<nettime> interview with Daniela Plewe

[Josephine Bosma <jesis@xs4all.nl>]

Daniela Plewe is an artist from Berlin. She studied philosophy,
literature, anthropology and video in Paris. We met at the Machine
Aesthetics workshop that was organised by V2 in the summer of '97,
where Daniela Plewe presented and discussed her work 'Ultima Ratio'.
We got into a bit of a discussion over it, as 'Ultima Ratio' is
supposed to be a dilemma solving machine. I thought building such a
machine is in essence ethically not a very healthy thing to do, and
as I think a lot about the fringes of art and ethics due to certain
developments in communication art/net.art, Daniela and I had a rather
deep discussion about ethical and artistic aspects of her work. Is art
always untouchable, more so then science, as it explores and uses
all possible aspects of the new technologies? Is the pure existence
of this artwork enough discussion of certain unpleasant excesses?
Maybe just see this interview as one of the quests for answers that
might never come.



Q: Your first exhibition was at Elektro in Berlin, which was kind
of Berlin underground, were you a part of them?

Daniela Plewe: No, but I was familiar with the people. I knew the Lux
Logos people and there was an exhibition before Elektro, called
Electronic Art Syndrom in '92. This is were I mostly met the electronic
art scene. That was not my first exhibition though. I did in 87 in
Leuven a very early interactive piece. It was a work together with
other people.

Q: Did you make machines from the beginning on?

Daniela Plewe: The last piece I was talking about was called
'the Electronic Sunset' and it had an interactive structure.
It was a light dimmed by the visitors coming and it was triggered
by a light sluice. I wouldn't call it a machine, but it had some
features of interactivity. I also made multiples out of acrilic,
little sculptures. Then going on from that to video and photography,
mostly in Paris, were I studied with Paul Fasier, not for a very
long time and not with very good french.. Then I started to do the
first program, called 'Satz Ersatz' in '92 again. Its the piece that
was shown at ELectronic Art Syndrom.

Q: You seem not so much somebody that is into the 'matter', into
the hardware, but very much into the poetics of the artpiece.
'Electronic Sunset' is a rather poetic title, last year at DEAF you
made this Elctronic Daydreaming Device and your new plan is a machine
to solve dilemma's. They all are very close to human emotions, one
can easily access them on an emotional level, is that important for
you?

Daniela Plewe: I think you are right, its true I am not coming from
the hardware, eventhough I had some aquaintance with people who were
very much into hardware. My first video installations were in the
machine art surroundings, like with BBM, people who would work with
Bastiaan Maris and Erik Hobijn. I was familiar with this idea and I
also was inspired by some of their approaches, especially this notion
of disfunctionality of machines.

Myself, I am more interested in maybe the issues of representation and
introspection and everything about the mind, the cognitive interests.
For me working with computers and machines has the direction of a
knowledge representation in the widest sense, so the mental aspect
is very prominent.
And within this of course also the emotional is important. If I would
have to stress it now I would say: my main interest is methodology.
I'm very interested in how people deal with the world or how do they
deal with life or the problems of life and everything we are faced
with and what are the methods which our culture offers us to cope
with problems. Doing machine art or the medium computer as a so called
universal machine (whatever that means) for me is very close to this
interest of implementing methodology and also reflecting methodology.
We talked about that at the presentation/discussion in V2, of course
with the hope that we could make a statement about what is the usual
case of methodology here and then have some sort of poetic 'turn' or
abuse or distortion of what is the normal case of method in the piece
of art.

Q: You have chosen a very specific medium to investigate this
methodology. Talking about computers as knowledge machines, the way
they handle knowledge of course is very limited when you compare it
to the human brain. In the workshop there were a lot of people comparing
computers to brains or vice versa, but I think we are far from or
beyond making that comparison. What you actually do is make a very
limited choice of which sides of methodology to use and how to handle
it. Are you aware of this?

