[ExI] destructive uploading was AI class
Eugen Leitl
eugen at leitl.org
Fri Sep 2 11:59:02 UTC 2011
On Thu, Sep 01, 2011 at 05:30:58PM -0700, Jeff Davis wrote:
> On Wed, Aug 31, 2011 at 1:57 AM, Eugen Leitl <eugen at leitl.org> wrote:
> > On Tue, Aug 30, 2011 at 04:47:30PM -0700, Jeff Davis wrote:
> >> Gene et al,
> >>
> >> It's my impression that the whole "destructive scan" meme is an
> >> artifact of the microtome-sliced sample-preparation method of electron
> >> microscopy. I can't imagine that this anti-diluvian
> >
> > Not just electron microscopy, there's no way to nondestructively
> > image a ~l sized chunk of glass. Fundamental physical limits
> > do not care about the technology vogue.
>
> Assuming that the "~l sized chunk of glass" above means "approximately
> liter sized chunk of glass", and refers to an "approximately liter
> sized chunk of vitrified (the "glass" in question) flesh (in
You assume correctly.
> particular, brain flesh), I respond with either, "Huh? So what?" or
I've had this conversation multiple times around 1995.
> "C'mon, acknowledge the given parameters." This assertion fails the
> test of relevance on two points: First, the fact that "there's no
> way", really is just a case of there's no way ***NOW***, which in
> discussions of future tech -- even near future tech -- is ALWAYS the
Do you know why the information-theoretic death is being called that?
> case, ...and trivial. Second, what's with the "liter sized" business?
> Too big to scan as a single unit? Won't be scanned as a single unit,
Yes, it won't be scanned as a single unit, precisely.
> and doesn't need to be. The flesh in question is comprehensively
> interpenetrated by a capillary system which reduces the bulk size of
> any scanning target to cubic micro- or nanometers, with scanning bots
The nanoagents in question are some um^3, and your numbers are
unfortunately wrong.
Rather think 100 um^3, and remember that the feature size is
at least 1 nm, and in some case you could need sampling down
to e.g. degree of phosphorylation of a biomolecule.
> surrounding the subject scanning volume from all angles. In this
Scanning using which wavelenghts or forms of radiation? Using
which sources, resulting in which damage, requiring which
signal acqusition times?
Instead of armwaving, do read up on the basic science, and then
come back with actual calculations.
> situation, with multi-spectral transmission microscopy there should be
> no limit (sub-atomic if you want to get fussy) to the resolution or
> data completeness/redundancy.
BULLSHIT.
> And regarding the gratuitous "fundamental physical limits", true
> enough, but then how often have we heard that what was yesterday
> considered impossible because of such "limits" is today found to be
> "doable" through some unanticipated "new" physics or artful
> workaround?
Fair enough, I pegged you for a religious person. I'm sorry I'll have
to bow out of a faith-based discussion. Life's too short as is.
> <snip>
>
> >> Please explain why it would be infeasible to clear the lumen of the
> >> circulatory system of the vitrified patient, replace it with LN2, and
> >
> > Clear = removing material. So it is fundamentally destructive,
>
> If you were a hack lawyer, I might say "good work". By the standards
> of good-faith discussion, however, it stinks. There's nothing the
> least bit destructive about clearing the circulatory lumen of
> solidified cryoprotectant.
You're trying to weasel yourself out by redefining destructive.
Removing material from your brain *is* destructive, my friend.
Using ionising radition resulting in cumulative structure degradation
*is* destructive, my friend.
You're trying to sell me an a dental job via anal access.
> > though one might quibble
> > it doesn't remove relevant structural information.
>
> The crux of the matter is hardly a "quibble". What's with you, Gene?
> You're waaaaay better than this. Just admit that you were wrong and
> that intra-luminal, trans-luminal non-destructive scan, with the
No such thing. Read up on the basic physics, and do the math.
