[Paleopsych] Edge: The Pancake People, or, "The Gods are Pounding My Head"
Premise Checker
checker at panix.com
Wed Apr 20 22:57:33 UTC 2005
The Pancake People, or, "The Gods are Pounding My Head"
http://www.edge.org/3rd_culture/foreman05/foreman05_index.html
______________________________________________________________________
But today, I see within us all (myself included) the replacement of
complex inner density with a new kind of self-evolving under the
pressure of information overload and the technology of the "instantly
available". A new self that needs to contain less and less of an inner
repertory of dense cultural inheritance--as we all become "pancake
people"--spread wide and thin as we connect with that vast network of
information accessed by the mere touch of a button.
THE PANCAKE PEOPLE, OR, "THE GODS ARE POUNDING MY HEAD" [3.8.05]
Richard Foreman
vs.
THE GÖDEL-TO-GOOGLE NET [3.8.05]
George Dyson
As Richard Foreman so beautifully describes it, we've been pounded
into instantly-available pancakes, becoming the unpredictable but
statistically critical synapses in the whole Gödel-to-Google net. Does
the resulting mind (as Richardson would have it) belong to us? Or does
it belong to something else?
THE REALITY CLUB: [11]Kevin Kelly, [12]Jaron Lanier, [13]Steven
Johnson, [14]Marvin Minsky ,[15] Douglas Rushkoff, [16]Roger Schank,
[17]James O'Donnell, [18]Rebecca Goldstein. respond to Richard Foreman
and George Dyson
___
Introduction
In early 2001, avant-garde playwright and director Richard Foreman,
called to enquire about Edge's activities. He had noticed the optimism
of the Edge crowd and the range of intellectual interests and
endeavors and felt that he needed to to begin a process to explore
these areas. Then 9/11 happened. We never had our planned meeting.
Several years have gone by and recently Foreman opened his most recent
play for his Ontological-Hysteric Theater at St. Marks Church in the
Bowery in New York City. He also announced that the play--The Gods Are
Pounding My Head--would be his last.
Foreman presents Edge with a statement and a question. The statement
appears in his program and frames the sadness of The Gods Are Pounding
My Head. The question is an opening to the future. With both, Foreman
belatedly hopes to engage Edge contributors in a discussion, and in
this regard George Dyson has written the initial response, entitled
"The Gödel-to-Google Net".
[19]--JB
RICHARD FOREMAN, Founder Director, Ontological-Hysteric Theater, has
written, directed and designed over fifty of his own plays both in New
York City and abroad. Five of his plays have received "OBIE" awards
as best play of the year--and he has received five other "OBIE'S" for
directing and for 'sustained achievement'.
[20]RICHARD FOREMAN 's [21]Edge Bio Page
_________________________________________________________________
THE PANCAKE PEOPLE, OR, "THE GODS ARE POUNDING MY HEAD"
A Statement
When I began rehearsing, I thought The Gods Are Pounding My Head would
be totally metaphysical in it's orientation. But as rehearsals
continued, I found echoes of the real world of 2004 creeping into many
of my directorial choices. So be it.
Nevertheless, this very--to my mind--elegiac play does delineate my
own philosophical dilemma. I come from a tradition of Western culture
in which the ideal (my ideal) was the complex, dense and
"cathedral-like" structure of the highly educated and articulate
personality--a man or woman who carried inside themselves a personally
constructed and unique version of the entire heritage of the West.
And such multi-faceted evolved personalities did not hesitate--
especially during the final period of "Romanticism-Modernism"--to cut
down , like lumberjacks, large forests of previous achievement in
order to heroically stake new claim to the ancient inherited land--
this was the ploy of the avant-garde.
But today, I see within us all (myself included) the replacement of
complex inner density with a new kind of self-evolving under the
pressure of information overload and the technology of the "instantly
available". A new self that needs to contain less and less of an inner
repertory of dense cultural inheritance--as we all become "pancake
people"--spread wide and thin as we connect with that vast network of
information accessed by the mere touch of a button.
Will this produce a new kind of enlightenment or
"super-consciousness"? Sometimes I am seduced by those proclaiming
so--and sometimes I shrink back in horror at a world that seems to
have lost the thick and multi-textured density of deeply evolved
personality.
But, at the end, hope still springs eternal...
___
A Question
Can computers achieve everything the human mind can achieve?
Human beings make mistakes. In the arts--and in the sciences, I
believe?--those mistakes can often open doors to new worlds, new
discoveries and developments--the mistake itself becoming the basis of
a whole new world of insights and procedures.
Can computers be programmed to 'make mistakes' and turn those mistakes
into new and heretofore unimaginable developments?
