Hans Moravec's Speech

0 visninger
Gå til det første ulæste indlæg

Jason Asbahr

ulæst,
4. okt. 1992 04.40.1504.10.1992
til

For those who enjoyed reading Bruce Sterling's ALA speech, which
was kindly posted not that long ago by Tom Maddox, here's the text
of the speech given earlier by Hans Moravec. Enjoy!

Jason Asbahr
asb...@uh.edu

---

Hi, thanks for the interest. The following is the text I wrote
before the ALA talk: the talk was similar. My work at Thinking
Machines (and the years preceding and following it) should result
in practical robots (that can at least clean) this decade, if
my ever more specific fantasies work out. Also, you might be
interested to hear that a sequel to "Mind Children" called
"The Age of Mind" is in the works. It should be out in about
a year (Bantam), if I work a little harder than thus far.

-- Hans

________________________________________________________________

Pigs in Cyberspace

Hans Moravec
Robotics Institute
Carnegie Mellon University
Pittsburgh, PA 15213
May 1992

Exploration and colonization of the universe awaits, but
earth-adapted biological humans are ill-equipped to respond to the
challenge. Machines have gone farther and seen more, limited
though they presently are by insect-like behavioral inflexibility.
As they become smarter over the coming decades, space will be
theirs. Organizations of robots of ever increasing intelligence
and sensory and motor ability will expand and transform what they
occupy, working with matter, space and time. As they grow, a
smaller and smaller fraction of their territory will be undevel
oped frontier. Competitive success will depend more and more on
using already available matter and space in ever more refined and
useful forms. The process, analogous to the miniaturization that
makes today's computers a trillion times more powerful than the
mechanical calculators of the past, will gradually transform all
activity from grossly physical homesteading of raw nature, to min
imum-energy quantum transactions of computation. The final fron
tier will be urbanized, ultimately into an arena where every bit
of activity is a meaningful computation: the inhabited portion of
the universe will transformed into a cyberspace.

Because it will use resources more efficiently, a mature
cyberspace of the distant future will be effectively much bigger
than the present physical universe. While only an infinitesimal
fraction of existing matter and space is doing interesting work,
in a well developed cyberspace every bit will be part of a relevant
computation or storing a useful datum. Over time, more compact
and faster ways of using space and matter will be invented,
and used to restructure the cyberspace, effectively increasing the
amount of computational spacetime per unit of physical spacetime.

Computational speedups will affect the subjective experience
of entities in the cyberspace in a paradoxical way. At first
glimpse, there is no subjective effect, because everything, inside
and outside the individual, speeds up equally. But, more subtly,
speedup produces an expansion of the cyber universe, because, as
thought accelerates, more subjective time passes during the fixed
(probably lightspeed) physical transit time of a message between a
given pair of locations--so those fixed locations seem to grow farther
apart. Also, as information storage is made continually more
efficient through both denser utilization of matter and more efficient
encodings, there will be increasingly more cyber-stuff between any
two points. The effect may somewhat resemble the continuous-creation
process in the old steady-state theory of the physical universe of
Hoyle, Bondi and Gold, where hydrogen atoms appear just fast enough
throughout the expanding cosmos to maintain a constant density.

A quantum-mechanical entropy calculation by Bekenstein
suggests that the ultimate amount of information that can be stored
given the mass and volume of a hydrogen atom is about a megabyte.
But let's be conservative, and imagine that at some point in the
future only "conventional" physics is in play, but every few atoms
stores a useful bit. There are about 10^56 atoms in the solar system.
I estimate that a human brain-equivalent can be encoded in
less than 10^15 bits. If a body and surrounding environment takes
a thousand times more storage in addition, a human, with immediate
environment, might consume 10^18 bits. An AI with equivalent
intelligence could probably get by with less, since it does without
the body-simulation "life support" needed to keep a body-oriented
human mind sane. So a city of a million human-scale inhabitants
might be efficiently stored in 10^24 bits. If the atoms of the
solar system were cleverly rearranged so every 100 could represent
a bit, then a single solar system could hold 10^30 cities--far more
than the number (10^22) of stars in the visible universe! Multiply
that by 10^11 stars in a galaxy, and one gets 10^41 cities per
galaxy. The visible universe, with 10^11 galaxies, would then have
room for 10^51 cities--except that by the time intelligence has
expanded that far, more efficient ways of using spacetime and encoding
data would surely have been discovered, increasing the number
much further.

Mind without Body?

