomere ends ?
<OcsRazor> these are from mice which only live 3 years tops
<BJKlein> or I guess the cells are not dividing
<OcsRazor> telomeres don't shorten in neurons
<BJKlein> so it would not matter
<OcsRazor> yep
<BJKlein> brain cells dont divide period right?
<OcsRazor> not neurons, there are glia in there too
<OcsRazor> they might divide, not sure of the literature on that
<BJKlein> what types of chemicals do you add to fix the food mixture.
<OcsRazor> something called glutamax, salts, and a sugar solution
<BJKlein> and you have to keep the temp stable im sure also
<OcsRazor> yep, 37 C on the nose
<BJKlein> ehh watch the sugar, to much sugar kills
<OcsRazor> we use a minimal media...
<MRAmes> I thought that was *fermented* sugar BJ.
<BJKlein> refined
<OcsRazor> which means just enough of everything
<OcsRazor> neurons like glucose though ;^)
<BJKlein> true.. complex sugar is better
<BJKlein> have you tried CR yet?
<BJKlein> on the neurons
<OcsRazor> that is essentially what they get
<John_Ventureville> So if I understand correctly (I got here after the movie started) this research is about keeping brain cells alive in a tissue culture, or is there more to it then that?
<OcsRazor> Check out my post in the forum ....
<BJKlein> John_Ventureville: http://www.imminst.o...p?act...&hl=&s=
<OcsRazor> neural implants John
<BJKlein> btw, John is over in AZ working on a Cryonics community
<John_Ventureville> wow!
<MRAmes> Ocs: How do you see the sectioning affecting the signal results? The fact that brain are lumps, and you are analysing slices... ?
<OcsRazor> it is just easier to experiment on slices...
<OcsRazor> there is a 3-d culture project underway here
<MRAmes> Yeah, but what is lost? Any ideas?
<OcsRazor> we know we are losing some significant strucutre
<OcsRazor> the brain is organized in layers...
<OcsRazor> with processing duties probably seperated by layers
<MRAmes> Glad to hear of the 3-D work. How are signals detected within your 3-D cultures?
<OcsRazor> on towers with electrodes along them
<MRAmes> Invasive probes?
<OcsRazor> these are still in the design phase...
<OcsRazor> two types in design in vivo and in vitro
<OcsRazor> in vivo is invasive, but in vitro we will grow
<OcsRazor> Ga Tech MEMS is putting them together
* MRAmes likes the sound of MEMS :)
<OcsRazor> they also have a circulatory system built in
<MRAmes> For nutrient delivery?
<OcsRazor> yep, just like a blood supply
<John_Ventureville> *dinner time- will be back*
<OcsRazor> bye
<MRAmes> wb
<BJKlein> OcsRazor do you know Mark Allen, PhD.. per chance
<OcsRazor> doesn't ring a bell
<BJKlein> he's working on a mems project... www.cardiomems.com
<OcsRazor> very cool
<BJKlein> the project leder is my brother-in-law there at tech
<OcsRazor> looking
<BJKlein> the device basically is a monitor for blood presure
<OcsRazor> some of our elec engs are probably involved
<BJKlein> yeh.. I dont see a personell list, but the project is animal testing now
<OcsRazor> some of the mems stuff is going to make our job a whole lot easier
<MRAmes> Ocs: Before I forget... thanks very much for dropping in an accepting a grilling... very good to talk to someone on the neuro R&D front-lines, as it were.
<OcsRazor> your welcome, I'll try and show for future firesides
* BJKlein duplicates MRAmes sentiment
<OcsRazor> Mark Allens office is across the street from me here
<BJKlein> as you can see, we're hungry for knowledge.. and you may feel a little used
<BJKlein> but we'll still respect you in the morning
<Eliezer> everything I need to know in life, I learned by killing smart people and eating their brains
<OcsRazor> intellectually ravaged
<MRAmes> Ocs: Tonight has been very business-like... the discussion is often somewhat lighter :) ... like that.
<OcsRazor> I don't mind
<MRAmes> :)
* BJKlein has instructed all members to elevate of get out
<Eliezer> incidentally, Ocs, which section of brain are you trying to read?
<Eliezer> visual?
<Eliezer> V1?
<OcsRazor> we use whole cortex or hippocampus
<OcsRazor> not specific right now
<Eliezer> but then... what could you be trying to decode?
<OcsRazor> in vivo people use visual and motor cortex
<BJKlein> and do you have to kill the little guys?
<Eliezer> or are you just looking for empirical regularities without any specific target to correlate them with?
<OcsRazor> we are trying to deduce the general laws of neural comp
<MRAmes> Signal classification.
<OcsRazor> we are sending and receiving signals, watching the networks grow and learn to process info
<Eliezer> but there'll always be some specific class of neuron you're using in any in vitro dish
<Eliezer> Purkinje neurons are not amacrine cells