______________________________
As Richard Foreman so beautifully describes it, we've been pounded
into instantly-available pancakes, becoming the unpredictable but
statistically critical synapses in the whole Gödel-to-Google net. Does
the resulting mind (as Richardson would have it) belong to us? Or does
it belong to something else?
[dysong.150.jpg]
THE GÖDEL-TO-GOOGLE NET
George Dyson
GEORGE DYSON, science historian, is the author of Darwin Among the
Machines.
[22]George Dyson's Edge Bio Page
___
THE GÖDEL-TO-GOOGLE NET
Richard Foreman is right. Pancakes indeed!
He asks the Big Question, so I've enlisted some help: the Old
Testament prophets Lewis Fry Richardson and Alan Turing; the New
Testament prophets Larry Page and Sergey Brin.
Lewis Fry Richardson's answer to the question of creative thinking by
machines is a circuit diagram, drawn in the late 1920s and published
in 1930, illustrating a self-excited, non-deterministic circuit with
two semi-stable states, captioned "Electrical Model illustrating a
Mind having a Will but capable of only Two Ideas."
[dyson.scan.jpg]
Machines that behave unpredictably tend to be viewed as
malfunctioning, unless we are playing games of chance. Alan Turing,
namesake of the infallible, deterministic, Universal machine,
recognized (in agreement with Richard Foreman) that true intelligence
depends on being able to make mistakes. "If a machine is expected to
be infallible, it cannot also be intelligent," he argued in 1947,
drawing this conclusion as a direct consequence of Kurt Gödel's 1931
results.
"The argument from Gödel's [theorem] rests essentially on the
condition that the machine must not make mistakes," he explained in
1948. "But this is not a requirement for intelligence." In 1949, while
developing the Manchester Mark I for Ferranti Ltd., Turing included a
random number generator based on a source of electronic noise, so that
the machine could not only compute answers, but occasionally take a
wild guess.
"Intellectual activity consists mainly of various kinds of search,"
Turing observed. "Instead of trying to produce a programme to simulate
the adult mind, why not rather try to produce one which simulates the
child's? Bit by bit one would be able to allow the machine to make
more and more `choices' or `decisions.' One would eventually find it
possible to program it so as to make its behaviour the result of a
comparatively small number of general principles. When these became
sufficiently general, interference would no longer be necessary, and
the machine would have `grown up'"
That's the Old Testament. Google is the New.
Google (and its brethren metazoans) are bringing to fruition two
developments that computers have been waiting for sixty years. When
John von Neumann's gang of misfits at the Institute for Advanced Study
in Princeton fired up the first 32 x 32 x 40 bit matrix of random
access memory, no one could have imagined that the original scheme for
addressing these 40,960 ephemeral bits of information, conceived in
the annex to Kurt Gödel's office, would now have expanded, essentially
unchanged, to address all the information contained in all the
computers in the world. The Internet is nothing more (and nothing
less) than a set of protocols for extending the von Neumann address
matrix across multiple host machines. Some 15 billion transistors are
now produced every second, and more and more of them are being
incorporated into devices with an IP address.
As all computer users know, this system for Gödel-numbering the
digital universe is rigid in its bureaucracy, and every bit of
information has to be stored (and found) in precisely the right place.
It is a miracle (thanks to solid-state electronics, and
error-correcting coding) that it works. Biological information
processing, in contrast, is based on template-based addressing, and is
consequently far more robust. The instructions say "do X with the next
copy of Y that comes around" without specifying which copy, or where.
Google's success is a sign that template-based addressing is taking
hold in the digital universe, and that processes transcending the von
Neumann substrate are starting to grow. The correspondence between
Google and biology is not an analogy, it's a fact of life. Nucleic
acid sequences are already being linked, via Google, to protein
structures, and direct translation will soon be underway.
So much for the address limitation. The other limitation of which von
Neumann was acutely aware was the language limitation, that a formal
language based on precise logic can only go so far amidst real-world
noise. "The message-system used in the nervous system... is of an
essentially statistical character," he explained in 1956, just before
he died. "In other words, what matters are not the precise positions
of definite markers, digits, but the statistical characteristics of
their occurrence... Whatever language the central nervous system is
using, it is characterized by less logical and arithmetical depth than
what we are normally used to [and] must structurally be essentially
different from those languages to which our common experience refers."
Although Google runs on a nutrient medium of von Neumann processors,
with multiple layers of formal logic as a base, the higher-level
meaning is essentially statistical in character. What connects where,
and how frequently, is more important than the underlying code that
the connections convey.
As Richard Foreman so beautifully describes it, we've been pounded
into instantly-available pancakes, becoming the unpredictable but
statistically critical synapses in the whole Gödel-to-Google net. Does
the resulting mind (as Richardson would have it) belong to us? Or does
it belong to something else?