Start with the concepts of telepresence and virtual reality.
You wear a harness that, with optical, acoustical, mechanical and
chemical devices controls all that you sense, and measures all of
your actions. Its machinery presents pictures to your eyes,
sounds to your ears, pressures and temperatures to your skin,
forces to your muscles and even smells and tastes for the remain
ing senses. Telepresence results when the inputs and outputs of
this harness connect to a distant machine that looks like a humanoid
robot. The images from the robot's two camera eyes appear
on your "eyeglass" viewscreens, and you hear through its ears,
feel through its skin and smell through its chemical sensors.
When you move your head or body, the robot moves in exact synchrony.
When you reach for an object seen in the viewscreens, the
robot reaches for the object, and when it makes contact, your muscles
and skin feel the resulting weight, shape, texture and temperature.
For most practical purposes you inhabit the robot's body--your sense
of consciousness has migrated to the robot's location, in a true
"out of body" experience.

Virtual reality retains the harness, but replaces the remote
robot with a computer simulation of a body and its surroundings.
When connected to a virtual reality, the location you seem to in
habit does not exist in the usual physical sense, rather you are
in a kind of computer-generated dream. If the computer has access
to data from the outside world, the simulation may contain some
"real" items, for instance representations of other people connected
via their own harnesses, or even views of the outside world, perhaps
through simulated windows.

One might imagine a hybrid system where a virtual "central
station" is surrounded by portals that open on to views of multiple
real locations. While in the station one inhabits a simulated
body, but when one steps through a portal, the harness link is
seamlessly switched from the simulation to a telepresence robot
waiting at that location.

The technical challenges limit the availability, "fidelity"
and affordability of telepresence and virtual reality systems
today--in fact, they exist only in a few highly experimental
demonstrations. But progress is being made, and its possible to
anticipate a time, a few decades hence, when people spend more time in
remote and virtual realities than in their immediate surroundings,
just as today most of us spend more time in artificial indoor
surroundings than in the great outdoors. The remote bodies we will
inhabit can be stronger, faster and have better senses than our
"home" body. In fact, as our home body ages and weakens, we might
compensate by turning up some kind of "volume control."
Eventually, we might wish to bypass our atrophied muscles and
dimmed senses altogether, if neurobiology learns enough to connect
our sensory and motor nerves directly to electronic interfaces.
Then all the harness hardware could be discarded as obsolete,
along with our sense organs and muscles, and indeed most of our
body. There would be no "home" experiences to return to, but our
remote and virtual existences would be better than ever.

The picture is that we are now is a "brain in a vat,"
sustained by life-support machinery, and connected by wonderful
electronic links, at will, to a series of "rented" artificial bodies
at remote locations, or to simulated bodies in artificial realities.
But the brain is a biological machine not designed to function
forever, even in an optimal physical environment. As it begins
to malfunction, might we not choose to use the same advanced
neurological electronics that make possible our links to the
external world, to replace the gray matter as it begins to fail?
Bit by bit our brain is replaced by electronic equivalents, which
work at least as well, leaving our personality and thoughts clearer
than ever. Eventually everything has been replaced by manufactured
parts. No physical vestige of our original body or brain remains, but
our thoughts and awareness continue. We will call this process, and
other approaches with the same end result, the downloading of a
human mind into a machine. After downloading, our personality is
a pattern impressed on electronic hardware, and we may then find
ways to move our minds to other similar hardware, just as a computer
program and its data can be copied from processor to processor.
So not only can our sense of awareness shift from place to
place at the speed of communication, but the very components of
our minds may ride on the same data channels. We might find our
selves distributed over many locations, one piece of our mind
here, another piece there, and our sense of awareness at yet an
other place. Time becomes more flexible--when our mind resides in
very fast hardware, one second of real time may provide a subjective
year of thinking time, while a thousand years of real time
spent on a passive storage medium may seem like no time at all.
Can we then consider ourselves to be a mind without a body? Not
quite.

A human totally deprived of bodily senses does not do well.
After twelve hours in a sensory deprivation tank (where one floats
in a body-temperature saline solution that produces almost no skin
sensation, in total darkness and silence, with taste and smell and
the sensations of breathing minimized) a subject will begin to
hallucinate, as the mind, somewhat like a television tuned to a
nonexistent channel, turns up the amplification, desperately looking
for a signal, becoming ever less discriminating in the theories it
offers to make sense of the random sensory hiss it receives. Even
the most extreme telepresence and virtual reality scenarios we have
presented avoid complete bodylessness by always providing the mind
with a consistent sensory (and motor) image, obtained from an actual
remote robot body, or from a computer simulation. In those scenarios,
a person may sometimes exist without a physical body, but never
without the illusion of having one.

But in our computers there are already many entities that
resemble truly bodiless minds. A typical computer chess program
knows nothing about physical chess pieces or chessboards, or about
the staring eyes of its opponent or the bright lights of a tournament.
Nor does it work with an internal simulation of those physical
attributes. It reasons instead with a very efficient and
compact mathematical representation of chess positions and moves.
For the benefit of human players this internal representation is
sometimes translated to a recognizable graphic on a computer
screen, but such images mean nothing to the program that actually
chooses the chess moves. For all practical purposes, the chess
program's thoughts and sensations--its consciousness--is pure chess,
with no taint of the physical, or any other, world. Much more
than a human mind with a simulated body stored in a computer, a
chess program is a mind without a body.