Daniela Plewe: Yes. Always when you work you are very aware of this.
I think its not a problem since I don't really feel very unhappy with
that I can impliment in the best case only very limited aspects like
for example this new formalism of Ultima Ratio.
It is very limited and it can only represent, which is actually a lot,
making in a limited syntax rational decisions, and they are fairly
complex. For example: I could not easily extend this formalism ad hoc
to maybe incorporate preferences and valuesystems, which would be
interesting in this context. If one could implement preferences of
goals, for example 'being happy', some people would say: to be happy
we want money and security and a family. Then you have sub-goals which
you reach by different steps. Now I have this problem: I can't
easily bring two formalisms together, which would cover that. It is a
limitation. I think it is for the recipient an even bigger problem. When
the machine is ready they ask them to do this and that...

The problem is people expect from machines much more if they don't know
how they actually work, while somebody that works with them is usually
much more modest then people that come by and say: why can't this machine
do anything we do? Its a problem we have nowadays, but science will
continue and...I am not naively optimistic about everything aritifial
intelligence will do, but it is still interesting to take it into art
and reflect it as far as possible.

Q: You say people are disappointed when things don't work the way they
expect them to work. At the same time people look up to machines and
are very easily capable of giving away responsibilities and leave them
to a machine, because the human being is lazy. So if you make a machine
that 'can solve dilemma's', in any kind of limited way, of course some
people might like to use it, without questioning it, without seeing the
doubleness of your poetic statement. Don't you think that is a danger?

Daniela Plewe: A danger would be that they just delegate decisions,
that they would rather trust the machines then trust themselves or so?
That' an interesting question. I think the machine will be so rough
that this will not happen so easily. From a practical point of view
I am not really afraid of this possibility. In principle I still
believe its a time phenomenon now. I think our overestimation of the
capabilities of machines is still around because of not understanding.
Of course it is dangerous if people...these auto-pilotes come to mind
now, the case where people rather trust the auto-pilote then the real
pilot. Actually I don't really really know what to say. I think the
human can fail and the machine can fail. I have problems to decide
there. Again about the filosofical question: no argument has really
convinced me that the human is not a machine as well, in a different
sense. A human can also make failures. That's the more filosofical
answer.

Q: The problem is when you start comparing humans to machines,
eventhough it is always possible (I am a materialist in this respect,
everything that happens has to do with some chemical processes, be it
in an interactive way between people or only inside the body)
bodies/humans are so much more complex then any machine at this moment,
that you cannot call them machinic.

Daniela Plewe: I agree completely, but its a matter of in principle.
In principle you can't exclude it that the human is something like a
machine. The second thing is will we ever be able to make it so
explicit, that you can describe it like a scientific theory. I am not
optimistic about this second condition. For me all the arguments that
say there is a difference between humans and machines were never really
convincing.

This whole discussion that is held everywhere and always pops up is
not so fruitful in my opinion, neither for understanding machine
aesthetics, nor to deal with machines here and there. I am not a
fan of this discussion anymore.

Q: But when you develop your machine, you're talking about artificial
intelligence, you call it artificial intelligence (AI). Does that
mean you work with neural networks, does that mean the machine learn
some things itself?

Daniela Plewe: I said that the formalism is actually coming from
the research field, which is called artificial intelligence. In the
empathetic way of AI, as you know it's a problematic definition: you
have to define intelligence etcetera. It is not workig with neural
networks, but since the eighties there are two paradigms in artificial
intelligence. One is sort of the neural or sub-symbolic and the other
one is symbolics. Neural networks belong to the sub-symbolic, mine is
working in the symbolic field. Learning can be modelled in both fields.
Neural networks work in a very different way then symbolic systems can
learn. I am thinking of how to get some self organisational features
in this project. If you call this learning, I would be happy if it were
learning. It would be learning in a reduced sense, it's not learning
something absolutely new I guess.