Turing proved that digital computers are able to answer most--but not
all--problems that can be asked in unambiguous terms. They may,
however, take a very long time to produce an answer (in which case you
build faster computers) or it may take a very long time to ask the
question (in which case you hire more programmers). This has worked
surprisingly well for sixty years.
Most of real life, however, inhabits the third sector of the
computational universe: where finding an answer is easier than
defining the question. Answers are, in principle, computable, but, in
practice, we are unable to ask the questions in unambiguous language
that a computer can understand. It's easier to draw something that
looks like a cat than to describe what, exactly, makes something look
like a cat. A child scribbles indiscriminately, and eventually
something appears that happens to resemble a cat. A solution finds the
problem, not the other way around. The world starts making sense, and
the meaningless scribbles are left behind.
"An argument in favor of building a machine with initial randomness is
that, if it is large enough, it will contain every network that will
ever be required," advised Turing's assistant, cryptanalyst Irving J.
Good, in 1958. Random networks (of genes, of computers, of people)
contain solutions, waiting to be discovered, to problems that need not
be explicitly defined. Google has answers to questions no human being
may ever be able to ask.
Operating systems make it easier for human beings to operate
computers. They also make it easier for computers to operate human
beings. (Resulting in Richard Foreman's "pancake effect.") These views
are complementary, just as the replication of genes helps reproduce
organisms, while the reproduction of organisms helps replicate genes.
Same with search engines. Google allows people with questions to find
answers. More importantly, it allows answers to find questions. From
the point of view of the network, that's what counts. For obvious
reasons, Google avoids the word "operating system." But if you are
ever wondering what an operating system for the global computer might
look like (or a true AI) a primitive but fully metazoan system like
Google is the place to start.
Richard Foreman asked two questions. The answer to his first question
is no. The answer to his second question is yes.
_________________________________________________________________
[kelly100.jpg]
Kevin Kelly:
"Can computers achieve everything the human mind can achieve?"
Computers can't, but the children of computers will.
[23]KEVIN KELLY is Editor-At-Large, Wired; Author, Out of Control: The
New Biology of Machines, Social Systems, and the Economic World.
_________________________________________________________________
The only way to deal with other people's brains or bodies sanely is to
grant them liberty as far as you're concerned, but to not lose hope
for them. Each person ought to decide whether to be a pancake or not,
and some of those pre-pancakes Foreman misses were actually vacuous
soufflés anyway. Remember? There are plenty of creamy rich three
dimensional digitally literate people out there, even a lot of young
ones. There is a lot of hope and beauty in digital culture, even if
the prevalent fog is sometimes heavy enough to pound your head.
[jaron100.jpg]
Jaron Lanier:
In the 1990s, I used to complain about the "suffocating nerdiness and
blandness" of Silicon Valley. This was how the pioneer days of Richard
Foreman's pancake personhood felt to me. I fled to live in New York
City precisely for the antidote of being around venues like Foreman's
Ontological-Hysterical Theater and his wonderful shows. But computer
culture broke out of its cage and swallowed Manhattan whole only a few
years later.
The very first articulate account of what information technology would
be like was written by E.M. Forster in 1909, in a story called "The
Machine Stops". The characters have something like the Internet with
video phones and the web and email and the whole shebang and they
become pancake people, but realize it, and long for a way out, which
ultimately involves a left-wing tableaux of the machine smashed. The
pancakes walk outside into a pastoral scene, and the story ends. We
don't really learn if they leaven.
Computer culture's reigning cool/hip wing of the moment, the free
software --or "open source"--movement, uses the idea of the Cathedral
as a metaphorical punching bag. In a famous essay by Eric Raymond
("The Cathedral and the Bazaar"), the Cathedral is compared
unfavorably to an anarchic village market, and the idea is that true
brilliance is to be found in the "emergent" metapersonal wisdom of
neo-Darwinian competition. The Cathedral is derided as a monument to a
closed, elitist, and ultimately constricting kind of knowledge. It's a
bad metaphor.
All this supports Foreman's pancake premise, but I recommend adding a
wing to Foreman's mental cathedral. In this wing there would be a
colossal fresco of two opposing armies. One one side, there would be a
group led by Doug Englebart. He'd be surrounded by some eccentric
characters such as the late Jef Raskin, Ted Nelson, David Gelernter,
Alan Kay, Larry Tesler, Andy van Dam, Ben Schneiderman, among others.
They are facing an opposing force made up of both robots and people
and mechochimeras.
The first group consists of members of the humanist tradition in
computer science, and are people that Foreman might enjoy. They are
not pancakes and they don't make others into pancakes. They are no
more the cause of mental shrinkage than the written word (despite
Plato's warnings to the contrary) or Gutenberg.
