From: Mitchell Porter (mitch@thehub.com.au)
Date: Wed Nov 19 1997 - 03:36:34 MST
This is my latest attempt to envision the future.
A copy will live at
http://www.thehub.com.au/~mitch/future/nov97-scenario.html
1) The "assembler breakthrough".
I recently posted to sci.nanotech that I expect the first
assemblers within five years. I would now add to that, that
I think most of the foreseeable uses of nanotech (that don't
require astronomical quantities of matter) will be possible
within two years of the fabrication of assemblers. My reasons:
i) Once we have assemblers, we can build new designs very quickly -
thus accelerating the construction stage in the design-and-test
cycle. Furthermore, those who are trying to design, say, a
new type of nanosystem will be able to use the Internet of five
years' time as a collaborative medium. I doubt that we can imagine
very clearly just what possibilities will exist, but we have
every reason to believe they will be awesome. Think of shared
3D virtual design-spaces, and realtime webcrawling agents making
spontaneous comments.
ii) Once we have assemblers, new sorts of intracellular probes
should be possible; our knowledge of cellular workings should
quickly become just about complete. In particular, this will
be true of neurons.
One consequence: our ability to model biological systems will
be improved, and therefore our ability to extrapolate the
consequences of proposed interventions (e.g. therapies and
augmentations). Even something with the complexity of the
fabled "cell-repair machine" might become feasible very quickly.
2) The noetics revolution.
"Noetics" is my term for whatever is going to emerge from the
current interest in "consciousness studies". I am skeptical
that current cognitive models of consciousness are adequate,
and that a description of the brain in which the finest grain
is "neuron firing" or "neuron not firing" will *ever* do
justice to the phenomenon. If the current interest in "quantum
mind" theories fares any better, it would not surprise me if
it does so in conjunction with new interpretations of quantum
theory which impute more structure to the world than the local
spatial relations which glue together the space-time manifold
of pre-quantum physics. I imagine a mature theory of noetics
to resemble a combination of the most abstract mathematics
and philosophical phenomenology (with its odd analyses of
the structure of conscious experience), grounded in a postquantum
physics equipped with concepts like David Finkelstein's "qets"
("quantum sets").
If there is anything to these ideas, one should expect them
to bear fruit very soon after the arrival of assemblers, if
not before, since (as argued above) newly detailed knowledge
of the brain and its neurons should definitely tell us
whether there's quantum coherence happening in there, and
equally, the burgeoning development of quantum technology
(quantum computation, quantum cryptography, quantum teleportation)
should help us sort through the various interpretations
and postquantum theories.
If the truth turns out to be not so far removed from what
we already understand - i.e. information processing by
locally deterministic physical systems - we can still expect
that our theories of mind will have a new robustness and
conviction, thanks to the new biological data.
3) Mind acceleration.
Fast minds whose subjective time passes at a thousand or a
million times human speed are generally anticipated to be
a consequence of mind uploading. As a skeptic regarding
the completeness of functionalist theories of mind, I'm a
skeptic regarding our current ideas of mind uploading.
But this doesn't mean that such schemes can't work at all.
Suppose (for example) that the "mind" - the "thing that
is conscious" - consists of a set of quantum-entangled
electrons, spread across millions of neurons which belong
to a particular phenotype optimized for quantum coherence,
and implementing some sort of quantum computer through
their superposed spin-flips. I see no reason why you
couldn't replace all the other neurons with nanocomputers,
in traditional Moravecian fashion, and then accelerate
the spin-flipping processes to the same degree, through
more fine-grained interventions at the quantum-chemical
level. This would be an incredible achievement by
contemporary standards, but it's not inconceivable.
So even if mind uploading doesn't work, mind acceleration
will still be a thinkable goal. I therefore predict that
that a "mind-acceleration underground" will emerge after
the assembler revolution, availing itself of the new
technology - and in all likelihood it *will* be an
underground. If governments ban psychedelics and human
cloning, what will they think of a technology meant to
accelerate mental processes by orders of magnitude?