<OcsRazor> I'm doing the classification right now...
<OcsRazor> by staining for certain cell types
<OcsRazor> the intitial work was just proof of concept
<OcsRazor> now we are getting specific
<MRAmes> I imagine it will get a lot more interesting when you can focus in on a *developed* brain, and snoop the signal live. When will that happen do you think?
<Eliezer> what kind of signals are you sending and receiving?
<OcsRazor> we are finding out what they "like"
<Eliezer> "like"?
<OcsRazor> what causes them to form networks and robust signal patterns
<OcsRazor> some stim never produces these patterns
<BJKlein> so Peter's goal is to find emergent properties... has this goal been writen down somewhere? or is there a mission statement for the project?
<BJKlein> sorry.. not Peter.. Steven's
<BJKlein> Potter
<OcsRazor> let me look at potter's site
<Eliezer> hm... okay, I can see what you mean by deducing general laws of neural comp
<OcsRazor> http://www.neuro.gatech.edu/potter/ is his site
<Eliezer> though I think it might be wiser to think of it as the specific laws of a specific kind of cortex, and generalize only when you find that the law empirically holds elsewhere
<OcsRazor> each of the projects has a description
<Eliezer> and "what maximally causes the network to form 'robust' signal patterns for our definition of 'robust'" is a very interesting fact, but not one that is necessarily identical to any design principle within the brain
<OcsRazor> I'm beginning the biological characterization by look at cell types and neurotransmitters
<Eliezer> but yeah, you could definitely use that to get at the laws of neural comp
<Eliezer> it would constrain future theories, at the very least
<OcsRazor> we want to start dissecting specific pieces some difficulties though
<OcsRazor> embryonic mouse brain is small!!!
<Eliezer> like finding a visual cell that responds to a horizontal bar doesn't mean this is a "horizontal bar detector", but it constrains your theory of what that visual cell does to allow the cell to respond to horizontal bars
<Eliezer> visual cortex researchers got bit on that a few times
<OcsRazor> each cell type definitely has a frequency and pattern it likes
<OcsRazor> that is going to be very important for our stuff
<Eliezer> but is "likes" necessarily the design principle used?
<Eliezer> the curve for "likes" is clearly going to play a very important role in neural wiring principles
<OcsRazor> seems to be...
<Eliezer> but it may not necessarily *be* that principle a la Hebb
*dreamingamoeba* the fonsecas are smarties :)
<OcsRazor> Hebbs rules are very good for general description...
<OcsRazor> but much more specific models are needed
<BJKlein> "Our group of over 40 researchers is busy advancing our understanding of fundamental neural processes such as motor control, learning, information processing, response to physical trauma, and the role of complex dynamics in the nervous system."
<MRAmes> Are you concerned that the cell types are not differentiated enough in the ebryonic tissue to make derterminations about specific neural types?
<OcsRazor> they differentiate in the dish...
<MRAmes> Or are you even focusing on specific types?
<OcsRazor> they are probably already genetically predestined
<MRAmes> Okay... are you able to identify the types in the dish?
<OcsRazor> Soon, next month in fact
<MRAmes> :)
<OcsRazor> That is one thing we are setting up the new scope for
<MRAmes> bleeding edge is right!
<OcsRazor> EL - what did you specifically about Hebbian cxns
<OcsRazor> mean specifically
<Eliezer> I mean that the hedonistic neuron model may be more complicated than that
<Eliezer> like... it started out that way, but then got more complicated; evolution elaborated on it
<Eliezer> everything evolution touches gets more complicated
<OcsRazor> it is pretty well known that specific neuron types will fire only if stimulated correctly
<OcsRazor> correct freq, pattern, etc
<Eliezer> with the upshot being that even if you have the response curve for what a neuron *most* likes, you may find that the neuron is getting a kind of input that it really likes, but not maximally likes, and moreover it's a special kind of input from the range of inputs it would really like that much
<MRAmes> cutoff E.
<Eliezer> it's evolution's infinite embroidery of everything that makes me worry that neural patterns may be opaque
<Eliezer> (no, it wasn't cutoff, it ended at "much")
<OcsRazor> exactly... we need a model of response curves for an entire network
<OcsRazor> with a variety of possible curves
