http://biz.yahoo.com/bw/010222/0112_3.html

The following presentation was made at the 2nd Annual NASA/DoD
Workshop on Evolutionary Hardware, in Palo Alto, California,
July 14, 2000. It gives a technical overview of CBM.

http://genobyte.com/slide1.html

http://foobar.starlab.net/~degaris/essays/KUL.html

http://ic-www.arc.nasa.gov/ic/psa/

http://foobar.starlab.net/~degaris/news/moduleimages.html

http://foobar.starlab.net/~degaris/news/political.html

The article is not yet online, but if you have a copy of the most
recent Red Herring zine, flip to page 152.  The article is written by
Mark Williams (markred@...) and titled "Mind reader".  The
sub-title reads "The next computer revolution asks, What is mind, and
what does it mean to be human?"  Nevertheless, interesting reading.

Hallo and happy Holidays .. cosmists(in favor of ultra intelligent
machines).
two items.
1. ArtilectWorld (http://artilect.org) has a new look.
2. a free ebook  http://www.anice.net.ar/intsyst/

So in closing,
ArtilectWorld has two sites
http://artilect.org
the Main Site
and
http://artilectworld.com
the store.. for artilect goodies

Doug

happy thanksgiving  = )


http://www.msnbc.com/news/493124.asp?bt=nm&btu=http://www.msnbc.com/tools/ne
wstools/d/news_menu.asp&cp1=1

http://www.zdnet.com/intweek/stories/news/0,4164,2643123,00.html

http://www.neuron.co.uk/

http://www.msnbc.com/news/475751.asp?bt=nm&btu=http://www.msnbc.com/tools/ne
wstools/d/news_menu.asp

http://www.justlikethat.com/client/philosophy_user/ai.html

http://www.inyouremail.com/computerezine/articles/20001005-2.shtml

David Bennett has written,

> The issues of how long and precisely which hardware are not all that
> important. We can predict that a computing device with sufficient power to
> be "intelligent" or "conscious" will appear within roughly 30-50 years from
> now. If it's 100 or 200 years, it really doesn't matter.

Well, it may not matter to YOU, but it sure as hell matters to the people who
are working day and night to develop artificial intelligence. If it takes a
hundred years, rather than thirty years, that will make a huge difference to
those who have invested their lives in the project.

> The missing piece for now is the programming. We have no idea how to build
> a machine that smart. One day I expect that problem to be solved, by humans
> working with machine assistance.

More than likely, I think, genetic programming and evolvable algorithms will
eventuate in artificial intelligence. We don't need to code the AI directly.
Once sufficient complexity is instilled in them, natural computational selection
can develop intelligence. After that, the AI can optimize itself and
autocatalytically increase its intelligence via positive feedback loops.

> 30 years down the track, when the machines are really smart but still
> friendly, we have a "3-mile Island". A military experiment to produce an
> artificial soldier capable of killing the enemy (humans) goes wrong and one
> (just one!) machine gets loose in a big city and kills 23,000 people before
> being stopped by a human with an anti-tank weapon. Sounds like a plot for a
> B-grade movie, doesn't it?

No, it sounds like I'd be on the side of the artificial soldier that could put a
stop to such idiotic scenarios.

> You can't control the technology and you can't be certain that nice humans
> will get to cull all the nasty machines. Expect a rough ride!

Expect continued disintegration at all levels of social life. Feminism is just
the tip of the iceberg.

--J. R.

This life is a test. It is only a test.
              Had this been an actual life, you would
              have received further instructions as
              to what to do and where to go.

J. R. Molloy wrote:

* Long before machines become smarter than humans (I don't think it will
happen
* all at once -- it's taken decades to go from vacuum tube computers to
integrated
* circuits) humans will have the chance to select the most cooperative
Artilects.

The issues of how long and precisely which hardware are not all that
important. We can predict that a computing device with sufficient power to
be "intelligent" or "conscious" will appear within roughly 30-50 years from
now. If it's 100 or 200 years, it really doesn't matter.

The missing piece for now is the programming. We have no idea how to build
a machine that smart. One day I expect that problem to be solved, by humans
working with machine assistance.