> subject remaining at vitrification temp, is a straightforward, trivial
> solution to the scanning requirement.
Doesn't exist.
> <snip>
>
> > You're in the lumen of a cleared vessel, with some 100 um vitrified
> > tissue between you and the other one. Which methods will you be
> > using to nondestructively scan these 100 um deep volume with
> > required 1 nm or better resolution, without degrading intermediate layers?
>
> Choose your method. Extracellular (the bots remain in the circulatory
WHICH method?
> lumen) Acoustic, Electromagnetic, or any other interactive
> transmission below destructive energy levels; or perhaps the bots
You have really no fucking idea what you're talking about, and it shows.
> exit the lumen, clear more intracellular cryoprotectant to provide
Which part of 'it's bulk vitrified tissue' you don't understand?
Which way are you going to image your cubic microns worth of stuff
without removing the stuff between what's between you and what
you're trying to image? Do you understand the words coming out
of my mouth?
> intracellular pathways, and scans with contact microscopy (STM).
Exactly, iterative ablation and proximal-probe. There is really no
other way, and if you don't like that for cosmetic reasons you can
stuff it.
> Even without contact microscopy I don't see resolution as a problem.
Show me the numbers. Signal acquisition times of years, not gigayears.
What is the wavelenght of your radition, and what it the accumulated
dosis. Numbers, or STFU.
> Recent advances have shown that wavelength-limited resolution is not
> operative at close range.
>
> > How long are you going to take?
>
> Can't really say. The scanning process will be limited by the rate at
> which the evolved heat can be dissipated. The heat generated by the
Haha, right.
> scanning process must be dissipated so as to prevent premature
> rewarming, but it seems to me that the liquid nitrogen environment and
> the circulatory system provide an effective waste heat transport
> system.
>
> > And why are you doing such ineffective
>
> Sez you.
>
> > contortions if all you
> > have to do is slice up the volume into sufficiently small
> > sections, and use known methods like interative ablation of
> > surface imaging, which allows you to speedly image at up to
> > atomic resolution *without* degrading underlying layers?
>
> Clearly, I am dealing with this in a cryonics patient context. (If
Of course.
> one were dealing with a living patient, who wanted to be uploaded but
> didn't want the biological self to be killed in the process, then the
> nanobots would be injected, conduct their information-gathering -- ie
> scan -- in the background, and then exit the subject silently,
> seamlessly.) As a cryonics patient living in a pre-upload society,
You're again failing to do the math, and there's no exiting, because
then you're diverging, unless you have shut down the biology and
never intend to resume it. Which is, you get it, pointless.
Once again to introduce a dose of reality into the cute polyannaland
you inhabit, assume cubic micron agent and micron grid spacing for
pickup. Whoops, that's a lot of volume, innit? Nevermind the infrastructure
to power it and to cool it. Do you think you're ambulatory?
Gundam suit ambulatory, maybe.
> where uploading is not yet a fact, but the concept is established,
As building numerical models doesn't require medical nanotechnolgoy
there are no revived cryonics patients unless there are numercal
animal models. In fact, it is far more likely that the only way
how cryonics patients will be recondstructed is as numerical models.
And be it because the old-fashion meat puppet humans no longer exist.
> with implementation considered possible, even likely, I have to
> anticipate the upload option and provide instructions to those caring
> for me in stasis, as to under what circumstances I might wish to be
> uploaded. Currently, those instructions are: "Upload me when
> feasible, but the biological self must not be damaged(ie no
> destructive scan).
I suggest you reformulate that as destructive scan meaning
*relevant information destruction*. Orelse, assuming your
wishes are binding, you're delaying your resurrection, possibly
indefinitely.
> > You could do dental work via anal access in principle, but
> > you'll probably agree it is more difficult that way.
>
> If one doesn't have one's head up one's ass. ;-}
> >
> >> from the inside, "infiltrating the brain" as Keith describes.