<Eliezer> or the gene regulatory networks controlling where a neuron reaches out, and models of the response of the gene regulatory networks to changing conditions inside the cell as a result of incoming neural patterns...
<Eliezer> then you'd really have a direct causal model of how neurons wire themselves as a function of firing patterns inside the network
<Eliezer> the part I worry about is where that direct causal model turns out not to be interpretable as "this neuron wants to fire as much as possible"
<Eliezer> somewhere a hundred million years back in its evolutionary history it was that simple
<Eliezer> but now, an incremental adaptation away from that starting point makes it search for a particular kind of input which makes it fire well
<Eliezer> but which is not the particular kind of input that makes it fire *most*, if you're testing all possible inputs in a lab
<Eliezer> among other things, because the algorithm you're using in the lab to find "good inputs" hits on inputs that the neuron would never naturally encounter in vivo
<OcsRazor> we are trying to model a nautral stim pattern
<MRAmes> Thus my question about live snooping....
<OcsRazor> in order to get our networks to behave similarly to what has been seen in vivo
<OcsRazor> we definitely don't want them to fire like crazy
<MRAmes> How did you determine the stim patterns?
<OcsRazor> what we would like is more of an orderly behavior
<OcsRazor> the stim patterns are determined by looking at recording from real neurons
<MRAmes> Yes... how was that recording made, was my question.
<OcsRazor> to see what freq, pattern, etc they are outputting
<OcsRazor> single electrode recording
<MRAmes> Ah. Okay.
<OcsRazor> old style electrophysiology
<MRAmes> yes
<OcsRazor> glass electrodes
<OcsRazor> EL - we should talk in person
<OcsRazor> you are here in Atlanta right?
<Eliezer> yep
<Eliezer> Feb's >H ATL doesn't have a definite date yet, though
<MRAmes> The correllations between the recorded stim patterns and your resultant in-vitro patterns might be interesting reading.
<MRAmes> Are you far enough along to have interesting results yet?
<OcsRazor> Just shaking down the tech still I'm afraid
<OcsRazor> the cat and monkey data is getting good though
<MRAmes> Well... best of luck with it, Ocs.
<OcsRazor> we will probably surpass them thoug
<OcsRazor> b/c our system is completely transparent in multiple sneses
<OcsRazor> senses
<OcsRazor> we can see exactly what cells we are attached to... and directly observe their behavior
<OcsRazor> I'm hoping there is going to be alot of cross talk between all the labs invivo and invitro
<MRAmes> Sleep time for me. Good night all.
* Eliezer goes sleep
<OcsRazor> you guys have worn me out too
<MRAmes> :)
<OcsRazor> EL send me that link to the Atlanta group again
<Eliezer> http://www.posthuman...nshumanATL.html
<OcsRazor> Danke Shoen
<Eliezer> Danke Schoen?
<OcsRazor> thanks
<OcsRazor> (in german)
<Eliezer> ah, arigatou gozaimasu
<OcsRazor> Ich spreche nicht Jananishc
<OcsRazor> japanisch
<Eliezer> watashi wa germango o wakarimaisen
<Eliezer> (I think)
<Eliezer> oh, wait, I'm supposed to be sleeping
<Eliezer> nevermind
<OcsRazor> this is getting really confused
<OcsRazor> night 3liezer
<OcsRazor> any other questions before I leave?
<John_Ventureville> I'm back from dinner!
<John_Ventureville> *I think the chat must be over now*
<John_Ventureville> SeedAI combined with this technology could prove interesting...
<OcsRazor> we need good AI for analysis
<John_Ventureville> Do you have goals/deadlines set for your experiment?
<John_Ventureville> One year/two year/five year plan?
<OcsRazor> 4 years to PhD thesis
<John_Ventureville> Ahhh....
<OcsRazor> will include a general understanding of neural comp
<John_Ventureville> What are your plans once you become a newly minted PhD?
<OcsRazor> Post Doc, proff somewhere, tech transfer into startup co.
<John_Ventureville> So eventually you will be making "the big bucks" hopefully.
* BJKlein kicks the embers
<OcsRazor> I made the big bucks, was OK, intellectual stim is more impt
<John_Ventureville> But intellectual stimulation WITH great stock options can't be beat. : )
<John_Ventureville> lol
<OcsRazor> ah for the heady days of the bubble...
<John_Ventureville> I remember a Simpsons episode where they had the stock on printed on toilet paper rolls...
<John_Ventureville> *printed on*
<John_Ventureville> Is biotech currently seen as more "solid" than the other tech areas?
<BJKlein> OcsRazor, what are you doing in Germany?
<OcsRazor> Nothing is solid anymore
<John_Ventureville> too bad
<OcsRazor> But some biotech is making somewhat of a comeback
<John_Ventureville> good
<OcsRazor> my money would be on HGS
<BJKlein> HGSI yep
