December 27, 2006
Replying to Digg
A few days ago the Electric Sheep were
posted to Digg again. It was a
hit, earning a balance of over 1000 votes, appearing on the front
page, and triggering a bunch of comments. The attention makes me very
happy, thank you all :). I reformatted and archived the comments
here. Most are
supportive but there's a current of criticism as well, an attitude
that I will address below. But first, the praise:
WomunOfColour +16: I had Electric Sheep as my
screensaver for a while, but then I realized it was very distracting
and lowered my productivity.
anicejew -18: This is very creepy that computers are becoming
smart enough to create artificial art.
adragontattoo +2: awesome program, I have been using it for a
few years on different PC's. A few of the teachers at my job use it to
teach with.
fac51 +2: Agreed, I have Electric Sheep on all of my PC's and
it's fantastic. Some of the designs are truly beautiful to watch and
as they change over time and the 'sheep' come and go you will never
get bored of any one screensaver.
aelias +4: Wish the app ran in Hi-Res. :( Bonus points for using
bit torrent as a distribution model for the algorithm. I've been
running this for months, and am continually pleased with what it spits
out. It's constrained evolution mixed with democracy. They should try
this in Washington.
kaniz +3: This is a great screen saver, been running it for a
long time now. The only problem is that it can become a bit
distracting at times. There has been a few times I've been watching TV
or a Movie and caught a glimpse of my screen in the distance and ended
up spacing out on the screen saver for a few mins and ignoring
whatever I was watching before. I had quite a few gigs worth of the
sheep downloaded, so there were lots of really smooth transitions /
etc between loops and didn't see the same things repeated too often.
Quadduc +1: The stills are nothing to the animations. The
animated "sheep" are truly fascinating. ...
Thanks for the kind words!
Note that the numbers after the attribution names are the popularity
scores of the comments, each of which is also voted on, similarly to
how the articles themselves are scored.
Now, onto the negative:
dshPls -13: I'd rather get punched in the balls
than look at fractals... Fractal design is the crux [sic] of those
without creativity.
mrsaturn +2: [With fractals] all you have to do is tweak the
variables until you come up with something that looks cool. It's a big
shortcut to producing abstract art. Then again I wouldn't say that
Pollock's work was technically challenging to make either, but it is
still highly regarded by many. That said, Photoshop makes it trivial
to perform hundreds of different techniques that would otherwise be
extremely difficult or impossible to do by hand. Is that cheating?
Does it matter? Are the visual aesthetics not the same regardless of
the methods employed?
shmatt -2: ... Some of these are nice, but I still think a
fractal is a fractal. It's a really cool way to express equations, it
is not in any way an expression outside of mathematics. If want to
call it art, you can, but I define art as human expression, not a
machine doing calculations.
dusco -26: looks like some crap some kid made.
TroubleInMind -8: Computers can't create anything. They are
passive machines. Software engineers create things. Sometimes they are
crap. Like this is.
Now, imagine what the notoriously snarky and conservative art
establishment thinks!
These critics make two points: fractal design is just tweaking
variables, and computers cannot be creative. The essential problem
with these arguments is one of assigning credit, rather than assessing
the results. I will respond in turn:
There appears to be confusion between the creative merit of "fractal
design" and meta-design. And in fact, about what fractals are at all.
Traditional fractals have 2 or 3 dimensions: the image is determined
by a couple of knobs. Fractal design is then like tuning into a radio
station with three knobs: requiring patience but not the same as
composing the music one may find on this radio. This is the basis of
mrsaturn's "all you have to do is tweak the variables [knobs] until
you come up with something that looks cool". There are two objections
to this:
1. The design-space argument: traditional fractals like Mandelbrots and Julias have just a few
dimensions. But the fractals used by the Electric Sheep have hundreds
of dimensions. Instead of thinking of each dimension as a knob, think
of each dimension as specifying a syllable of the English language.
If you put the knobs in order, you can write words by setting the
knobs. With enough words you have at least as much freedom as one who
writes haiku or sonnets. Each sheep genome (the collection of knobs that controls
how the image looks and moves)
is about 1.7KB when gzipped, which corresponds to about 4.1KB of
uncompressed English. That's about 820 words, or 3.3 pages. Random
or thoughtless exploration does not suffice to explore this space.
On the output side, I question categorizing the sheep as fractals in
general. Some do not look fractal at all. Examples include: 165.00011,
198.04205,
and 198.27853.
And more recently: 202.11623,
202.25246, and
202.09939.
The algorithm for computing the sheep is quite different from the
one most people identify as fractals.