The only way to deal with other people's brains or bodies sanely is to
grant them liberty as far as you're concerned, but to not lose hope
for them. Each person ought to decide whether to be a pancake or not,
and some of those pre-pancakes Foreman misses were actually vacuous
soufflés anyway. Remember? There are plenty of creamy rich three
dimensional digitally literate people out there, even a lot of young
ones. There is a lot of hope and beauty in digital culture, even if
the prevalent fog is sometimes heavy enough to pound your head.
If Foreman is serious about quitting the theater, he will be missed.
But that's not a reason to offer computers, arbiters on their own of
nothing but insubstantiality, the power to kick his butt and pound his
head. The only reality of a computer is the person on the other side
of it.
[24]JARON LANIER is a Computer Scientist and Musician.
_________________________________________________________________
...the kind of door-opening exploration that Google offers is in fact
much more powerful and unpredictable than previous modes of
exploration. It's a lot easier to stumble across something totally
unexpected--but still relevant and interesting--using Google than it
is walking through a physical library or a bookstore. A card catalogue
is a terrible vehicle for serendipity. But hypertext can be a
wonderful one. You just have to use it the right way.
[Johnson,Steven100.jpg]
Steven Johnson:
I think it's a telling sign of how far the science of information
retrieval has advanced that we're seriously debating the question of
whether computers can be programmed to make mistakes. Rewind the tape
8 years or so--post-Netscape, pre-Google--and the dominant complaint
would have been that the computers were always making mistakes: you'd
plug a search query into Alta-Vista and you'd get 53,000 results, none
of which seemed relevant to what you were looking for. But sitting
here now in 2005, we've grown so accustomed to Google's ability to
find the information we're looking for that we're starting to yearn
for a little fallibility.
But the truth is most of our information tools still have a fuzziness
built into them that can, in Richard Foreman's words, "often open
doors to new worlds." It really depends on how you choose to use the
tool. Personally, I have two modes of using Google: one very directed
and goal-oriented, the other more open-ended and exploratory.
Sometimes I use Google to find a specific fact: an address, the
spelling of a name, the number of neurons estimated to reside in the
human brain, the dates of the little ice age. In those situations, I'm
not looking for mistakes, and thankfully Google's quite good at
avoiding them. But I also use Google in a far more serendipitous way,
when I'm exploring an idea or a theme or an author's work: I'll start
with a general query and probe around a little and see what the oracle
turns up; sometimes I'll follow a trail of links out from the original
search; sometimes I'll return and tweak the terms and start again.
Invariably, those explorations take me to places I wasn't originally
expecting to go--and that's precisely why I cherish them. (I have a
similar tool for exploring my own research notes--a program called
DevonThink that lets me see semantic associations between the
thousands of short notes and quotations that I've assembled on my hard
drive.)
In fact, I would go out on a limb here and say that the kind of
door-opening exploration that Google offers is in fact much more
powerful and unpredictable than previous modes of exploration. It's a
lot easier to stumble across something totally unexpected--but still
relevant and interesting--using Google than it is walking through a
physical library or a bookstore. A card catalogue is a terrible
vehicle for serendipity. But hypertext can be a wonderful one. You
just have to use it the right way.
[25]
STEVEN JOHNSON, columnist, Discover; Author: Emergence: The Connected
Lives of Ants, Brains, Cities, and Software.
_________________________________________________________________
I don't see any basic change; there always was too much information.
Fifty years ago, if you went into any big library, you would have been
overwhelmed by the amounts contained in the books therein.
Furthermore, that "touch of a button" has improves things in two ways:
(1) it has change the time it takes to find a book from perhaps
several minutes into several seconds, and (2) in the past date usually
took many minutes, or even hours, to find what you want to find inside
that book--but now, a Computer can help you can search through the
text, and I see this as nothing but good.
[minsky100.jpg]
Marvin Minsky:
Mr. Foreman complains that he is being replaced (by "the pressure of
information overload") with "a new self that needs to contain less and
less of an inner repertory of dense cultural inheritance" because he
is connected to "that vast network of information accessed by the mere
touch of a button."
I think that this is ridiculous because I don't see any basic change;
there always was too much information. Fifty years ago, if you went
into any big library, you would have been overwhelmed by the amounts
contained in the books therein. Furthermore, that "touch of a button"
has improves things in two ways: (1) it has change the time it takes
to find a book from perhaps several minutes into several seconds, and
(2) in the past date usually took many minutes, or even hours, to find
what you want to find inside that book--but now, a Computer can help
you can search through the text, and I see this as nothing but good.
Indeed, it seems to me that only one thing has gone badly wrong. I do
not go to libraries any more, because I can find most of what I want
by using that wonderful touch of a button! However the copyright laws
have gotten worse--and I think that the best thoughts still are in
books because, frequently, in those ancient times, the authors
developed their ideas for years well for they started to publicly
babble. Unfortunately, not much of that stuff from the past fifty
years is in the public domain, because of copyrights.
