Eliezer S. Yudkowsky writes:
> I get the impression that Billy Brown is talking about post-Singularity
> entities, while Eugene Leitl is talking about _Nanomedicine_. These are
First of all, I am talking about current technologies, applied to
people who will be dying now, and in the near future, and simple
biophysics of cryobiology, which are essentially constrained by
natural laws. At the patient-receiving end, I'm assuming
molecular-resolution quantitative dissassembly and computronium
(=ultimative computers made from matter, or a close enough liking to
it) plus optimal software to boot, so holding and processing that much
data is not a bottleneck. Time constraints are also not an issue,
since the system never leaves vitrified state during processing.
Alone these assumptions brings the whole discussion out of the realm
of science/technology into the technicolor LaLaLand, but I'm cool with
that.
> radically different levels of technology. By way of making the point
> (rather than suggesting technologies to bet your life on):
>
> 1) I've seen, oh, at least three different proposals for
> femtotechnology, which is to say machines built on the scale of nucleons
> rather than atoms. Given that kind of 'tech, it would be easy enough to
> read out the position of every single atom in the frozen brain.
This isn't the issue. You can achieve atomic resolution (well, that's
stretching it not too much, see Shao's NaCl picture) with cryo
AFM. This is today's technology:
http://www.people.Virginia.EDU/~js6s/zsfig/figureindex.html
I don't see any showstopping reasons why you can't combine imaging
with destructive disassembly of kg scale vitrified biological
objects. (Again, remember, we engage in absurd optimism here. This is
a gedanken, not a scientific discussion. Nano nano uber alles. You've
been warned).
> 2) The space of possible human minds is much smaller than the space of
> possible brains. Given enough intelligence, it would be possible to
> distinguish plausible reconstructions from implausible ones.
I will buy the Omega Point only when I've been personally introduced
to him. So far, total persona space is sufficiently large to render
total resurrection scenario intractable, even is Someone should be
benevolent enough to attempt it (I wouldn't, unless there wasn't
anything else worthwhile to do). 10^20 ops Aleph_zero ops are not.
> So yeah, under those circumstances, I think the grain size preserved by
> dropping your head in a bucket of liquid nitrogen contains all the
> relevant information about the brain, as I said before in greater
> detail. Whether you can get that information back out using
I am the only one that believes that information erasure in course of
natural system evolution is possible? Who guarantees you the perfectly
reversible Hamiltonian? To go to the mental wanking level (the usual
level of discussion when it comes to cryonics), do two quantum states
never decay to a single quantum state? So that when you reverse the
time you see a bifurcation, which is impossible to resolve without
resorting to other information? (Which has been meanwhile carried away
by photons to remote areas of space).
Not that these points are relevant in Real Life(tm), it's just a
gedanken showing that information erasure _can_ occur. I'm very
certain there is lots of information erasure going on in a lot of
natural processes, e.g. when you expose extremely anisotropic media to
extreme dehydration plus nonlinear compression at extremely high
pressures, in a fractal pattern (I need to quantify this, but
informally the ice structure obtained via freeze-subst looks like a
fractal to me).
Reynolds number is simply inapplicable here. This is not a macrosopic
homogenous system (that's what some people seem to have between their
ears), after all.
What I'm saying nobody knows what exactly is going on in there. At gut
level, it looks bad, to several people. Future research will tell us,
how bad. Echoing Merkle is exactly less than helpful, because I'd
rather be a live pessimist then a dead optimist. Does this really make
no sense?
> micron-scale robots following a blind reconstructive pattern is another issue.
>
> It's worth remembering that the design for a six-degrees-of-freedom
> manipulator specified in _Nanosystems_ would have been about 4 million
> atoms in size, while the actual atomic-scale specification worked out by
> Drexler seven years later contained only 2,596 atoms(*). So one might
> justifiably expect that the devices described in _Nanosystems_ or
> _Nanomedicine_ are the nanotechnological equivalent of vacuum tubes...
>
> (*) http://www.imm.org/Parts/Parts2.html
I am not talking about the limitations ot future technology here. I
thought I made that point very clear, but apparently I haven't.
This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:04:07 MDT