Rubin,

[ Pobierz całość w formacie PDF ]
Artificial Intelligence and Human Nature
Charles T. Rubin
What awaits is not oblivion but rather a future which, from our
present vantage point, is best described by the words “postbiological” or
even “supernatural.” It is a world in which the human race has been
swept away by a tide of cultural change, usurped by its own artificial
progeny.
–Hans Moravec,
Mind Children
We are dreaming a strange, waking dream; an inevitably brief
interlude sandwiched between the long age of low-tech humanity on the
one hand, and the age of human beings transcended on the other … We
will find our niche on Earth crowded out by a better and more compet-
itive organism. Yet this is not the end of humanity, only its physical
existence as a biological life form.
–Gregory Paul and Earl D. Cox,
Bey
ond
Humanity
T
he cutting edge of modern science and technology has moved, in its aim,
beyond the relief of man’s estate to the elimination of human beings. Such fan-
tasies of leaving behind the miseries of human life are of course not new; they
have taken many different forms in both ancient and modern times. The chance
of their success, in the hands of the new scientists, is anyone’s guess. The most
familiar form of this vision in our times is genetic engineering: specifically, the
prospect of designing better human beings by improving their biological sys-
tems. But even more dramatic are the proposals of a small, serious, and accom-
plished group of toilers in the fields of artificial intelligence and robotics. Their
goal, simply put, is a new age of post-biological life, a world of intelligence with-
out bodies, immortal identity without the limitations of disease, death, and unful-
filled desire. Most remarkable is not their prediction that the end of humanity
is coming but their wholehearted advocacy of that result. If we can understand
why this fate is presented as both
necessary
and
desirable
, we might understand
something of the confused state of thinking about human life at the dawn of this
new century—and perhaps especially the ways in which modern science has shut
itself off from serious reflection about the good life and good society.
Charles T. Rubin
is a professor of political science at Duquesne University. An earlier version
of this essay was presented at “The Ethical Dimensions of Biotechnology,” a conference organized
by the Henry Salvatori Center for the Study of Individual Freedom in the Modern World,
Claremont McKenna College.
88 ~ T
HE
N
EW
A
TLANTIS
Copyright 2003. All rights reserved. See
for more information.
The Road to Extinction
T
he story of how human beings will be replaced by intelligent machines goes
something like this: As a long-term trend beginning with the Big Bang, the evo-
lution of organized systems, of which animal life and human intelligence are rel-
atively recent examples, increases in speed over time. Similarly, as a long-term
trend beginning with the first mechanical calculators, the evolution of comput-
ing capacity increases in speed over time and decreases in cost. From biological
evolution has sprung the human brain, an electro-chemical machine with a great
but finite number of complex neuron connections, the product of which we call
mind or consciousness. As an electro-chemical machine, the brain obeys the laws
of physics; all of its functions can be understood and duplicated. And since com-
puters already operate at far faster speeds than the brain, they soon will rival or
surpass the brain in their capacity to store and process information. When that
happens, the computer will, at the very least, be capable of responding to stim-
uli in ways that are indistinguishable from human responses. At that point, we
would be justified in calling the machine intelligent; we would have the same evi-
dence to call it conscious that we now have when giving such a label to any con-
sciousness other than our own.
At the same time, the study of the human brain will allow us to duplicate its
functions in machine circuitry. Advances in brain imaging will allow us to “map
out” brain functions synapse by synapse, allowing individual minds to be dupli-
cated in some combination of hardware and software. The result, once again,
would be intelligent machines.
If this story is correct, then human extinction will result from some combi-
nation of transforming ourselves voluntarily into machines and losing out in the
evolutionary competition with machines. Some humans may survive in zoo-like
or reservation settings. We would be dealt with as parents by our machine chil-
dren: old where they are new, imperfect where they are self-perfecting, contin-
gent creatures where they are the product of intelligent design. The result will
be a world that is remade and reconstructed at the atomic level through nan-
otechnology, a world whose organization will be shaped by an intelligence that
surpasses all human comprehension.