Then we have a partnership. First we have machines much dumber than we are.
then not so dumb. Then nearly as smart. Then smarter. Then much smarter.
And so on.

Assume we successfully cull "beserker" tendencies early on. The machines
are dumb but friendly. Experimentation continues. We have regulation and
control procedures. Investors are making heaps of money.

30 years down the track, when the machines are really smart but still
friendly, we have a "3-mile Island". A military experiment to produce an
artificial soldier capable of killing the enemy (humans) goes wrong and one
(just one!) machine gets loose in a big city and kills 23,000 people before
being stopped by a human with an anti-tank weapon. Sounds like a plot for a
B-grade movie, doesn't it?

There's lots of scenarios and maybe this one is not the right one. I just
don't go for the warm-and-cuddlies. There are too many bad people out there
and too many incentives to make machines of all sorts, including bad ones.
You can't control the technology and you can't be certain that nice humans
will get to cull all the nasty machines. Expect a rough ride!

Regards
David Bennett
[ POWERflex Corporation     Developers of PFXplus ]
[ Tel:  +61-3-9888-5833     Fax:  +61-3-9888-5451 ]
[ E-mail: sales@...   support@... ]
[ Web home: www.pfxcorp.com   me: dmb@... ]

Take a look at this microbot that can move individual cells.
Expect similar machines (nanobots) to move individual molecules in a few years.

http://www.ifm.liu.se/Applphys/ConjPolym/research/micromuscles/mr/microrobot.htm
l

James Berge writes,

> But, how safe will we be in a world where the only chance we
> have is to program these genetic algorithms with a basic
> moral 'code' and hope for the best. One nanomachine with the
> ethics of a Charles Whitman could do a lot more damage than
> a lunatic in a Texas bell tower

We don't know that. What we do know is that Artilects are on the way, whether we
want them or not.

Check out:
http://www.business2.com/content/magazine/indepth/2000/09/12/17734

> (Sometimes these discussione out here in the net seem like
> the swan song of a civilization that is past its zenith and
> is fast approaching the end of the curve....)

Sometimes I think it would be better to give the world to Artilects and be done
with it.
People really piss me off sometimes.

--J. R.

David Bennett asks,

> * Any unfriendly AI agents (unlike human children) could simply be
> terminated.
>
> By whom? Humans would be neither smart enough nor quick enough. Machines
> would do the culling, and therefore...

Long before machines become smarter than humans (I don't think it will happen
all at once -- it's taken decades to go from vacuum tube computers to integrated
circuits) humans will have the chance to select the most cooperative Artilects.

> * This would result in a population of AIs with docility and compliance as
> part of their genetic code.
>
> ...the result would be a population with XXX and YYY as part of their
> genetic code (where XXX and YYY are as yet unknown factors desired by the
> machines doing the culling, probably not including either docility or
> compliance.)

Why *not* including docility or compliance?

> * Moravec's Mind Children could obviously number in the millions from the
> start, because as soon as one is developed, it could be duplicated ad
> infinitum. With an unlimited supply of genetically programmed AIs, their
> evolution could be guided and directed as experimenters see fit. The
> socialization of AI would consequently be far easier than the socialization
> of humans.
>
> I agree, but it would be machine experimenters, machine socialisation and
> machine society, not human.

That's in the future. For now it's all human.

> Inherent in this proposal is the concept that machines are never as smart
> as we are, and therefore we remain as their gods. That position is
> untenable.

In the future... once again. For now that position is not only tenable, it's
unavoidable.
(Unless you have a real Artilect that you've been hiding.)

> This entire thread of discussion leads to the conclusion that eventually
> the machines get to be far smarter than we are, and therefore we can no
> longer control them.
>
> Our one and only chance is like bringing up children. Give them some good
> values to start with, and then hope and pray that they do the right thing.

The difference is that with machines, we don't need to just "hope and pray" --
we can immediately terminate any Mind Children (Moravec's term) that fail to
adequately cooperate.

--J. R.

ozze@... thought you would be interested in this article at
Salon.com (http://www.salon.com/).


