[Synapse elist] elist Digest, Vol 4, Issue 2
Paul Brown
paul at paul-brown.com
Fri Apr 4 12:23:14 CST 2008
Hi Greg - good to hear from you again.
These are important questions.
First a context - the robots we are working with may have 20-30
"neurons" in their "brain". Some very primitive worms have brains
this simple - an ant has 100,000 neurons, a wasp 250,000 and a human
about 16,000,000,000! So, to expect a robot like this to be able to
make a work of art that might compare with a human-made work is
hopelessly optimistic and not really the point. So it's a case of
small steps. Can we create an automaton that is capable of
exhibiting interesting behaviour. For example when it's drawing
recognising the crossing of another line and modifying it's behaviour
accordingly?
So, yes - the "judgement" or "value" issue. This is another
quagmire! The idea of evolving "observers" who can make judgements
has been pioneered by people like Rob Saunders (Rob - are you
monitoring this discussion?). So we evolve two automata - one to do
stuff and another to evaluate whether the stuff is "interesting" or
"of value". We maybe get lucky and end up with an alien intelligence
- and one that we might not comprehend.
The history of the arts gives us good evidence that we, the humans,
are pretty awful at selecting value in contemporary arts - Bach,
Cezanne, Van Gogh, Satie... the list is long - were all dismissed as
amateurs during their lifetimes (and in the case of Bach for almost 2
centuries after). And look at the status of graffiti - London
transport recently destroyed a Banksy worth over £100,000. So how
can we be expected to recognise not just a contemporary work by a
human (which arguable is hard enough - just look at all the crap in
the worlds "leading" art galleries) but work by a possible alien?
So personally I put these issues to one side. My ambition for the
DrawBots project is to make a group of robots that can manifest co-
operative drawing behaviour that people will find "interesting" and
will be prepared to place in the context of an art exhibition space.
As an artist I see what I'm doing as part of a tradition going back
centuries and as a modernist I firmly believe that the present is
built on the foundations of the past. However for most of my career
most people in the mainstream artwork have not considered what I do
as art. They perceive me as some eccentric dabbler who doesn't
really "understand" what real art is. I see them as lost souls
following some cult of fashion and novelty - their head still stuck
in the 1960's and the paradigm of the "shock of the new" - the
mainstream artwork has always been hideously conservative.
So I'm not particularly concerned about big questions like "is it
art" but more with little questions like "is it interesting" and
"does it have the potential to be taken further".
Gosh, I really must get on with doing my accounts! The Tax Office
are threatening to fine me!
Paul
On 4 Apr 2008, at 09:47, Greg Hooper wrote:
> Hi everyone, apologies if I am a little naive with respect to this
> field, but I wish to learn more about it and hopefully participate in
> the future. If art is going to have a distinctive meaning then there
> must be some criterion for judging something as art as against not-art
> (or a little bit art or slightly art or....). If machines make art
> then who or what is going to make the art classification. If it is
> people/humans who make that judgement then the human is not out of the
> loop. If humans are excluded from making the art judgement then there
> are no criteria from which to judge the actions or output of the
> machines. All outputs may or may not be art from the robot's viewpoint
> - our view is irrelevent. However, if the machines have some sort of
> value system built in by which they judge whether an output is art,
> then that is again letting in the human. The only way to remove the
> human and allow the robot autonomy would be to give the robot some
> sort of adaptive/ evolving value system. But we would be quite limited
> in our ability to recognise the robot response as being 'satisfaction'
> or 'self awareness' of their robotic output as art.
> This is the general problem of recognition of 'alien intelligence'. If
> something really is alien how close can we get to understanding it?
> How do we make our understanding of 'the alien' closer than a
> redescription of observable behaviour?
>
>
>
> On Fri, Apr 4, 2008 at 7:18 AM, wade marynowsky
> <wadox at laudanum.net> wrote:
>> Hello list
>>
>> I am interested in discussing the two points in inverted comers by
>> Leonel and Paul
>>
>> taking the "human out of the loop". and art that can "make itself".
>>
>> to what extent has this ever happened? as it is always the creators
>> rules within the system which governs the work
>>
>> maybe complex systems like CA etc can express this ..........
>>
>> my robotic art blog is available at
>> http://marynowsky.wordpress.com/
>>
>> regards
>> wade
>>
>>
>>
>> On 02/04/2008, at 1:30 PM, elist-request at synapse.net.au wrote:
>>
>>> Send elist mailing list submissions to
>>> elist at synapse.net.au
>>>
>>> To subscribe or unsubscribe via the World Wide Web, visit
>>> http://lists.synapse.net.au/mailman/listinfo/elist
>>> or, via email, send a message with subject or body 'help' to
>>> elist-request at synapse.net.au
>>>
>>> You can reach the person managing the list at
>>> elist-owner at synapse.net.au
>>>
>>> When replying, please edit your Subject line so it is more specific
>>> than "Re: Contents of elist digest..."