Nearly all the elements of this story are problematic. They often involve near
metaphysical
speculation about the nature of the universe, or
technical
speculation
about things that are currently not remotely possible, or
philosophical
speculation
about matters, such as the nature of consciousness, that are topics of perennial
dispute. One could raise specific questions about the future of Moore’s Law, or
the mind-body problem, or the issue of evolution and organized complexity. Yet
while it may be comforting to latch on to a particular scientific or technical rea-
son to think that what is proposed is impossible, to do so is to bet that we under-
S
PRING
2003 ~ 89
Copyright 2003. All rights reserved. See
for more information.
stand the limits of human knowledge and ingenuity, which in fact we cannot
know in advance. When it comes to the feasibility of what might be coming, the
“extinctionists” and their critics are both speculating.
Nevertheless, the extinctionists do their best to claim that the “end of
humanity … as a biological life form” is not only possible but
necessary
. It is
either an evolutionary imperative or an unavoidable result of the technological
assumption that if “we” don’t engage in this effort, “they” will. Such arguments
are obviously thin, and the case that human beings ought to assist enthusiasti-
cally in their own extinction makes little sense on evolutionary terms, let alone
moral ones. The English novelist Samuel Butler, who considered the possibili-
ty that machines were indeed the next stage of evolution in his nineteenth-cen-
tury novel
Erewhon
(“Nowhere”), saw an obvious response: his Erewhonians
destroy most of their machines to preserve their humanity.
“Just saying no” may not be easy, especially if the majority of human beings
come to desire the salvation that the extinctionist prophets claim to offer. But so
long as saying no (or setting limits) is not impossible, it makes sense to inquire
into the goods that would supposedly be achieved by human extinction rather
than simply the mechanisms that may or may not make it possible. Putting aside
the most outlandish of these proposals—or at least suspending disbelief about
the feasibility of the science—it matters greatly whether or not we reject, on
principle, the promised goods of post-human life. By examining the moral case
for leaving biological life behind—the case for merging with and then becoming
our machines—we will perhaps understand why someone might find this
prospect appealing, and therefore discover the real source of the supposed imper-
ative behind bringing it to pass.
Wretched Body, Liberated Mind
I
n their work
Beyond Humanity: Cyberevolution and Future Minds
, evolutionary
biologist Gregory Paul and artificial intelligence expert Earl D. Cox put the case
for human extinction rather succinctly: “First we suffer, then we die. This is the
great human dilemma.” As the extinctionists see it, the problem with human life
is not simply suffering and death but the tyranny of desire: “I resent the fact,”
says Carnegie Mellon University roboticist Hans Moravec, “that I have these
very insistent drives which take an enormous amount of effort to satisfy and are
never completely appeased.” Inventor Ray Kurzweil anticipates that by 2019 vir-
tual sex, performed with the aid of various mechanisms providing complete sen-
sory feedback, will be preferred for its ability “to enhance both experience and
safety.” But this is clearly only the beginning of the story:
Group sex will take on new meaning in that more than one person can simul-
taneously share the experience of one partner … (perhaps the one virtual
body will reflect a consensus of the attempted movements of the multiple
90 ~ T
HE
N
EW
A
TLANTIS
Copyright 2003. All rights reserved. See
for more information.
partners). A whole audience of people—who may be geographically dis-
persed—could share one virtual body while engaged in sexual experience
with one performer.
Neither Moravec nor Kurzweil can be dismissed as mere cranks, even if their
judgment can rightfully be called into question. Moravec has been a pioneer in the
development of free-ranging mobile robots, particularly the software that allows
such robots to interpret and navigate their surroundings. His work in this area is
consistently supported both by the private sector and by government agencies like
NASA, the Office of Naval Research, and the Defense Advanced Research Projects
Agency. His 1988 book,
Mind Children: The Future of Robot and Human Intelligence
,
is perhaps the ur-text of “transhumanism,” the movement of those who actively
seek our technology-driven evolution beyond humanity. Kurzweil is the 1999
National Medal of Technology winner, deservedly famous for his work developing
optical character recognition systems. He invented the first text-to-speech sys-
tems for reading to the blind and created the first computer-based music synthe-
sizer that could realistically recreate orchestral instruments.