So, in my view, it is not the gods, but Foreman himself who has been
pounding on his own head. Perhaps if he had stopped longer to think,
he would have written something more sensible. Or on second thought,
perhaps he would not--if, in fact, he actually has been replaced.
[26]MARVIN MINSKY is a mathematician and computer scientist; Cofounder
of MIT's Artificial Intelligence Laboratory; Author, The Society of
Mind.
_________________________________________________________________
We give up the illusion of our power as deriving from some notion of
individual collecting data, and find out that having access to data
through our network-enabled communities gives us an entirely more
living flow of information that is appropriate to the ever changing
circumstances surrounding us. Instead of growing high, we grow wide.
We become pancake people.
[rushkoff100.jpg]
Douglas Rushkoff:
I don't think it's the computer itself enabling the pancake people,
but the way networked computers give us access to other people. It's
not the data--for downloaded data is just an extension of the wealthy
gentleman in his library, enriching himself as a "self." What creates
the pancake phenomenon is our access to other people, and the
corresponding dissolution of our perception of knowledge as an
individual's acquisition.
Foreman is hinting at a "renaissance" shift I've been studying for the
past few years.
The original Renaissance invented the individual. With the development
of perspective in painting came the notion of perspective in
everything. The printing press fueled this even further, giving
individuals the ability to develop their own understanding of texts.
Each man now had his own take on the world, and a person's storehouse
of knowledge and arsenal of techniques were the measure of the man.
The more I study the original Renaissance, the more I see our own era
as having at least as much renaissance character and potential. Where
the Renaissance brought us perspective painting, the current one
brings virtual reality and holography. The Renaissance saw humanity
circumnavigating the globe; in our own era we've learned to orbit it
from space. Calculus emerged in the 15th Century, while systems theory
and chaos math emerged in the 20th. Our analog to the printing press
is the Internet, our equivalent of the sonnet and extended metaphor is
hypertext.
Renaissance innovations all involve an increase in our ability to
contend with dimension: perspective. Perspective painting allowed us
to see three dimensions where there were previously only two.
Circumnavigation of the globe changed the world from a flat map to a
3D sphere. Calculus allowed us to relate points to lines and lines to
objects; integrals move from x to x-squared, to x-cubed, and so on.
The printing press promoted individual perspectives on religion and
politics. We all could sit with a text and come up with our own,
personal opinions on it. This was no small shift: it's what led to the
Protestant wars, after all.
Out of this newfound experience of perspective was born the notion of
the individual: the Renaissance Man. Sure, there were individual
people before the Renaissance, but they existed mostly as parts of
small groups. With literacy and perspective came the abstract notion
the person as a separate entity. This idea of a human being as a
"self," with independent will, capacity, and agency, was pure
Renaissance--a rebirth and extension of the Ancient Greek idea of
personhood. And from it, we got all sorts of great stuff like the
autonomy of the individual, agency, and even democracy and the
republic. The right to individual freedom is what led to all those
revolutions.
But thanks to new emphasis on the individual, it was also during the
first great Renaissance that we developed the modern concept of
competition. Authorities became more centralized, and individuals
competed for how high they could rise in the system. We like to think
of it as a high-minded meritocracy, but the rat-race that ensued only
strengthened the authority of central command. We learned compete for
resources and credit made artificially scarce by centralized banking
and government.
While our renaissance also brings with it a shift in our relationship
to dimension, the character of this shift is different. In a
holograph, fractal, or even an Internet web site, perspective is no
longer about the individual observer's position; it's about that
individual's connection to the whole. Any part of a holographic plate
recapitulates the whole image; bringing all the pieces together
generates greater resolution. Each detail of a fractal reflects the
whole. Web sites live not by their own strength but the strength of
their links. As Internet enthusiasts like to say, the power of a
network is not the nodes, it's the connections.
That's why new models for both collaboration and progress have emerged
during our renaissance--ones that obviate the need for competition
between individuals, and instead value the power of collectivism. The
open source development model, shunning the corporate secrets of the
competitive marketplace, promotes the free and open exchange of the
codes underlying the software we use. Anyone and everyone is invited
to make improvements and additions, and the resulting projects--like
the Firefox browser--are more nimble, stable, and user-friendly.
Likewise, the development of complementary currency models, such as
Ithaca Hours, allow people to agree together what their goods and
services are worth to one another without involving the Fed. They
don't need to compete for currency in order to pay back the central
creditor--currency is an enabler of collaborative efforts rather than
purely competitive ones.
For while the Renaissance invented the individual and spawned many
institutions enabling personal choices and freedoms, our renaissance
is instead reinventing the collective in a new context. Originally,
the collective was the clan or the tribe--an entity defined no more by
what members had in common with each other than what they had in
opposition to the clan or tribe over the hill.