- - - - - - - - - - - -

Robots "R" us
By Janelle Brown

http://www.salon.com/tech/books/2000/09/14/robo_sapiens/index.html

The evil machine is mankind's favorite collective apocalyptic fantasy: from the
robotic femme fatale Futura of Fritz Lang's "Metropolis"  through the rebellion
of HAL in "2001" to the dystopian visions of "The Matrix" and "Terminator," we
love the shivering idea that by building robots, we are in fact constructing our
own demise. Imagine if artificially intelligent robots decided they didn't like
pandering to our pathetic flesh and turned against us: It's a titillating
fantasy, far enough away to seem unlikely to happen in your lifetime yet close
enough to draw that acrid taste of fear.

<p>"Robo sapiens: Evolution of a New Species" takes this notion and runs with
it. Part photo essay and part robotics primer, with a healthy dose of fatalistic
futurism, journalist Faith D'Aluisio and photographer Peter Menzel have
assembled an accessible guide to the field of robotics. They start out with a
bang -- the shocking cover image of an eerily fetus-like robot head is possibly
the most disturbing photo ever seen on a coffee table book -- and manage to turn
interviews with over a hundred of the geekiest humans around the world into a
curious peek at the future that will satisfy both the layman and the engineer
alike.

<p>The question that is at the crux of "Robo sapiens" is posited right up front,
as a comparison to the evolution of aviation: "We accelerated from the Wright
stuff to the right stuff so quickly that the question inevitably arises: How
long -- or how short -- a time will it be to the next step, Robo sapiens?"

- - - - - - - - - - - -

*In this message, responses are prefaced by ^^^^^


------Original Message------
From: David Bennett <dmb@...>
To: "'Artilect@egroups.com'" <Artilect@egroups.com>
Sent: September 6, 2000 4:02:17 AM GMT
Subject: RE: [Artilect] Digest Number 54

James Berge wrote:

* I am certain that if we feed the right information into
one
* of these super-super computers that they will indeed
arrive
* at various scientific theories that have been until know,
* outside the realm of human discovery. However, I also
* believe that since the information used will be derived
from the sum of
human knowledge, I am certain that the computer could
explain itself to a
human in such a away as to be understood (unless it was
playing games with
the human).

I fear you either underestimate the problem or overestimate
our
intelligence.

^^^^^Well, perhaps a bit of both. However, it concerns me
that we ^^^^^may be giving these technologies 'control'
rather than ^^^^^'responsibility'. When responsibility
becomes control, we ^^^^^have abdicated our role as masters
of our domain (! pun ^^^^^quite unintended) and we face the
consequences.

It is indeed true that many scientific theories can be
explained, up to a
point, in much simplified language. That does not mean that
a layman can
"understand" relativity or Quantum Mechanics, but given a
suitable
framework they can understand some of the concepts and
follow some of the
consequences. And perhaps our best and brightest could do
that too, if the
machines ever bother to explain their new discoveries in
really simple
terms.

^^^^^Again I agree, to a point. But, having said that, are
we ^^^^^asking these technologies to problem solve for our
own ^^^^^enlightenment or for our comfort? I mean, what use
is a ^^^^^scientific conclusion without an innate
understanding of the ^^^^^processes and methods involved? I
would be wary of asking a ^^^^^scientist a question, getting
an answer and then not being ^^^^^able to have it explained
to me. I fear we will be in a ^^^^^situation where we accept
advances made by (as opposed to ^^^^^IN) Ai, in the belief
that if it's in our 'best interest' it ^^^^^must be good, no
questions asked and no explanations needed. ^^^^^I'm a
consultant, not a physicist but I still understand the
^^^^^rudiments of Einstein's ToR and I know there are
scientists ^^^^^who understand it's esoterica in terms way
over my ^^^^^head, but nonetheless understood by the human
mind.

I am convinced that many of the concepts, proofs,
techniques, processes and methodologies we are currently
using are already beyond (unassisted) human abilities. 200
years ago a scientist could explain the whole of science to
any educated layman; 100 years ago some serious mathematics
was needed in
many areas; 50 years ago specialisation restricted
scientists largely to
their own fields and now, most advanced science is
impossible without
computer assistance. It's getting harder and it's happening
fast!