<OcsRazor> I'll be visiting friends in germany this summer
<John_Ventureville> What is HGSI?
<OcsRazor> Haseltine is a (not so) closet immortalist
<dreamingamoeba> biotech is too slow imo
<OcsRazor> John HGS is Human Genome Sciences
<John_Ventureville> ok
<OcsRazor> Bill Haseltine is chairman and CEO
<OcsRazor> brilliant business plan
<BJKlein> yes. not shy about physical immortality either
<John_Ventureville> How is his business plan brilliant?
<BJKlein> Saul Kent of LEF.org is another one
<OcsRazor> I know Saul personally
<BJKlein> cool
<John_Ventureville> really!
<OcsRazor> almost worked for him last summer
<John_Ventureville> very cool
<OcsRazor> at 21CM
<John_Ventureville> What kind of work would you have done?
<OcsRazor> they wanted someone permanent though and I was heading back to grad school
<OcsRazor> I would have worked on neuro cryo stuff
<John_Ventureville> I should have guessed that!
<John_Ventureville> lol
<OcsRazor> rabbit brain stasis
<OcsRazor> with Greg Fahy , who is another great guy
<BJKlein> OcsRazor, do you know where timeship will be built? by chance.
<OcsRazor> last I heard, Arizona desert, but that may have changed
<John_Ventureville> Over the Thanksgiving weekend I got to take part in a vitrification neurosuspension at Alcor, which for me was a fascinating experience being my first time.
<John_Ventureville> THE ARIZONA DESERT!!!
<OcsRazor> who performed the susp?
<BJKlein> when was the last you heard?
<OcsRazor> ~8 mnths
<John_Ventureville> Hugh Hixon and Dr. Jose were the main people.
<OcsRazor> I met hugh a few times
<John_Ventureville> The rumor I heard about the Timeship was that it would be built near Disney World.
<John_Ventureville> *so the masses could see and adore it*
<OcsRazor> they wanted a site for the timeship that was extremely geologically stable, not subject to flooding, etc
<OcsRazor> maybe they changed their minds
<OcsRazor> Florida would be perfect for marketing
<BJKlein> I sent them an email about Birmingham,
<John_Ventureville> I wish they would build it right across the highway from Ventureville!
<John_Ventureville> *we would seem awfully normal by comparison!*
<John_Ventureville> lol
<BJKlein> low taxes, good location, proximity to medical facilities, airport, non-nuclear target, etc
<OcsRazor> where is Ventureville?
<John_Ventureville> Mayer, Arizona
<John_Ventureville> one hour from Phoenix
<crw> how far from Tucson?
<John_Ventureville> I would have to consult a map
<crw> don't bother, i'll ask someone from Tucson. :)
<John_Ventureville> cool
<OcsRazor> Bruce, did Max and Natasha move to TX?
<BJKlein> not sure..
<John_Ventureville> Bruce, do you really think the southern/Bible belt mindset of the people in your city would accept the Timeship project??
<BJKlein> last I heard they were in CA
<John_Ventureville> I believe they did.
<John_Ventureville> I could be wrong.
<OcsRazor> Just before I left they had a going away party for them
<BJKlein> we're very hospitable
<John_Ventureville> I could imagine lots of picketers....
<BJKlein> I heard Natasha had cancer recently
<John_Ventureville> with signs reading "You can't avoid hell in a freezer!"
<OcsRazor> I'm sorry to hear that
<BJKlein> far as I know she's better
<OcsRazor> Good
<BJKlein> john, the old south is dead ;)
<OcsRazor> So why did you settle in AL BJ
<BJKlein> my family is from here
<BJKlein> I grew up in GA actually, south GA Americus
<BJKlein> but, as the song says, "My Home in in Alabama"
<OcsRazor> I've been around...
<BJKlein> you know Plains, GA Jimmy Carter hometown.. Americus is a few miles from there
<OcsRazor> Ft lauderdale, Cape Canaveral, Gainesville, Pittsburgh, LA and now atlanta
<OcsRazor> Driven by on the way south
<OcsRazor> family down in FL
<BJKlein> where do you call home
<BJKlein> ahh.. Ft Laud?
<OcsRazor> the Neurolab
<BJKlein> heh
<OcsRazor> ;^)
<BJKlein> the net
<BJKlein> cybertown, usa
<Guest> nick/John_Ventureville
<BJKlein> population 2 billion and growing
<OcsRazor> probably G'ville, FL feels like home to me
<OcsRazor> but family is all over
<OcsRazor> I need to re-establish my web presence
<John_Ventureville> Something on par with Eliezer or Anders Sandberg?
<BJKlein> http://www.imminst.o...ut/advisory.php
<OcsRazor> kuhl, thanks
<BJKlein> and now you'll see the "Volunteer's Forum" on the forum page
<OcsRazor> want to set up an info site for the Better than Original parts mentality
<BJKlein> Gracias
<BJKlein> if you need space let me know
<John_Ventureville> sounds good
<OcsRazor> I don't think anyone except Natasha has tried to really express this concept
<John_Ventureville> Anders has to a point
<OcsRazor> outdated though