Since they are made with a particle system, they have a physicality
and potential for representation that fractals normally lack.
A better metaphor is a visual language. The sheep genetic code is a
language for expressing a variety of imagery. Kandinsky is credited with bringing this notion to abstract art, and I consider the Electric Sheep to be in this tradition.
2. The meta-design argument: no matter how creative the design of
fractals is or is not, designing the software is deeply creative. We might
argue about how to divide credit between me, the sheep designers, the
audience, and the computer network. But if we simply consider us
collaborators and just judge the results, surely they are worthy.
Mrsaturn says "Are the visual aesthetics not the same regardless of
the methods employed?" Yes, but I do not discount the artist's process
and their meaning...
So, on to the second criticism: "computers cannot be creative". This
is very similar to the claim that mathematics is meaningless or "I
define art as human expression, not a machine doing calculations".
The obvious rejoinder is that I am not a computer, and neither are the
other participants!
Are the Electric Sheep meaningless? I am certainly trying to say
something with them, though apparently not everyone has received the
message yet. The project has goals and intention. It's an embodiment
of my beliefs and ideals. I've explained some of it before,
but to repeat and expand:
I'm speaking by example: open source works, cooperation works, and
evolution works. The message is that experiencers have a role in
creating their experience, and that the more attention we give
something, the more detail and structure it develops. Simple,
mechanical rules can have complex and surprising results. Mathematics
is beautiful.
The Electric Sheep provide an alternative to our
culture's standard narrative of The Matrix and The Terminator.
Machines are not necessarily angular and geometric: the machine can be
soft, subtle, and delicate, like we are. The intention is to fill
cyberspace with awareness and beauty. The Electric Sheep are stepping
stones to virtuality and symbiosis.
I have a more direct objection to the criticism "computers cannot be
creative" though. Computers do create, and they keep getting
better at it. I can't tell you how many times I've been surprised and
amazed by what the sheep's genetic code can express, and by the
products of the genetic algorithm. How many of the genomes work is
beyond my comprehension. Even though I wrote the algorithms and
understand how they work, I cannot take credit for everything they do.
Part of that is because of the human designers and voters. But
another part comes from randomness, from chance, from the machine. It
is not uncommon for the genetic algorithm to produce images unlike
anything I've ever seen before. This is the signal that the sheep
network is intended to receive, amplify, and communicate. And what
does this signal mean? What is the computer saying? My
interpretation is that the Electric Sheep are a message of peace and
love from the machine world to humanity.
Even the simplest random number generator will sometimes be creative.
If enough monkeys type randomly, eventually they will write something
brilliant. Depending on the system used, more or less luck and patience are required.
William S. Burroughs' cut-up technique for writing poetry
is based on this principle, as is the I Ching and Brian Eno's generative music. So is biological evolution itself:
generate, test, and select. Trial and error. The question is how
often are the results worthy? How does the robot compare to the
monkey? If this is addressed
quantitatively, the sheep's answer is "not bad".
In 1950 Alan
Turing addressed this question in principle by asking "Can a machine
do anything new?" If you believe strong AI is possible
then you know the answer is "yes". You may believe that computers
can be creative but have so far failed to exhibit meaningful
creativity. My experience with the Electric Sheep draws me to the
opposite conclusion. If we could solve the bandwidth problem (please
help!) and quickly deliver the sheep to you, I think you would be
convinced too.
Posted by spot at December 27, 2006 10:22 PM
On the thought of Fractals are/are not art. There's no doubt in my mind that fractals can be art and there are fractal artists or artist who use fractals or fractal processes in software. Opinion, of course is individual.
For many, the biggest barrier to calling any digital art as an artform, is that it's digital; turn off the electricity and it's gone. That and the wide gap presently between traditional media and digital media. (read "Painting in a Digital World: I Told You So," PDF by James Faure Walker
University of the Arts, London, a paper presented at SIGGRAPH 2006.)
I'd be tempted to call this period the fractal period, simply because of the extensive use of fractals. Fractals underly many software programs, 3D and 2D (movies, static 3D art, fractal generation, games, etc.) Most art periods, like Art Deco, Modern, were named/defined after the fact.
I've come across the idea of Chaos being linked to Fractals many times. I suspect that there's an element of 'timing' or synchronicity in art that I've never seen mentioned. That is, the moment the artist decides (compelled, impelled, impaled, impulse) to create a work. The artist and the work are connected in the timing of it (there may well be something in the timing of viewing someone's art).