Networks give us a new understanding of our potential relationships to
one another. Membership in one group does not preclude membership in a
myriad of others. We are all parts of a multitude of overlapping
groups with often paradoxically contradictory priorities. Because we
can contend with having more than one perspective at a time, we
needn't force them to compete for authority in our hearts and
minds--we can hold them all, provisionally. That's the beauty of
renaissance: our capacity to contend with multiple dimensions is
increased. Things don't have to be just one way or directed by some
central authority, alive, dead or channeled. We have the capacity to
contend with spontaneous, emergent reality.
We give up the illusion of our power as deriving from some notion of
individual collecting data, and find out that having access to data
through our network-enabled communities gives us an entirely more
living flow of information that is appropriate to the ever changing
circumstances surrounding us. Instead of growing high, we grow wide.
We become pancake people.
[27]DOUGLAS RUSHKOFF is a media analyst; Documentary Writer; Author,
Media Virus.
_________________________________________________________________
As to Dyson's remarks: "Turing proved that digital computers are able
to answer most but not all¤ programs that can be asked in unambiguous
terms." Did he? I missed that. Maybe he proved that computers could
follow instructions which is neither here nor there. It is difficult
to give instructions about how to learn new stuff or get what you
want. Google's "allowing people with questions to find answers" is
nice but irrelevant. The Encyclopedia Britannica does that as well and
no one makes claims about its intelligence or draws any conclusion
whatever from it. And, Google is by no means an operating system--I
can't even imagine what Dyson means by that or does he just not know
what an operating system is?
[schank100.jpg]
Roger Schank:
I am constantly astounded by people who use computers but who really
don't understand them at all when I hear people talk about artificial
intelligence (AI) . I shouldn't be surprised by most folk's lack of
comprehension I suppose, since the people inside AI often fail to get
it as well. I recently attended a high level meeting in Washington
where the AI people and the government people were happily dreaming
about what computers will soon be able to do and promising that they
would soon make it happen when they really had no idea what was
involved in what they were proposing. So, that being said, let me talk
simply about what it would mean and what it would look like for a
computer to be intelligent.
Simple point number 1: A smart computer would have to be able to
learn.
This seems like an obvious idea. How smart can you be if every
experience seems brand new? Each experience should make you smarter
no? If that is the case then any intelligent entity must be capable of
learning from its own experiences right?
Simple point number 2: A smart computer would need to actually have
experiences. This seems obvious too and follows from simple point
number 1. Unfortunately, this one isn't so easy. There are two reasons
it isn't so easy. The first is that real experiences are complex, and
the typical experience that today's computers might have is pretty
narrow. A computer that walked around the moon and considered
seriously what it was seeing and decided where to look for new stuff
based on what it had just seen would be having an experience. But,
while current robots can walk and see to some extent, they aren't
figuring out what to do next and why. A person is doing that. The best
robots we have can play soccer. They play well enough but really not
all that well. They aren't doing a lot of thinking. So there really
aren't any computers having much in the way of experiences right now.
Could there be computer experiences in some future time? Sure. What
would they look like? They would have to look a lot like human
experiences. That is, the computer would have to have some goal it was
pursuing and some interactions caused by that goal that caused it to
modify what it was up to in mid-course and think about a new strategy
to achieve that goal when it encountered obstacles to the plans it had
generated to achieve that goal. This experience might be
conversational in nature, in which case it would need to understand
and generate complete natural language, or it might be physical in
nature, in which case it would need to be able to get around and see,
and know what it was looking at. This stuff is all still way too hard
today for any computer. Real experiences, ones that one can learn
from, involve complex social interactions in a physical space, all of
which is being processed by the intelligent entities involved. Dogs
can do this to some extent. No computer can do it today. Tomorrow
maybe.
The problem here is with the goal. Why would a computer have a goal it
was pursuing? Why do humans have goals they are pursuing? They might
be hungry or horny or in need of a job, and that would cause goals to
be generated, but none of this fits computers. So, before we begin to
worry about whether computers would make mistakes, we need to
understand that mistakes come from complex goals not trivially
achieved. We learn from the mistakes we make when the goal we have
failed at satisfying is important to us and we choose to spend some
time thinking about what to do better next time. To put this another
way, learning depends upon failure and failure depends upon having had
a goal one care's about achieving and that one is willing to spend
time thinking about how to achieve next time using another plan. Two
year olds do this when they realize saying "cookie" works better than
saying "wah" when they want a cookie.
The second part of the experience point is that one must know one has
had an experience and know the consequences of that experience with
respect to one's goals in order to even think about improving. In
other words, a computer that thinks would be conscious of what had
happened to it, or would be able to think it was conscious of what had
happened to it which may not be the same thing.