> >
> > Because latter assumes a living system,
> > instrumented with a
> > rather large volume of nanoagents, which will cause volume
> > inflation and need to be done slowly so the biology can deal
> > with it.
>
> By latter, do you mean Keith, or "Infiltrating the brain". In any
The idea of in-vivo gradual uploading is old, and has been countlessly
reinvented, so attaching particular names to it is pointless ego-stroking.
> case, you have some assumptions here -- "large volume" and "slowly" --
> which may or may not be too large or too slow. In any event, without
> more detail, I can neither agree nor disagree. That said, I'm working
> the cryopatient upload/revivification issue.
You should produce a written document I hope, which we can then
tear apart (he cackled mainiacally, rubbing his hands).
> >
> > Once you're vitrified, you do not want to go above devitrification
> > temperature, as it would introduce the damage you worked so hard
> > to avoid,
>
> Trivial. Adjust the stasis temp to accommodate waste heat removal
Of course it's trivial, I'm saying it that you have to work at -130 C or
below (depends on Tg).
> rate. Melting point of LN2 is 63.15 deg K. That gives you 14 degrees
> K to work with.
>
>
>
> > and you're dealing with a brittle glass object.
>
>
> > You can
> > of course introduce a fractal heat exchange infrastructure and
> > flash-devitrify it, but then you've made your job much harder,
> > as you will have to remove cryoprotectants *and* keep the
> > tissue from unraveling as it no longer has homeostasis working
> > for you. In other words, as soon as you no longer process the
> > tissue at cryogenic temperatures you a) introduce irreversible
> > information destruction *without any need* b) make time critically
> > relevant, again *without any need*.
>
> This comment suggests that, perhaps due to some disconnect, we're not
> working the same problem. All my proposed scanning would be done with
> the patient fully and continuously vitrified.
Excellent. We can at least agree on something.
> >
> > This is obviously stupid, and this will not be done as long
> > as there are other, much easier ways. See the anal dental.
> >
> >> Then Gene, you make the comment below:
> >>
> >> " ...there isn't any medical nantechnology, nor will there be any for
> >> anyone reading this message right now."
> >>
> >> We don't have the nanotech at the moment. Well duh, but more to the
> >
> > This means your only option is vitrification, or maybe fixation and
> > embedding.
>
> As in cryo-suspended.
>
> >
> >> point, so what? Scanning-for-upload will be the "back end" of the
> >> freeze-wait-reanimate process, so the fact that we don't have the
> >> reanimation tech -- scanning included -- right now is both well
> >> established(in our circles) and trivially irrelevant. Anyone "reading
> >
> > If it was trivially irrelevant I wouldn't have to write these
> > missives every year, or so. Not that these do any good.
>
> Or, if you accepted that it is in fact trivially irrelevant, then you
> could discard these missives and get on with pursuits more
> constructive than serial nay-saying. But I still think you're a
> stalwart fellow.
>
> >> this message right now", if vitrified (cryonically preserved) will
> >> be,...well... preserved. They will remain essentially unchanged,
> >> essentially indefinitely, and consequently, will be there waiting when
> >> the tech arrives.
> >
> > Anyone who hears uploading implies she will be suspended while alive
> > by magical nano, using incremental in vivo scanning/substitution.
>
> Suspended by magical nano? What does that mean? Why not suspension
Walk-in-uploading-while-u-wait-and-drink-coffee. In maybe 30 years. Yes, people are really
silly.
> by conventional vitrification. And how -- or perhaps "why" --
> should "in vivo scanning/substitution" be a part of the suspension
> process? "Un-suspended" with nano -- magical(Clarkian) or otherwise
> -- I could understand, but suspension?
Not suspended, uploaded.
> >
> > This ain't going to happen. Unless you've bought into the Singularity
> > cult, world without end, amen. At which point you've become a part
> > of the problem.