I do have to mention what seems to me the main competition
in the race towards fast minds, and that is the "de novo"
fast mind - nonhuman, and produced by design and/or
evolutionary engineering. Again, even if I'm right that
consciousness is not just neural information-processing,
there's no reason why artificial consciousness can't be
created in artificial systems with the right characteristics.
It may well be easier to start from scratch, than to
transform the human brain piecemeal into a faster medium.
One more comment. For those who always use as their mental
medium the fastest available, there may never be any
Singularity. Superintelligences - "Powers" - will be extremely
complex entities. Even if we had "wish machines" which could
grant any wish that was formulated with sufficient exactness,
I think we'd have trouble getting them to produce a
superintelligence, since our notions are not developed
enough to frame the wish precisely. If this is true, then
genuine intelligence increase (as opposed to simple
increase of basic speed) will be a slow affair, in the
subjective time of the entities at work on the problem.
The Singularity, according to this conception, will only
ever happen to those entities which miss out on mind
acceleration, and who consequently experience the changes
in subjectively brief time.
4) Picotechnology and space-time engineering
The possibility of "picotechnology" (technology on nuclear
and smaller scales) is sometimes rejected on the grounds
that, at least outside a neutron star, all you can do with
nucleons is to make the existing spectrum of nuclei. But I
don't think we can safely assume that if anything else
is possible, we'll have to go to space to obtain the
necessary conditions. Look at recent speculations on
"Q-balls", the hypothesized supersymmetric counterparts
to nucleons. Unlike quarks, "superquarks" are spin-zero
bosons and can form arbitrarily large aggregates. The Q-ball
theorists suggest that "the laws of physics" would be
different in their interior; more specifically, that
particle masses, lifetimes and interactions would be
different there. Or consider Bose-Einstein condensation.
Here is a way to get atoms to pile up on top of each other.
Could one use this to obtain densities approaching those
of exotic states of matter like neutronium? Finally,
consider that theorists of the inflationary universe
(Alan Guth and Andrei Linde) have said that in principle
one would need only a few kilograms of mass-energy to
create a "basement universe". (No doubt they would argue
that one would need a lot more, in order to concentrate
those kilograms of mass-energy appropriately. I can't
help but wonder whether the necessary apparatus really
could fit in a basement.)
There are any number of reasons for a fast-mind society
to explore such possibilities. One might hope to gain
space, energy, or computation from endlessly propagating
chains or networks of space-time bubbles, modelled
on the cosmology of self-reproducing universes that
has obtained attention recently. One might hope to
accelerate oneself to even greater speeds, perhaps as
far as one unit of "Planck time" per tick of the
subjective clock (that would be a clock speed of about
10^43 Hz). One might hope to make wormholes, link up
"known space" with them, and send them out into the
universe (or into newly fashioned universes) as
probes.
A notion that intrigues me: as subjective time speeds up,
subjective space expands. For a mind that is faster by
a factor of a thousand, Earth is a thousand times bigger,
in that it takes a thousand times longer to circumnavigate.
There might well be a practical argument for the creation
of wormholes here - to prevent the disintegration of
the network of accelerated minds from "subjective expansion".
Another option which may be significant in an era of
space-time engineering is out-and-out Separation:
migration into a designer universe, followed by the
severing of all ties with the parent universe.
5. "Post-spatial" existence
What are the limits? String theory and its successor
M theory already describe conditions under which the effective
dimensionality of spacetime changes, along with the inventory
of fields and their interactions. Can we go further and
imagine "realms" which are not spaces at all? Think of the
variety of form possible amongst sets, or data structures.
What sort of structure could we choose to give to the world
we inhabit - or to the "phenomenal world" which we perceive?
Let me make more concrete the thinking behind this idea.
What if it were the case that spatial adjacency and
quantum entanglement were two instances of the same sort
of thing, namely an ultimate elementary relation of
"connectedness"? One might fashion something resembling a
wormhole by engineering an appropriate entanglement of
two physical systems; I gather that such a setup is a
necessary precursor of quantum teleportation.
Once one starts thinking of the physical world as a sort
of web of connected space-time points, it's a short step
to imagining webs without the gridlike structure of
space. But I don't know how to take this idea further
at present.
-mitch
http://www.thehub.com.au/~mitch
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:45:08 MST