Simple point number 3: Computers that are smart won't look like you
and me.
All this leads to the realization that human experience depends a lot
on being human. Computers will not be human. Any intelligence they
ever achieve will have to come by virtue of their having had many
experiences that they have processed and understood and learned from
that have helped them better achieve whatever goals they happen to
have.
So, to Foreman's question: Computers will not be programmed to make
mistakes. They will be programmed to attempt to achieve goals and to
learn from experience. They will make mistakes along the way, as does
any intelligent entity.
As to Dyson's remarks: "Turing proved that digital computers are able
to answer most but not all¤ programs that can be asked in unambiguous
terms." Did he? I missed that. Maybe he proved that computers could
follow instructions which is neither here nor there. It is difficult
to give instructions about how to learn new stuff or get what you
want. Google's "allowing people with questions to find answers" is
nice but irrelevant. The Encyclopedia Britannica does that as well and
no one makes claims about its intelligence or draws any conclusion
whatever from it. And, Google is by no means an operating system--I
can't even imagine what Dyson means by that or does he just not know
what an operating system is?
People have nothing to fear from smart machines. With the current
state of understanding of AI I suspect they wont have to even see any
smart machines any time soon. Foreman's point was about people after
all and people are being changed by the computer's ubiquity in their
lives. I think the change is, like all changes in the nature of man's
world, interesting and potentially profound, and probably for the
best. People may well be more pancake-like, but the syrup is going to
very tasty.
[28]ROGER SCHANK is a Psychologist & Computer Scientist; Author,
Designing World-Class E-Learning.
_________________________________________________________________
I have trouble imagining what students will know fifty years from now,
when devices in their hands spare them the need to know multiplication
tables or spelling or dates of the kings of England. That probably
leaves us time and space for other tasks, but the sound of the gadgets
chasing us is palpable. What humans will be like, accordingly, in 500
years is just beyond our imagining.
[odonnell100.jpg]
James O'Donnell:
Can computers achieve everything the human mind can achieve? Can they,
in other words, even make fruitful mistakes? That's an ingenious
question.
Of course, computers never make mistakes--or rather, a computer's
"mistake" is a system failure, a bad chip or a bad disk or a power
interruption, resulting in some flamboyant mis-step, but computers can
have error-correcting software to rescue them from those. Otherwise, a
computer always does the logical thing. Sometimes it's not the thing
you wanted or expected, and so it feels like a mistake, but it usually
turns out to be a programmer's mistake instead.
It's certainly true that we are hemmed in constantly by technology.
The technical wizardry in the graphic representation of reality that
generated a long history of representative art is now substantially
eclipsed by photography and later techniques of imaging and
reproduction. Artists and other humans respond by doing more and more
creatively in the zone that is still left un-competed, but if I want
to know what George W. Bush looks like, I don't need to wait for a
Holbein to track him down. We may reasonably expect to continue to be
hemmed in. I have trouble imagining what students will know fifty
years from now, when devices in their hands spare them the need to
know multiplication tables or spelling or dates of the kings of
England. That probably leaves us time and space for other tasks, but
the sound of the gadgets chasing us is palpable. What humans will be
like, accordingly, in 500 years is just beyond our imagining.
So I'll ask what I think is the limit case question: can a computer be
me? That is to say, could there be a mechanical device that embodied
my memory, aptitudes, inclinations, concerns, and predilections so
efficiently that it could replace me? Could it make my mistakes?
I think I know the answer to that one.
[29]JAMES O'DONNELL is a classicist; cultural historian; Provost,
Georgetown University; Author, Avatars of the Word.
_________________________________________________________________
The complexity suddenly facing us can feel overwhelming and perhaps
such souls as Lugubrioso's will momentarily shrink at how much they
must master in order to appropriate this complexity and make it their
own. It's that shrinkage that Lugubriosos is feeling, confusing his
own inadequacy to take in the new forms of knowing with the inadequacy
of the forms themselves. Google doesn't kill people, Rosa admonished
him. People kill people.
[goldstein100.jpg]
Rebecca Goldstein:
I admit that I'm of two distinct minds on the question posed by
Richard Foreman as to whether the technological explosion has led to
an expansion or a flattening of our selves. In fact, a few years ago
when I was invited to represent the humanities at Princeton
University's celebration of the centenary of their graduate studies, I
ended up writing a dialogue to express my inner bifurcation. My way of
posing the question was to wonder whether the humanities, those
"soul-explorations," had any future at all, given that the soul had
been all but pounded out of existence, or in any case pounded into a
very attenuated sort of existence.