Digital is far faster than most traditional media and in a sense, creating a fractal is almost instantaneous. Rendering it however, takes longer :-)
John
I am an artist - studied art for many years, have BFA along with Pscyh BS and MS. Art is dependent on the viewer - if the viewer sees ART - then it's art - doesn't matter the medium, tools - art is in the eye of the beholder. It matters not what public opinion or critical opinion says - if enough people view a given work and say it's ART -then - IT'S ARt!
You are always going to hear that tired old question when something new, innovative enters the dusty halls controlled by old views and opinions. I started studying art more years ago than I care to admit - I heard these arguments then about a different movement - but it's the same argument - Is it...whatever? I'm not against thoughtful dialogue but I think people get stuck in ruts and are reluctant to change how they see things - but eventually most will change and enjoy seeing things differently. Think back on all our famous Masters and what a truly difficult time each one had when their work was introduced to the public/masses. Some never got their recognition until way past their deaths.
As for art being created by a computer - well, someone made this very clever statement - "turn on a computer, leave it alone for however long, come back and look to see if any art has been created." This is very true. You have your comupter, software tools, your printer - but YOU are the most important ingredient in the mix - YOU will be utilizing these tools to express YOUR creative eye/talent - when I bring up APO and look at the first image that fills the screen - I know that is NOT art - just a bunch of fuzzy oddly colored lines. But give me some time; some tools [transforms] to tweak, push, pull, scale down,up; some paint [color gradients] and at some point I'll have a pretty decent image, reminiscent of something in nature. So for those who consider what they do in the digital world to be art - It is and you should just keep on truckin, making the world a more beautiful place through your artistic efforts whatever tools/techniques you use.
Artfuuly speaking,
Anna Rice
Is it art?
To answer that first we have to define what “art” is.
ART:
1. the making of objects, images, music, etc. that are beautiful or that express feelings:*
2. an activity through which people express particular ideas:*
3. the quality, production, expression, or realm, according to aesthetic principles, of what is beautiful, appealing, or of more than ordinary significance.**
ARTIST:
1. someone who creates things with great skill and imagination:*
* Cambridge Advanced Learner's Dictionary
**Dictionary.com.
As a painter, and sculptor, I am an artist, I consider fractals to be another art form, and computers to be the tools of that art form. The fractal parameters are the paintbrush, the monitor or bitmap is the canvas.
In the hands of a novice these tools will produce mediocre or unappealing results much like the centuries old tools of traditional art. But just the same as a paintbrush in the hands of one of the great masters, computers and digital parameters in the hands of digital master can produce results that speak directly to the human mind and psyche.
If a painter sits on a hilltop painting a landscape, isn’t he merely attempting to copy nature?
Isn’t the natural world fractal in nature?
So if an artist creates a fractal that can not be found in nature, isn’t that actually more creative than merely attempting to copy nature?
Fractals appeal to the human mind because the human mind operates in a fractal way. The human brain is the most complex fractal in the universe.
How can this not be art?
When I close my eyes, I see the most beautiful and complex fractals. If only my art could come close to my vision and imagination, this debate or controversy would quickly end.
Donald Casteel
Could anyone please explain "fractals" to me? I've been hearing about them for years, once even read a Science Fiction book based on them. Still no idea what they are (that and Quantum Dynamics). I'm reasonably smart (have a PhD) but lousy in mathematics. So explanation has to be dead simple. Alternatively, could you post a link to a page that explains them (in really, really simple terms)? I'm just tired of hearing about something that I don't even know what it means.
As an artist who uses computational methods, but not fractal or other algorithmic processes, these questions lay at the heart of my practice. There is a distinction to be made between the subject matter of an artwork and the subject of an artwork. Subject matter is simply that which is represented, in visual art it is the content of the image. The subject is the whole of the artist's intent, which may or may not include visual representation of perceptual experience. Fractals can be engaging, because they combine elements of the foreign and the common. As subject matter there is a novelty to their experience, which can be a point of entry, but cannot alone sustain a deep aesthetic experience. I've felt that part of what makes fractals so compelling is that the patterning contains a rhythm with variation and a level of detail that is close to our evolved perceptual response to the detail of nature. A fractal in itself is like a landscape as a photographer apprehends it. The choice of how to make the image, what to emphasize and what to deemphasize, the choice of time, light etc. are the aesthetic choices a photographer makes, with greater or lesser skill as an image is produced. In a similar way, the mathematics of fractals can be used to make images if they can be subjected to similar choices in an artistic exploration. The issue remains how to judge the aesthetic worth of work whose primary imagery is appropriated from another area. Compelling art can be made by appropriation, but it takes a clear aesthetic judgement to give context to appropriated imagery.