>>>
>>>
>>> Today's Topics:
>>>
>>> 1. Re: Bioart + Robotics (Paul Brown)
>>> 2. Artificial Creativity (Leonel Moura)
>>> 3. Re: Bioart + Robotics (kirsty boyle)
>>>
>>>
>>> --------------------------------------------------------------------
>>> --
>>>
>>> Message: 1
>>> Date: Tue, 1 Apr 2008 12:49:39 +1000
>>> From: Paul Brown <paul.brown.art.technology at gmail.com>
>>> Subject: Re: [Synapse elist] Bioart + Robotics
>>> To: elist at synapse.net.au
>>> Message-ID: <80A95984-D619-45BA-B6E3-E678404C14C9 at gmail.com>
>>> Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
>>>
>>> Hi everyone - I first trained as a painter at art school in
>>> Manchester, UK in the mid 1960's when I saw an exhibition at the ICA
>>> in London that changed my life!. It was of course Cybernetic
>>> Serendipity in 1968 and, from that moment on I was hooked. From
>>> 1974
>>> my practice has been based almost exclusively in the computational
>>> domain.
>>>
>>> Another major influence was an issue of Scientific American (October
>>> 70) when Martin Gardner's Mathematical Recreations column featured
>>> John Conway's cellular automaton - The Game of Life.
>>>
>>> I went back to university in the 1970's to learn about computers and
>>> was lucky to get a place at the Slade School of Art, University
>>> College London when it had the pioneering Experimental Dept. with
>>> it's own minicomputer (a 32kb system that was larger than a fridge
>>> and cost as much as a house). The Slade postgraduate school was
>>> dominated by the Systems Group who were involved with ideas of
>>> art as
>>> process and autonomy.
>>>
>>> Since then I have been developing ideas about art that can "make
>>> itself". From the earliest days I included cellular automata in my
>>> work.
>>>
>>> In 2000 I was artist-in-residence at the Centre for Computational
>>> Neuroscience and Robotics (CCNR) at the University of Sussex
>>> (where I
>>> now have a visiting position). More recently I have been working
>>> together with colleague at the CCNR on a project that is attempting
>>> to use evolutionary robotics to create robots that can draw.
>>>
>>> So, that's me! I'm looking forward to this discussion and would
>>> like
>>> to thank Vicki for setting it up and inviting me.
>>>
>>> Paul
>>>
>>> ====
>>> Paul Brown - based in OZ Dec 07 - Apr 08
>>> mailto:paul at paul-brown.com == http://www.paul-brown.com
>>> OZ Landline +61 (0)7 5443 3491 == USA fax +1 309 216 9900
>>> OZ Mobile +61 (0)419 72 74 85 == Skype paul-g-brown
>>> ====
>>> Visiting Professor - Sussex University
>>> http://www.cogs.susx.ac.uk/ccnr/research/creativity.html
>>> ====
>>>
>>>
>>>
>>>
>>>
>>> ------------------------------
>>>
>>> Message: 2
>>> Date: Tue, 01 Apr 2008 10:17:05 +0100
>>> From: Leonel Moura <leonel.moura at mail.telepac.pt>
>>> Subject: [Synapse elist] Artificial Creativity
>>> To: elist at synapse.net.au
>>> Message-ID: <mailman.2.1207103402.6200.elist at synapse.net.au>
>>> Content-Type: text/plain; charset="us-ascii"; format=flowed
>>>
>>>
>>>
>>> Hi all,
>>> since I have started working with cellular automata, ant algorithms,
>>> swarm intelligence
>>> and later robotics my main purpose was to produce an Artificial
>>> Creativity (AC)
>>> in a similar sense as we talk about an Artificial Intelligence.
>>> That is the capacity of a machine or a robot to generate an original
>>> "aesthetical expression" with a minimum of human intervention.
>>>
>>> Robots are for that matter one of the best choices, as they
>>> evolve in
>>> the physical world,
>>> and, more important, they can gather directly from the environment
>>> the data needed
>>> to perform such a task. This ability helps to take the "human out of
>>> the loop".
>>> Critical condition if we want to speak about true AC or even true
>>> AI.