Q: I am sorry to nag on about some of the finer details you let
your machine do, for instance the last two projects: daydreaming
and solving dilemma's. There are some very important things missing
in the way the machines approach their 'jobs': coincidence, the
influence of the environment on a person, the state of being at that
certain moment in time of this person and of course coincidence in
time which a machine does not know. That is why I asked you whether
you work with neural networks because in that way these things could
maybe develop in.

Daniela Plewe: There is nothing accidental in neural networks, its
only that even the people that build them don't really have a theory
how to explain what happens exactly. Like Nick Baginsky also said:
"Its like a black box still." With some things you kind of get the
right output, but you don't know exactly what happens. The thing about
the branch I am working with is the symbolic part. It is very
transparant, every step you take is defined in the logic system, in
the explicit symbolic thing. You have different problems, like how do
you get their poesis in.

Now I don't believe that arbitrary things or random functions are
the solutions for computer arts to get into the poetics, because it
has been quite a tradition to do that and to thereby capture something
like nature, something that is not so appearantly deterministic, even
when if you look at it precise it is somehow generating and therefore
predictable. As I said: I don't have a solution, but I don't think the
simple implementation of random functions in a computerpiece (and I am
actually quite polemic about this) for me is not the way and maybe not
in general.

In Musers Service also I tried to do the opposite: I forced the user to
introspect on this process: how he got from the one idea to the next
idea. If he would have an association like 'love' and 'hate', he would
have to think about what is actually the relationship. "What kind of
operation did I do when I came from 'love' to 'hate'?" I would offer
them a very simple list with some options. One was maybe contradiction,
so you could classify this relation love/hate as a contradiction. I am
just saying that because I use now this knowledge again, which I call
sort of a methodological reflection, to make these chains of
associations. You could give in a preference, saying:" I want a chain
were there are mainly contradictions for example and only a few
disjunctions, which was another option. You would gain these chains of
assciations from musers service.

I am just saying that because there I avoided to use randomness. I
could of course have it build up chains between two words at random.
I avoided that by saying:"Lets define it exactly, lets investigate the
methodology, lets go to introspection rather then randomness.

On the panel we were talking about: how could one create poesis? I am
now thinking in this direction: with colours for instance you could have
functions and set relations between the functions. A painter is setting
colours to eachother and is having there his oppositions or his
contrasts and in that some kind of poesis or effect is created. Maybe
with combining functions one could also create some poesis, which is
hopefully more then each single piece.
This is how I try to create something that is not so painfully
deterministic, of which we have problems saying: this is art.

Q: Did your machines give you any new views? Were you ever surprised
 by your machines in such a way that you could say: "Yes, this is
'machine aesthetics'?

Daniela Plewe: I don't know what machine aesthetics exactly is but
I was of course always surprised by each piece. 'Musers' Service' for
instance makes chains of associations. I remember one stage where I
had entered my personal database, only entries of my associations,
and then the machine would connect the things in a way that I was
very happy I wouldn't think like that. Eventhough it was
all my knowledge, I was glad I would not function so awkward.

'Satz Ersatz', a project before this, was literature on a screen. It
was a very simple program which substituted simultaneously and
synonymous all words. Synonyms are very important, or meaning
preserving I call it. House..building..these symantic fields were
going through the texts, and these words would be substituted. This
enabled you to do experiments like having a story told in slang,
after which it would be translated into the language of the law,
of the courtroom. The story would be the same, but the language would
be altered. It is chopping and changing all the time, funny to see...

Q: But do you think there is such a thing as machine aesthetics, an
aesthetics that we do not recognise immediately as aesthetics, but
which is developed purely by the machine itself?

Daniela Plewe: I think there will be something, even if we have
problems right now to pin it down to some criteria. I guess that for
example the notion of functionality will be a central criterium.
You will see art pieces, which actually, since it's a 'prozentorale
maschine', artpieces will be a certain function of some machine.
For me this will be even more important then the notion of
interactivity. This is for my taste overemphasised, maybe because
I am interested in computer-arts in the real literal sense, rather
then in multi media or communication projects. I am very interested
in the medium computer and that is also why I forced myself to study
logics. For me the language of the computer is somehow logics.