<John_Ventureville> but Natasha has an artist's gift for doing such things
<BJKlein> others do dance around the idea.. but natasha does talke specifically about replacing bodyparts.. and the pics are compelling
<OcsRazor> It could be even more so, one of my roomates is an industrial designer
<John_Ventureville> I do wish her site was more technically oriented
<OcsRazor> I'm trying to get him to do engineering sketches of some of my concepts
<John_Ventureville> Your friend should contact her
<BJKlein> yes... i'd like to read her book though..
<BJKlein> about the history or transh
<John_Ventureville> *or work with you*
<OcsRazor> I know Natasha personally
<BJKlein> of*
<John_Ventureville> I have her book.
<BJKlein> nice
<John_Ventureville> She is a cool lady.
<John_Ventureville> I finally got to meet her at Extro-5.
<OcsRazor> she is very independent in the way she works though
<John_Ventureville> oh
<BJKlein> Max did a good job on CNN here lately(a few months ago) btw
<OcsRazor> don't know if she would work with someone else
<OcsRazor> full time?
<BJKlein> interview
<BJKlein> the Ted W. cryonics thing
<dreamingamoeba> i remember that
<OcsRazor> I saw it ;^)
<OcsRazor> he was great
<John_Ventureville> I never did see it!
<John_Ventureville> Is the video of it archived anywhere?
<BJKlein> probably our biggest media event thus far heh
<John_Ventureville> I heard he kicked butt!
<OcsRazor> the bio(d)ethicist they had on was a moron
<BJKlein> the transcript is
<BJKlein> Leon Kass?
<BJKlein> no..
<dreamingamoeba> i saw pinker on techtv once
<BJKlein> someone else.. yeh
<dreamingamoeba> that was fun
<John_Ventureville> I have already read that.
<dreamingamoeba> they reran the show today actually
<John_Ventureville> I want to see the body language.
<John_Ventureville> graggg!!!
<John_Ventureville> *and I missed it*
<OcsRazor> Who is Pinker?
<dreamingamoeba> steven pinker
<OcsRazor> the neurobiologist?
<dreamingamoeba> he's a cog sci researcher @ mit
<dreamingamoeba> nope
<dreamingamoeba> he's really big now
<BJKlein> i'll look around for the videa John_Ventureville.. don't know if it's aval
<dreamingamoeba> wrote 4 great books
<OcsRazor> yep thats him
<dreamingamoeba> oh
<dreamingamoeba> neurobiologist
<dreamingamoeba> heheh
<dreamingamoeba> he teaches psych mainly
<dreamingamoeba> John_Ventureville: theres a 2hr streaming movie of pinker on the web. it's very entertaining
<OcsRazor> Big picture guy, fun to read
<dreamingamoeba> introducing his latest book
<dreamingamoeba> yea
<OcsRazor> How the Mind Works is prob his best book
<dreamingamoeba> ya
<dreamingamoeba> you read it?
<OcsRazor> yep
<BJKlein> link?
<dreamingamoeba> nice!
<John_Ventureville> Bruce, Amoeba, thank you.
<BJKlein> Ok..Take Care John_Ventureville
<dreamingamoeba> http://web.mit.edu/m...ors/pinker.html
<OcsRazor> bye John
<dreamingamoeba> bye
<OcsRazor> my fun class I'm taking is...
<OcsRazor> Philosophical Issues in Computation
<dreamingamoeba> i like cognitive neuroscience the most
<dreamingamoeba> need to borrow gazzaniga's book from the library soon
<John_Ventureville> Am I going some where?
<John_Ventureville> lol
<dreamingamoeba> lol
<BJKlein> whoops
<OcsRazor> Cog is fun, but they need to make more of effort at technical achievement
<OcsRazor> starting to get good MRI data
<dreamingamoeba> more bottom up
<OcsRazor> we are hoping to act like a bridge b/w molecular and cog neuro
<dreamingamoeba> approaches*
<dreamingamoeba> good :)
<BJKlein> brb
<dreamingamoeba> John_Ventureville: you have realplayer correct?
<John_Ventureville> yes
<dreamingamoeba> :)
<John_Ventureville> did you find something?
<John_Ventureville> : )
<dreamingamoeba> i pasted the link
<dreamingamoeba> http://web.mit.edu/m...ors/pinker.html
<John_Ventureville> cool!
<OcsRazor> Does Pinker still have the goofy long hair?
<John_Ventureville> thank you
<dreamingamoeba> yes
<dreamingamoeba> lol
<OcsRazor> Its his trademark ;^)
<dreamingamoeba> no problem
<OcsRazor> He sticks out like a sore thumb at the annual neuro meeting
<dreamingamoeba> OcsRazor: is there a librarian assigned to your lab
<dreamingamoeba> research librarian
<dreamingamoeba> or something of the sort
<OcsRazor> we have a tech who logs all our stuff
<dreamingamoeba> ah
<OcsRazor> and keeps track of all the books and papers we use
<OcsRazor> he is building a database for web access right now
<dreamingamoeba> he's not a student?
<OcsRazor> applying to grad school I believe
<OcsRazor> full time tech
<dreamingamoeba> k
<OcsRazor> right now
<OcsRazor> I'm pretty toasty, been in the lab for 9 hours, I'm going to head home
<dreamingamoeba> ouch
<dreamingamoeba> take care
<OcsRazor> Night All