100 years from how, machines will be the only ones who
(which?) understand
the most advanced theories. We shall be satisfied if we can
pick up the
gist in Scientific American.

^^^^^I couldn't have said it better myself!

Regards
David Bennett
[ POWERflex Corporation     Developers of PFXplus ]
[ Tel:  +61-3-9888-5833     Fax:  +61-3-9888-5451 ]
[ E-mail: sales@...   support@... ]
[ Web home: www.pfxcorp.com   me: dmb@... ]

Sincerely Yours,
James Berge
San Francisco, CA
415.551.9744 - direct/voicemail
415.430.2179 x1375 - fax
james.berge@... - email


______________________________________________
FREE Personalized Email at Mail.com
Sign up at http://www.mail.com/?sr=signup

As a consultant in the IT industry, I support the research
and development that is taking place, across the spectrum.
To seek answers to our questions is not wisdom, widsom comes
from what we do with the answers we receive.

But, how safe will we be in a world where the only chance we
have is to program these genetic algorithms with a basic
moral 'code' and hope for the best. One nanomachine with the
ethics of a Charles Whitman could do a lot more damage than
a lunatic in a Texas bell tower

I encourage people not to be so arrogant as to believe that
these machines will always be our servants. Mainly because I
weep for the future if our destiny is in the hands of
machines with untenable ethics and a moral code so basic
that we wouldn't trust it with a Marine (no offense meant to
any leathernecks out there, Semper Fi)

(Sometimes these discussione out here in the net seem like
the swan song of a civilization that is past its zenith and
is fast approaching the end of the curve....)

------Original Message------
From: David Bennett <dmb@...>
To: "'Artilect@egroups.com'" <Artilect@egroups.com>
Sent: September 7, 2000 3:25:07 AM GMT
Subject: RE: [Artilect] Digest Number 55


J. R. Molloy wrote:

* AFAIK, Asimov never considered the possibility of genetic
programming and
evolvable machines which could compete against each other to
reach higher
levels of IQ. With thousands (or millions and billions) of
artificially
intelligent agents battling each other to reproduce, all
humans would need
to do is to cull the herd, so to speak.

Once intelligent machines are as common as bacteria, an
attempt to cull the
herd would have roughly the same success rate as do
antibiotics. You would
locally kill of the undesirable elements, and encourage
others to fill the
void by natural selection.

* Any unfriendly AI agents (unlike human children) could
simply be
terminated.

By whom? Humans would be neither smart enough nor quick
enough. Machines
would do the culling, and therefore...

* This would result in a population of AIs with docility and
compliance as
part of their genetic code.

....the result would be a population with XXX and YYY as
part of their
genetic code (where XXX and YYY are as yet unknown factors
desired by the
machines doing the culling, probably not including either
docility or
compliance.)

* Moravec's Mind Children could obviously number in the
millions from the
start, because as soon as one is developed, it could be
duplicated ad
infinitum. With an unlimited supply of genetically
programmed AIs, their
evolution could be guided and directed as experimenters see
fit. The
socialization of AI would consequently be far easier than
the socialization
of humans.

I agree, but it would be machine experimenters, machine
socialisation and
machine society, not human.

Inherent in this proposal is the concept that machines are
never as smart
as we are, and therefore we remain as their gods. That
position is
untenable.

This entire thread of discussion leads to the conclusion
that eventually
the machines get to be far smarter than we are, and
therefore we can no
longer control them.

Our one and only chance is like bringing up children. Give
them some good
values to start with, and then hope and pray that they do
the right thing.

Regards
David Bennett
[ POWERflex Corporation     Developers of PFXplus ]
[ Tel:  +61-3-9888-5833     Fax:  +61-3-9888-5451 ]
[ E-mail: sales@...   support@... ]
[ Web home: www.pfxcorp.com   me: dmb@... ]

Sincerely Yours,
James Berge
San Francisco, CA
415.551.9744 - direct/voicemail
415.430.2179 x1375 - fax
james.berge@... - email


______________________________________________
FREE Personalized Email at Mail.com
Sign up at http://www.mail.com/?sr=signup

J. R. Molloy wrote:

* AFAIK, Asimov never considered the possibility of genetic programming and
evolvable machines which could compete against each other to reach higher
levels of IQ. With thousands (or millions and billions) of artificially
intelligent agents battling each other to reproduce, all humans would need
to do is to cull the herd, so to speak.