So now, imagine a future world where programs that do chess,
mathematics, physics, engineering, art, business or whatever, have
grown up to become at least as clever as the human mind. Imagine
also the most of the inhabited universe has been converted to a
computer network--a cyberspace--where such programs live, side by
side with downloaded human minds and accompanying simulated human
bodies. Suppose that all these entities make their living in
something of a free market way, trading the products of their
labor for the essentials of life--in this world memory space and
computing cycles. Some entities do the equivalent of manual work,
converting undeveloped parts of the universe into cyberspace, or
improving the performance of existing patches, thus creating new
wealth. Others work on physics or engineering problems whose so
lutions give the developers new and better ways to construct
computing capacity. Some create programs that can become part of
one's mental capacity. They trade their discoveries and inventions
for more working space and time. There are entities that specialize
as agents, collecting commissions in return for locating
opportunities and negotiating deals for their clients. Others
act as banks, storing and redistributing resources, buying and
selling computing space, time and information. Some we might
class as artists, creating structures that don't obviously result
in physical resources, but which, for idiosyncratic reasons, are
deemed valuable by some customers, and are traded at prices that
fluctuate for subjective reasons. Some entities in the cyberworld
will fail to produce enough value to support their requirements
for existence--these eventually shrink and disappear, or merge with
other ventures. Others will succeed and grow. The closest present
day parallel is the growth, evolution, fragmentation and consolidation
of corporations, whose options are shaped primarily by their economic
performance.

A human would likely fare poorly in such a cyberspace.
Unlike the streamlined artificial intelligences that zip about,
making discoveries and deals, reconfiguring themselves to efficiently
handle the data that constitutes their interactions, a
human mind would lumber about in a massively inappropriate body
simulation, analogous to someone in a deep diving suit plodding
along among a troupe of acrobatic dolphins. Every interaction
with the data world would first have to be analogized as some
recognizable quasi-physical entity: other programs might be presented
as animals, plants or demons, data items as books or treasure
chests, accounting entries as coins or gold. Maintaining such
fictions increases the cost of doing business, as does operating
the mind machinery that reduces the physical simulations into mental
abstractions in the downloaded human mind. Though a few humans
may find a niche exploiting their baroque construction to
produce human-flavored art, more may feel a great economic
incentive to streamline their interface to the cyberspace.

The streamlining could begin with the elimination of the
body-simulation along with the portions of the downloaded mind
dedicated to interpreting sense-data. These would be and replaced
with simpler integrated programs that produced approximately
the same net effect in one's consciousness. One would
still view the cyber world in terms of location, color, smell,
faces, and so on, but only those details we actually notice would
be represented. We would still be at a disadvantage compared with
the true artificial intelligences, who interact with the cyberspace
in ways optimized for their tasks. We might then be
tempted to replace some of our innermost mental processes with
more cyberspace-appropriate programs purchased from the AIs, and
so, bit by bit, transform ourselves into something much like them.
Ultimately our thinking procedures could be totally liberated from
any traces of our original body, indeed of any body. But the
bodiless mind that results, wonderful though it may be in its clarity
of thought and breadth of understanding, could in no sense be
considered any longer human.

So, one way or another, the immensities of cyberspace will be
teeming with very unhuman disembodied superminds, engaged in affairs
of the future that are to human concerns as ours are to
those of bacteria. But, once in a long while, humans do think of
bacteria, even particular individual bacteria seen in particular
microscopes. Similarly, a cyberbeing may occasionally bring to
mind a human event of the distant past. If a sufficiently powerful
mind makes a sufficiently large effort, such recall could
occur with great detail--call it high fidelity. With enough fidelity,
the situation of a remembered person, along with all the
minutiae of her body, her thoughts, and feelings would be perfectly
recreated in a kind of mental simulation: a cyberspace
within a cyberspace where the person would be as alive as
anywhere. Sometimes the recall might be historically accurate, in
other circumstances it could be artistically enhanced: it depends
on the purposes of the cybermind. An evolving cyberspace becomes
effectively ever more capacious and long lasting, and so can support
ever more minds of ever greater power. If these minds spend
only an infinitesimal fraction of their energy contemplating the
human past, their sheer power should ensure that eventually our
entire history is replayed many times in many places, and in many
variations. The very moment we are now experiencing may actually
be (almost certainly is) such a distributed mental event, and most
likely is a complete fabrication that never happened physically.
Alas, there is no way to sort it out from our perspective: we can
only wallow in the scenery.

--
Jason Asbahr 116 E. Edgebrook #603
asb...@uh.edu Houston, Texas 77034
ne...@tree.egr.uh.edu (NeXTmail) (713) 743-6995 voice
asb...@tree.egr.uh.edu (NeXTmail) UH NeXT Campus Consultant

Svar alle
Svar forfatteren
Videresend
0 nye indlæg