My one character, dubbed Lugubrioso, had a flair for elaborate
phraseology that rivaled the Master's, and he turned it to deploring
the loss of the inner self's solemn, silent spaces, the hushed
corridors where the soul communes with itself, chasing down the
subtlest distinctions of fleeting consciousness, catching them in
finely wrought nets of words, each one contemplated for both its
precise meaning and euphony, its local and global qualities, one's
flight after that expressiveness which is thought made surer and
fleeter by the knowledge of all the best that had been heretofore
thought, the cathedral-like sentences (to change the metaphor) that
arose around the struggle to do justice to inexhaustible complexity
themselves making of the self a cathedral of consciousness.
(Lugubrioso spoke in long sentences.)
He contemplated with shuddering horror the linguistic impoverishment
of our technologically abundant lives, arguing that privation of
language is both an effect and a cause of privation of thought. Our
vocabularies have shrunk and so have we. Our expressive styles have
lost all originality and so have we. The passivity of our image-heavy
forms of communication--too many pictures, not enough words,
Lugubrioso cried out pointing his ink-stained finger at the popular
culture--substitutes in an all-too-pleasant anodyne for the rigors of
thinking itself, and our weakness for images encourages us to reduce
people, too--even our very own selves--to images, which is why we are
drunk on celebrity hood and feel ourselves to exist only to the extent
that we exist for others.
What is left but image when the self has stopped communing with
itself, so that in a sad gloss on Bishop Berkeley's apothegm, our esse
has become percipi, our essence is to be perceived? Even the torrents
of words posted on "web-related locations" (the precise nature of
which Lugubrioso had kept himself immaculately ignorant) are not words
that are meant for permanence; they are pounded out on keyboards at
the rate at which they are thought, and will vanish into oblivion just
as quickly, quickness and forgetfulness being of the whole essence of
the futile affair, the long slow business of matching coherence to
complexity unable to keep up, left behind in the dust.
My other character was Rosa and she pointed out that at the very
beginning of this business that Lugubrioso kept referring to, in
stentorian tones, as "Western Civilization," Plato deplored the
newfangled technology of writing and worried that it tolled the death
of thought. A book, Plato complained in Phaedrus, can't answer for
itself. (Rosa found the precise quotation on the web. She found
'stentorian,' too, when she needed it, on her computer's thesaurus.
She sort of knew the word, thought it might be "sentorian," or
"stentorious"but she'll know where to find it if she ever needs it
again, a mode of knowing that Lugubrioso regards as epistemologically
damnable.)
When somebody questions a book, Plato complained, it just keeps
repeating the same thing over and over again. It will never, never, be
able to address the soul as a living breathing interlocutor can, which
is why Plato, committing his thoughts to writing with grave
misgivings, adopted the dialogue form, hoping to approximate something
of the life of real conversation. Plato's misgivings are now
laughable--nobody is laughing harder than Lugubrioso at the thought
that books diminish rather than enhance the inner life--and so, too,
will later generations laugh at Lugubrioso's lamentations that the
cognitive enhancements brought on by computers will make of us less
rather than more.
Human nature doesn't change, Rosa tried to reassure Lugubrioso,
backing up her claims with the latest theories of evolutionary
psychology propounded by Steven Pinker et al. Human nature is
inherently expansive and will use whatever tools it develops to grow
outward into the world. The complexity suddenly facing us can feel
overwhelming and perhaps such souls as Lugubrioso's will momentarily
shrink at how much they must master in order to appropriate this
complexity and make it their own. It's that shrinkage that Lugubriosos
is feeling, confusing his own inadequacy to take in the new forms of
knowing with the inadequacy of the forms themselves. Google doesn't
kill people, Rosa admonished him. People kill people.
Lugubrioso had a heart-felt response, but I'll spare you.
[30]REBECCA GOLDSTEIN is a philosopher and novelist; Author,
Incompleteness.
References
11. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#kelly
12. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#lanier
13. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#johnson
14. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#minsky
15. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#rushkoff
16. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#schank
17. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#odonnell
18. http://www.edge.org/3rd_culture/foreman05/foreman05_index.html#goldstein
19. http://www.edge.org/3rd_culture/bios/brockman.html
20. http://www.edge.org/3rd_culture/bios/foreman.html
21. http://www.edge.org/3rd_culture/bios/foreman.html
22. http://www.edge.org/3rd_culture/bios/dysong.html
23. http://www.edge.org/3rd_culture/bios/kelly.html
24. http://www.edge.org/3rd_culture/bios/lanier.html
25. http://www.edge.org/3rd_culture/bios/johnson.html
26. http://www.edge.org/3rd_culture/bios/minsky.html
27. http://www.edge.org/3rd_culture/bios/rushkoff.html
28. http://www.edge.org/3rd_culture/bios/schank.html
29. http://www.edge.org/3rd_culture/bios/odonnell.html
30. http://www.edge.org/3rd_culture/bios/goldstein.html
More information about the paleopsych
mailing list