Once intelligent machines are as common as bacteria, an attempt to cull the
herd would have roughly the same success rate as do antibiotics. You would
locally kill of the undesirable elements, and encourage others to fill the
void by natural selection.

* Any unfriendly AI agents (unlike human children) could simply be
terminated.

By whom? Humans would be neither smart enough nor quick enough. Machines
would do the culling, and therefore...

* This would result in a population of AIs with docility and compliance as
part of their genetic code.

...the result would be a population with XXX and YYY as part of their
genetic code (where XXX and YYY are as yet unknown factors desired by the
machines doing the culling, probably not including either docility or
compliance.)

* Moravec's Mind Children could obviously number in the millions from the
start, because as soon as one is developed, it could be duplicated ad
infinitum. With an unlimited supply of genetically programmed AIs, their
evolution could be guided and directed as experimenters see fit. The
socialization of AI would consequently be far easier than the socialization
of humans.

I agree, but it would be machine experimenters, machine socialisation and
machine society, not human.

Inherent in this proposal is the concept that machines are never as smart
as we are, and therefore we remain as their gods. That position is
untenable.

This entire thread of discussion leads to the conclusion that eventually
the machines get to be far smarter than we are, and therefore we can no
longer control them.

Our one and only chance is like bringing up children. Give them some good
values to start with, and then hope and pray that they do the right thing.

Regards
David Bennett
[ POWERflex Corporation     Developers of PFXplus ]
[ Tel:  +61-3-9888-5833     Fax:  +61-3-9888-5451 ]
[ E-mail: sales@...   support@... ]
[ Web home: www.pfxcorp.com   me: dmb@... ]

James Berge wrote:

* I am certain that if we feed the right information into one
* of these super-super computers that they will indeed arrive
* at various scientific theories that have been until know,
* outside the realm of human discovery. However, I also
* believe that since the information used will be derived from the sum of
human knowledge, I am certain that the computer could explain itself to a
human in such a away as to be understood (unless it was playing games with
the human).

I fear you either underestimate the problem or overestimate our
intelligence.

It is indeed true that many scientific theories can be explained, up to a
point, in much simplified language. That does not mean that a layman can
"understand" relativity or Quantum Mechanics, but given a suitable
framework they can understand some of the concepts and follow some of the
consequences. And perhaps our best and brightest could do that too, if the
machines ever bother to explain their new discoveries in really simple
terms.

I am convinced that many of the concepts, proofs, techniques, processes,
methodologies we are currently using are already beyond (unassisted) human
abilities. 200 years ago a scientist could explain the whole of science to
any educated layman; 100 years ago some serious mathematics was needed in
many areas; 50 years ago specialisation restricted scientists largely to
their own fields and now, most advanced science is impossible without
computer assistance. It's getting harder and it's happening fast!

100 years from how, machines will be the only ones who (which?) understand
the most advanced theories. We shall be satisfied if we can pick up the
gist in Scientific American.

Regards
David Bennett
[ POWERflex Corporation     Developers of PFXplus ]
[ Tel:  +61-3-9888-5833     Fax:  +61-3-9888-5451 ]
[ E-mail: sales@...   support@... ]
[ Web home: www.pfxcorp.com   me: dmb@... ]

James Berge wrote,
> Lest the zeal with which we anticipate the future act as a
> soporific and we do not intsitute standards of responsible
> stewardship. Asimov was right and knew what we would be
> facing.

AFAIK, Asimov never considered the possibility of genetic programming and
evolvable machines which could compete against each other to reach higher levels
of IQ. With thousands (or millions and billions) of artificially intelligent
agents battling each other to reproduce, all humans would need to do is to cull
the herd, so to speak. Any unfriendly AI agents (unlike human children) could
simply be terminated. This would result in a population of AIs with docility and
compliance as part of their genetic code. Moravec's Mind Children could
obviously number in the millions from the start, because as soon as one is
developed, it could be duplicated ad infinitum. With an unlimited supply of
genetically programmed AIs, their evolution could be guided and directed as
experimenters see fit. The socialization of AI would consequently be far easier
than the socialization of humans.