>
> Notions of a "singularity" -- and uploading -- are thoroughly
> entertaining and entirely speculative. I wait with pleasant
It's good that you don't expect the nice people/AI or the
future, yet unborn to save our asses. Unfortunately it's us
that has to save our asses, and we're so far failing spectacularly.
> anticipation for the next "episode".
>
> >> But I don't have to explain this to you, Gene. You're fully versed in
> >> this business -- theory and practice -- having worked with 21st
> >> Century Medicine maybe a decade ago helping to develop ice-blockers
> >> for the vitrification process. Correct me if I'm wrong.
> >>
> >> So I'm befuddled. (Scratches head in befuddlement.) Why are you,
> >> overqualified as you are in this business, making negative,
> >> misleading, and coyly irrelevant comments? Still pissy about the
> >
> > If you have technical arguments why you think above does not apply,
> > I'll be very happy to hear them.
> >
> >> knuckle-dragging atavism and luddite obstructionism of your fellow
> >> humans? Can't help you there. Well,...actually...
> >
> > Don't hint at it, spill the beans!
>
> Offlist and privately, only. I've hinted at this several times on
Sure, I'll even sign the NDA.
> the list, to see if anyone would notice. No one has shown any
> interest. I know how to make the future you so fiercely crave, emerge
> at a rapidly accelerated pace, and I'm willing to tell you about it,
> with a strict promise of confidentiality. It's been my experience
Granted.
> however, that people generally dismiss ideas that aren't their own.
> And of course, I'm a crackpot.
Welcome to the club. It's part of the job description, here.
> Best, Jeff Davis
>
> "Everything's hard till you know how to do it."
> Ray Charles
>
>
> >>
> >> On Tue, Aug 30, 2011 at 9:00 AM, Eugen Leitl <eugen at leitl.org> wrote:
> >> > On Tue, Aug 30, 2011 at 06:16:57AM -0700, Keith Henson wrote:
> >> >> On Mon, Aug 29, 2011 at 11:45 PM, Adrian Tymes <atymes at gmail.com> wrote:
> >> >>
> >> >> snip
> >> >>
> >> >> > If the uploading process is destructive and one-way,
> >> >>
> >> >> Can you think of any reason uploading would need to be destructive and one-way?
> >> >
> >> > Anyone reading this will only be uploaded as a nice chunk of
> >> > vitrified tissue
> >>
> >>
> >> > which will be presectioned and destructively
> >> > scanned in a massively parallel fashion, abstracting nonrelevant
> >> > data along the way. It is destructive, but not necessarily
> >> > one-way, though it is effectively one-way as nobody will bother
> >> > fabricating slow and expensive meat puppets.
> >> >
> >> >> I know Hans Moravec proposed that decades ago and it became a
> >> >> pervasive meme, but it's an awful concept and a serious marketing
> >> >> burden for transhumanism.
> >> >
> >> > Doesn't affect cryonics, as this side of suspension process is
> >> > resurrection-agnostic.
> >> >
> >> >> I have reasons to think it would be a lot harder than infiltrating the
> >> >> brain with nanotech monitoring posts and learning how to emulate it.
> >> >
> >> > It would be a lot harder since there isn't any medical nantechnology,
> >> > nor will there be any for anyone reading this message right now.
> >> >
> >> > --
> >> > Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
> >> > ______________________________________________________________
> >> > ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
> >> > 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
> >> > _______________________________________________
> >> > extropy-chat mailing list
> >> > extropy-chat at lists.extropy.org
> >> > http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
> >> >
> >>
> >> _______________________________________________
> >> extropy-chat mailing list
> >> extropy-chat at lists.extropy.org
> >> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
> > --
> > Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
> > ______________________________________________________________
> > ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
> > 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
> > _______________________________________________
> > extropy-chat mailing list
> > extropy-chat at lists.extropy.org
> > http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
> >
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
--
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
More information about the extropy-chat
mailing list