#2 caliban

  • Admin, Advisor, Director
  • 9,150 posts
  • 581
  • Location:UK

Posted 22 January 2003 - 11:44 PM

Very nice chat. Just when I thought about turning my back on these... did not even slide of the dreaded tangent to much.

Regarding the chat topic: We are having Professor Kevin Warwick over to give a talk in a few weeks time.
Anything that I should quiz him about on someone's behalf?


------------------------------------------

My sudden departure: Sorry, computer froze and I no time to log back in. - Apologies.

@Ocsrazor: As I already wrote to BJ, very glad to have you on board here!

#3 Bruce Klein

  • Topic Starter
  • Guardian Founder
  • 8,794 posts
  • 242
  • Location:United States

Posted 23 January 2003 - 12:54 AM

Caliban,

The credit goes to having an excellent source of info in Peter Possaro. He kept us all focused :)

If you get a chance:

Ask Kevin what is his ultimate goal of his work may be? longer life, immortality, etc?

#4 caliban

  • Admin, Advisor, Director
  • 9,150 posts
  • 581
  • Location:UK

Posted 23 January 2003 - 01:00 AM

Hmm I see the above link is not clickable.
copypaste if you are interessted:

http://www.cyber.rdg...index.htm?00057
or
http://www.cyber.rdg...e/K.Warwick.htm

#5 Bruce Klein

  • Topic Starter
  • Guardian Founder
  • 8,794 posts
  • 242
  • Location:United States

Posted 23 January 2003 - 01:32 AM

Sorry html is turned off in the chat archive. It has to do with a tech glitch.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users