Moravec and Kurzweil share a deep resentment of the human body: both the
ills of fragile and failing flesh, and the limitations inherent to bodily life, includ-
ing the inability to fulfill our own bodily desires. Even if we worked perfectly, in
other words, there are numerous ways in which that “working” can be seen as
defective because we might have been better designed in the first place.
Take, for example, the human eye. Why is it made out of such insubstantial
materials? Why is its output cabled in such a way as to interfere with our vision?
Why is it limited to seeing such a narrow portion of the electro-magnetic spec-
trum? Of course, we think we know the answers to all such questions: this is the
way the eye evolved. Again and again, chance circumstances favored some muta-
tions over others until we have this particular (and doubtless transitory) config-
uration. Little wonder that it all seems rather cobbled together. But, the extinc-
tionists claim, we have also evolved an intelligent capacity to guide evolution.
Leaving aside all metaphysical speculation that such an outcome is the point of
the process, we can at least see whether the ability to guide evolution will confer
survival advantages or not. Having eyes, we do not walk around blindfolded.
Having the ability to guide evolution, we might as well use it.
In short, if human beings are simply mechanisms that can be improved, if our
parts are replaceable by others, then it matters little whether they are construct-
ed biologically or otherwise. That much applies to the life of the
body
. But what
about the life of the
mind
? Not only does that life arise from the biological mech-
anism of the brain, but what we experience through that mechanism is, the
extinctionists argue, already virtual reality. We have no knowledge of the real
world; we have only our brain’s processing of our body’s sensory inputs.
Consciousness is radically subjective and essentially singular. We infer it in oth-
S
PRING
2003 ~ 91
Copyright 2003. All rights reserved. See
or more information.
ers (e.g., neighbors, pets, zoo animals) from outward signs that seemingly corre-
spond to inward states we experience directly. Getting computers to show such
outward signs has been the holy grail of artificial intelligence ever since Alan
Turing invented his famous test of machine intelligence, which defines an intel-
ligent machine as one that can fool a judge into thinking that he is talking to a
human being.
Although subsequent thinkers may have developed a more sophisticated pic-
ture of when artificial life should be considered
conscious
, the guiding principle
remains the same: there is no barrier to defining the life of the mind in a way that
makes it virtually indistinguishable from the workings of computers. When all
is said and done, human distinctiveness comes to be understood as nothing other
than a particular biological configuration; it is, like all such configurations, a
transitory event on an evolutionary scale. From this point of view it becomes
difficult to justify any grave concern if the workings of evolution do to us what
they have done to so many other species; it becomes rank “speciesism” to think
that we deserve anything different.
The Temptations of Artificial Life
Y
et the extinctionists are not content to show why, like everything else, human
beings will be replaced or why the world might be better off without us. They aim
to show why human beings
should
be replaced. If we are troubled by limits and
imperfection, decay and death, we can imagine a world where intelligence has power
enough to create something better.
Central to the extinctionist project of perfecting—and thus replacing—human life
as we know it is not only the belief that our bodies are nothing more than poorly
designed machines, but that our identity is something that can exist independent of our
given body. As Moravec describes it, the essence of a person is “the pattern and the
process going on in my head and body, not the machinery supporting that process. If
the pattern is preserved, I am preserved. The rest is jelly.” In a similar vein, Kurzweil
paints a picture of how we will progressively live in closer communion with machine
intelligence; how we will create “virtual avatars” that will allow us to “multitask”; how
the coming “age of spiritual machines” will allow us, among other things, to attend
meetings and enjoy sexual encounters at the same time. From here it is a short step
to the ultimate goal: scanning the brain, duplicating its circuitry in hardware and soft-
ware, and translating ourselves into robotic form (with adequate backups, of course).
In this view, there is no reason why these post-human robots should have
human form; actually, many reasons why they should not. Moravec imagines
something he calls a “bush robot,” a collection of millions of sensory-manipula-
tive arms ranging in size from huge to nano-scale. Imagine a hand where each
of the fingers had fingers, and those fingers had fingers, scaled across many
orders of magnitude from a micron to a meter:
92 ~ T
HE
N
EW
A
TLANTIS
Copyright 2003. All rights reserved. See
for more information.
[ Pobierz całość w formacie PDF ]

Rubin,

Tematy