jean_poole on Thu, 9 Feb 2006 10:48:20 +0100 (CET)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Cyborg Cockroach interview



--- or rather, an interview with Garnet Hertz, creator of the 
infamous 'Cockroach Controlled Mobile Robot'...
upped here >> http://www.skynoise.net/2006/02/08/cyborg-roach-man/
--
jp

At the University of California Irvine, 1x Garnet Hertz is ploughing his way
through a Visual Studies PhD program, with a particular focus on "the quest for
living machines". Aside from exhibiting a dead = frog with a miniature web server
embedded in it ( allowing web visitors to re-animate it's limbs! ), this quest
has more recently lead him to using a living roach atop a modified trackball to
control a 3 wheeled robot, infra-red sensors providing navigation feedback to the
cockroach, with the hope of creating a pseudo-intelligent system with the
cockroach as CPU! Visit www.conceptlab.com for much more info, knowing you'll be
in fine company ( one of Arnold Schwarzenegger's close aides was given a demo
recently).

 >>What attracts you to artificial intelligence?

It's more of a repulsion than an attraction.  The proponents of "strong" AI like
Kurzweil that believe that machines will (or already are) literally intelligent
makes me laugh.  Strong AI is a joke: it's a re-hashing of the age-old desire to
endow inanimate technology with life.  Yesterday it was idols made of stone:
today it's silicon.

"Soft" AI: that machines can solve difficult problems as expert systems is a
slightly different matter, though.  MapQuest wayfinding and other similar
technologies are quite amazing computer programs.  I may be off the mark here,
but I think it's an anthropomorphic stretch to think of these systems in the same
breath as carbon-based lifeforms.

AI is a catch-phrase that many undergraduate (and graduate) students in computer
science fling out to their parents to try to impress them when asked "What do you
want to study?".  In my estimations, few of these students have a clear
understanding about the field of AI - other than it's a holy grail and that
they're wanting to join the priestly caste to pursue it.

 >>How has the AI concept morphed over time, and do you still find it
useful?

I think the concept of AI has basically remained the same for several thousand
years.  I've thrown out the holy pursuit of AI with other religious
proclamations.

 >>What do you think of the proposal that we will grow, rather than
design, artificial intelligence?

Whose proposal is this?

 >>Much writing about AI gives me a kind of 'emperor has no clothes'
feeling. Does the building ( or growing) of an AI system that can fool a human /
pass the Turing test, while fun and impressive, necessarily mean we have created
intent / sentience / consciousness?

I won't really get into the Turing Test, but in my opinion, a written on-screen
conversation is only a minuscule portion of bandwidth compared to the full range
of human experience.  Any supercomputer in any era would very quickly fail the
Turing Test if you only followed the terminal's cables back to reveal a
supercomputer mainframe (or, perhaps, a hot-rodded C=3D64).  In other words, the
Turing Test as it's currently understood, only plays the game of intelligence
with conversation on a computer screen, and not in the full bandwidth of human
life, vision and sense perception.

Or, put differently by Rodney Brooks, humans spent millions and millions of years
just figuring out to locomote. Only a few thousand years have been spent on
figuring out language.  So, maybe the Turing Test has it all wrong and is
skipping over the harder "common sense" problem: how to walk.  In other words,
maybe a better test of intelligence is to stick the supercomputer mainframe on
one side of a freeway and get it to run across in heavy traffic without getting
hit.

 >>Wearing your Dorkbot hat ("People doing strange things with electricity" ),
what projects / developments point out interesting ways forward with energy
efficiency and the mammoth task of reducing our energy footprint in the 21st C?

There haven't really been a lot of demos on this topic, although Tom Jennings
(inventor of FidoNet and first webmaster for Wired Magazine, among many other
incredible feats) presented on converting his car to run on alternative fuels.

 >>And your Re-Animator hat - how complex a dead creature would you be prepared
to insert a webserver inside, for limb-animating website visitors?

I think I'm sort of done with the Damien Hirst line of work.

What sort of hopes do you have for the evolution of your (Giant Madagascan
Hissing) "Cockroach Controlled Mobile Robot"?

The main concept in this project is to see if I could get a robot to be somewhat
intelligently controlled by a cockroach.  In other words, take a mobile robot and
basically replace the CPU with an insect.  To get this to work "properly" isn't a
very easy task: biological "meat" is hard to predict.  I ended up making a robot
that is quite temperamental and non-functional, sort of like a lot of grad
students

I know.


( sidenote by jp: check spiders as remote controlled iraq surveillance comic at
electric sheep? by patrick farley? http://www.e-sheep.com/ )

What did you glean from the Australian Symbiotica workshop?

I think it's a good idea to have hands-on experience with something before
forming a firm opinion on it.  In the case of genetic and tissue engineering,
this workshop was a way that I could see part of what all the fuss was about. In
many ways, it's a lot less glamorous than I had initially expected.

Which ballpark concerns about biotechnology and genetic engineering, are
legitimate and misplaced?

That's a tough one, and it's something that I honestly don't spend a mountain of
time thinking about.  However, it's interesting to note that many American
consumers don't seem to be overly concerned about GMO produce in comparison to EU
folk.  Maybe the American fast (and pre-prepared) food industry paved a path for
consumers to be generally more accepting of food being highly processed... I'm
not sure.

 >>What technologies that mimic biology ( Biomimetic systems )
interest you at the moment?

None, really.

 >>What do swarms of insects have to teach us?

Time to load up on bug repellent.

 >>And finally, what fuel will our future robot-driving cock-roach
overlords be using?

Charred human flesh left over after we've all killed each other.

jean poole




#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime@bbs.thing.net