--J. R.

J. R.;
I agree that we need to nurture the evolution of nonorganic
and eventually organic (non-human) computing platforms. The
ease with which they can accomplish the most laborious and
tedious tasks certainly helps me in my career.
I do believe that in light of the certainty of these (near)
future developments that the governments, companies and
educational institutions world-wide that are engaged in this
research and development enact as standard the Foresight
Institute 's suggestions on responsible development and
application of these technologies.

Lest the zeal with which we anticipate the future act as a
soporific and we do not intsitute standards of responsible
stewardship. Asimov was right and knew what we would be
facing.

------Original Message------
From: "J. R. Molloy" <jr@...>
To: <Artilect@egroups.com>
Sent: September 5, 2000 1:30:11 AM GMT
Subject: Re: [Artilect] Digest Number 49


> J.R.;
> I admit, that was funny if a bit chilling in the enormity
of
> its potential realization....
> James

As I see it, we should all do what we can to accelerate the
emergence of
greater-than-human AI, because superhuman intelligence (SI)
will help us to
answer questions and solve problems that have plagued
humanity for centuries.

--J. R.

"I cannot fear the night, for I have loved the stars."
--Astronomer's epitaph

Sincerely Yours,
James Berge
San Francisco, CA
415.551.9744 - direct/voicemail
415.430.2179 x1375 - fax
james.berge@... - email


______________________________________________
FREE Personalized Email at Mail.com
Sign up at http://www.mail.com/?sr=signup

*********** BEGIN FORWARDED MESSAGE  ***********
There are several advantages to powering a robot with ambient
vegetation.  The energy content of vegetation is relatively high -- plant
carbohydrates (sugars and starches) have about the same energy density
as lithium batteries; digestive mechanisms can be made light and
compact (as nature testifies); and vegetation is very widely
distributed in both time and space.  Several labs around the world
are exploring the option, which so far is understood as tapping
into the electron transport chain stimulated by the microbial
fermentation of some carbohydrate.  These electrons are then
organized around one end of the circuit, making in effect a
microbial fuel cell.  Machines competent to extract energy from
this source could live off the land unattended for years at a time.

Applications are not hard to imagine.   For the military, guns that
live in the forest; for consumers, gardening robots that rake, weed,
trim, and prune, and feed themselves with the results of their labor.
Environmental managers might have a need for autonomous mobile sensor
platforms that could both respond to requests to visit given sectors
and then establish a working presence in those sectors for long
periods. Perhaps the most ambitious application would be designing
ecological links as needed, such as a machine for detecting and eating
"foreign species" (however that is to be defined), or accelerating
nutrient cycles.

Such a device would need a sense of taste to make sure it did
not accidently eat something toxic to its fermenters;  the problem of
long-term autonomous maintenance remains unsolved; some policy needs
to be worked out wrt robot poop.  However, there do not seem to be any
shoestoppers.  See www.gastrobots.com.


*********** END FORWARDED MESSAGE  ***********

> J.R.;
> I admit, that was funny if a bit chilling in the enormity of
> its potential realization....
> James

As I see it, we should all do what we can to accelerate the emergence of
greater-than-human AI, because superhuman intelligence (SI) will help us to
answer questions and solve problems that have plagued humanity for centuries.

--J. R.

"I cannot fear the night, for I have loved the stars."
--Astronomer's epitaph

David Bennett patiently pointed out,
> It wasn't a joke. It was a SciFi short story, published maybe 30 years ago,
> by a well-known author (whose name I now forget). It was very short: 2-3
> pages.

Thank you for that piece of information. I've been repeating this little story
for decades, and always thought of it as a gag. I first heard it in San
Francisco in the late sixties. Nice to know where it first originated.

Cheers,

--J. R.

"I can't understand why people are frightened of new ideas.
I'm frightened of the old ones."       --John Cage