>>>
>>> In this perspective RAP (the Robotic Action Painter) is a good
>>> example:
>>> <http://www.leonelmoura.com/rap.html>http<http://
>>> www.leonelmoura.com/rap.html>://www.leonelmoura.com/rap.html
>>>
>>>
>>> Leonel
>>> http://www.leonelmoura.com/
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> ------------------------------
>>>
>>> Message: 3
>>> Date: Wed, 2 Apr 2008 02:20:04 +0200
>>> From: "kirsty boyle" <kirsty at karakuri.info>
>>> Subject: Re: [Synapse elist] Bioart + Robotics
>>> To: elist at synapse.net.au
>>> Message-ID:
>>> <da2393170804011720y52992335s767561160ae3cf08 at mail.gmail.com>
>>> Content-Type: text/plain; charset=ISO-8859-1
>>>
>>> Hi,
>>> I originally trained as a puppeteer.. but always held down day jobs
>>> working with technology. I spent 4 years working at telecom/telstra
>>> automating and configuring telecommunication networks. I then
>>> worked
>>> in animation and film. I studied electrical and mechanical
>>> engineering (mechatronics/robotics), but found it quite hard and
>>> depressing as the focus was on the automotive industry. For many
>>> years I worked at the ABC in engineering, figuring it was a good
>>> place
>>> to be to use my technological knowledge in a creative environment.
>>>
>>> In 2002 I had an opportunity to travel to Japan and study with
>>> Master
>>> Tamaya SHOBEI, a Karakuri craftsman. Karakuri is a mechanical doll
>>> making tradition which can be seen as the origin of the current day
>>> robotics industry in Japan. I was motivated to meet him as I had
>>> heard about the tradition but there was little written about it in
>>> English. I have a website which is an attempt to document the
>>> craft -
>>> http://www.karakuri.info.
>>>
>>> Whilst in Japan I also started studying with Professor Yoshikazu
>>> SUEMATSU, who founded the Suematsu Robotics Laboratory at Nagoya
>>> University. He is considered the world's foremost authority on
>>> Karakuri, and was a pioneer of foveated robotic vision systems. He
>>> was also a student of Masahiro MORI (more of which I will write
>>> about
>>> below).
>>>
>>> I thought after initially spending time in Japan I would return
>>> and be
>>> making my own mechanical dolls, but instead it has taken me on a
>>> completely different path. I have continued studying the tradition
>>> and documenting it so as to share the knowledge with people
>>> outside of
>>> Japan. I currently have only a small amount of information up on my
>>> website but it astounds me how much interest this has created. I
>>> often get frustrated thinking about how just because something
>>> hasn't
>>> been expressed in English it might not actually exist. Currently
>>> I am
>>> Master SHOBEI's only student, which never ceases to amaze me.
>>>
>>> At the moment I am an artist in residence and visiting researcher at
>>> the AI Lab, University of Zurich - http://www.ifi.uzh.ch/ailab/ I
>>> was invited here by Professor Rolf Pfiefer, who featured my
>>> website in
>>> the AI Lectures from Tokyo, considered the first global, fully
>>> interactive, videoconferencing-based lecture series at the
>>> University
>>> of Tokyo - http://tokyolectures.org/
>>>
>>> At the lab I am building an embodied robot inspired by Karakuri. I
>>> have a blog about my ANAT supported residency here -
>>> http://www.anat.org.au/blog/boyle/
>>>
>>> One topic I am hoping to raise for discussion is as artists, how we
>>> can we shape the future aesthetic appearance/movement of robots? Or
>>> how do we want our robots to look and act?
>>>
>>> For many years I have been studying the 'Uncanny Valley' theory
>>> proposed by Masahiro MORI (original paper in English is here -
>>> http://www.androidscience.com/theuncannyvalley/proceedings2005/
>>> uncannyvalley.html)
>>> I am keen for any comments relating to the Uncanny Valley as I am
>>> preparing to speak on a panel themed around Virtual Humans at FMX
>>> next
>>> month - http://www.fmx.de.
>>>
>>> And just out of interest - what is your favourite robot, and why?
>>>
>>> Many thanks,
>>> Kirsty
>>>
>>>
>>> ------------------------------
>>>
>>> _______________________________________________
>>> elist mailing list
>>> elist at synapse.net.au
>>> http://lists.synapse.net.au/mailman/listinfo/elist
>>>
>>>
>>> End of elist Digest, Vol 4, Issue 2
>>> ***********************************
>>
>> _______________________________________________
>> elist mailing list
>> elist at synapse.net.au
>> http://lists.synapse.net.au/mailman/listinfo/elist
>>
> _______________________________________________
> elist mailing list
> elist at synapse.net.au
> http://lists.synapse.net.au/mailman/listinfo/elist
====
Paul Brown - based in OZ Dec 07 - Apr 08
mailto:paul at paul-brown.com == http://www.paul-brown.com
OZ Landline +61 (0)7 5443 3491 == USA fax +1 309 216 9900
OZ Mobile +61 (0)419 72 74 85 == Skype paul-g-brown
====
Visiting Professor - Sussex University
http://www.cogs.susx.ac.uk/ccnr/research/creativity.html
====
More information about the eList
mailing list