On the hardware level I would love to have the possibility that
there is the rational machine, the rational hardware. That it is
rational itself. Could one build irrational hardware, I asked a
friend of mine, but he said it was bullshit. So I guess its not
possible. On the software I am also interested in logics, because
I believe that is the machinic language. That is not primarily
images and sound and cd-rom or whatever..all these strange notions
which are around and which people so much connect to computerart,
like hypermedia or multimedia. Its understandable this is now fashion,
but its not really computeraesthetics. Its just old media put into it,
sort of combined, but its not computer adequate aesthetics.

Q: Do you think there is too much focus on new developments of
toys and hypes then exploring the possible content of each layer more?

Daniela Plewe: I am not in the position to give a normative statement.
I see these tendencies, and for myself I say I am not interested in
it and that's enough. I will not judge it. 'Content' sounds like
communication projects again. That is not what I am aiming at either.

Q: You have no interest at all of doing anything in a communication
medium like the internet?

Daniela Plewe: 'Musers' service' was as a prototype not in the
internet, but then since it is an open database which wants to grow,
I thought it would be nice to impliment it on the internet and
follow how people from remote places let to grow. So this was in
a sense a communication project, because it took the entries of
all the people and did something with them. The same will happen
with 'Ultima Ratio', where in the ideal case, people also will enter
their data, perhaps even through the internet they can gain, the data
base can still grow. There will be a communication situation, where
discourse is taking place, but what I am doing is automisation of
discourse, which is in a sense transhuman. Its not emotional, its
more complete since it will consider all the pro's and cons, which
as I said we can't always do. So I see my projects in a sense as
communication projects, but abstraction of communication rather then
haing somebody here and somebody there and exchanging natural language
emails or working on something. It is actually about communication,
eventhough Ultima Ratio has, since its about decision making, that
you can see it eiter as an individual decision making process, like
your inner monologue: "should I kill or shouldn't I?". You can however
also see it as collective decisionmaking. It has a lot of
political implications about democracy and whatever the notion
argumentation implies, since that is sort of the democratic basis
as we were told in school. We will have to make a statement on that
also.

Q: Where will you be presenting Ultima Ratio and when?

Daniela Plewe: It must be created and I am at the stage where I sort
of gather the people who work on it and the financial situation is not
too bad, but we haven't worked on it. There is nothing yet planned.

Q: What is the Method Mutant?

Daniela Plewe: Its another aspect or maybe just an addition of what
we talked about before: media aesthetics. For me the artist is somebody
that should mutate the methods of how we live. I think within his own
life probaly he mutates the normal lifestyle. For example many artists
don't follow this rule that one should care about money a lot, so they
just neglect this issue and are very unhappy from time to time. I see
that as a kind of mutation which of course has a chance of some very
good offspring. Also in art one could see it as a kind of mutation.
Artpieces give statements to culture and alter certain ways how culture
functions.

Q: That is the artist as the Method Mutant.

Daniela Plewe: The artist in his life is a kind of Method Mutant, but
also the artpiece is a mutation of methodology. I understand
computerart, or my personal computer art as method mutation.


plewe@freebse.contrib.de
plewe@is.in-berlin.de

http://www.icf.de/CAC/Artists/b/a/.start.html
(this url will expire soon, as icf will diappear,
contact Daniela for further information)

http://www.kbs.uni-hannover.de/~schroede/abstracts/sch_aa98.html

http://www.v2.nl/events/machineaesthetics/projects/plewe.html


*



---
#  distributed via nettime-l : no commercial use without permission
#  <nettime> is a closed moderated mailinglist for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@icf.de and "info nettime" in the msg body
#  URL: http://www.desk.nl/~nettime/  contact: nettime-owner@icf.de