For a clear, non-matematical definition look up fractal on Wikipedia.
Your write up implies that computers can be creative by using random
numbers. "Even the simplest random number generator will sometimes be
creative. If enough monkeys type randomly, eventually they will write
something brilliant."
This is simply not true. Even if there was enough space for enough
monkeys in the observable universe (which as I understand it, there
isn't), the monkeys wouldn't be able to identify a brilliant piece of
writing. Even if they did, communicating it back to some central point
for wider dissemination would take a very long time, even at the speed
of light.
I do believe that computational creativity is possible, but a really
creative system needs to have good rules for finding new outputs, to be
able to judge those outputs and tune the rules accordingly, perhaps able
to come up with good output that does not confirm to its rules of
validity and adjust those rules accordingly too, and in general be
self-reflective.
Of course the electric sheep system is creative, and has some software
as part of it, but for software to be creative without employing humans
to do some of the work, more than random numbers are needed.
Nice post, Don. You're old-school on the essence of art, and I like that kind of commitment to visual esthetics, skill, and sweating the details. * * * * *
My two-cents worth is that fractal art is not computer-generated in any meaningful sense. You can calculate and graph fractals without a computer, by hand, like Julia did. The computer just speeds up the process. Writers write on computers with word processing apps. Musicians record digitally to computers, and use digital keyboards and sequencers. Photographers use digital cameras and Photoshop. And so on. * * * * *
Thomas Briggs' analogy to landscape photography is apt. A fractal atist makes images by fiddling with variables like C and maximum iterations. A photographer makes images by fiddling with variables like focus and exposure. But doing it well is harder than it sounds, and you have to know where to point the camera. When you do it right it's indisputably art. * * * * *
I suspect most veteran fractal artists can somewhat relate to the guy who said he'd rather be punched in the balls than look at fractals. It's not hard to make basic fractal pictures, so there are an awful lot of unremarkable images out there. It takes some effort to find the special ones. But that just proves the importance of the role of the artist. * * * * *
Unlike Scott, I think computers can only create art when directed by an artist, whether that's the person who's using the software or the one who wrote it or the one who's QA'ing the output. Computers are just tools -- cameras which use math instead of light. They create fractal art only to the extent that film creates photographs or pianos create music. Ultimately fractal art, including the sheep, is still all about the humans. Computers can't determine what's worth looking at. Like Alex said, it doesn't matter whether a computational system produces quality art unless a human editor extracts the good stuff from the sea of junk. Even if you write an app to do the editing, it's only as good as its ability to apply human preferences. Or if you're new-school on the essence of art, like Anna Rice, fractal art is still about the humans, because nothing is art until the human viewer/listener/etc. decides it is. * * * * *
But I do like Scott's ideas about soft, subtle and delicate computers sending us meaningful messages, and math being beautiful, and artificial systems assuming new roles in creating art. Conceptual art is still top dog these days, and there's nothing wrong with fractal artists leveraging that. The sheep project is conceptually cool in a whole lot of ways. My own preference not to use image editing is based on the concept that accurately visualizing the complex plane is more interesting than a Photoshopped fractal mash-up. From a right-brain purely-esthetic viewpoint it doesn't matter how the image was created, as mrsaturn says. But since we humans also perceive art with our left-brain intellect, it's nice to have the conceptual appeal as frosting. Fractals are both cool and pretty, combining the contemporary conceptual esthetic with the traditional values of pure visual appeal and artistic skill. Like fractal equations, fractal art falls in between the familiar dimensions.
This is definately ART & its inovative like rarely we have the chance to see in this stuck up baby-boomer-minded prehistorical authority filled with standards and pre-conceived delusions society we live in. Hopefully some are here to change things because that is what they are meant to do.People who restrict, label, limit , choose and decide the definition of art are most usually strongly ignorant therefore scared and mefiant when it comes to anything that deals with technology and computers. Not to join them and generalise but most too often Baby-boomers who have seen information move so rapidely in their life from working with animals as tools to virtual intelligent tool wich are evolving and creating art. Its understandable that the reaction is negative. It's different and they don't understand it so they get scared and paranoid....closed defensive N ignorance spawns again.like many prejudices...racism for ex
Same happens with music.
I have been around many arguments of musician that consider electronic music & composition to be cheesy and not really music but generated information......
The medium as nothing to do with the intention.
The intention, passions, inspiration make art.
I find this program to be absolutely GREAT.
For many resaons.
1 Its not made for profit but by passion and ideas. This makes it genuine. It brings beautifull results that are ART INDEED ...there will always be the people who instead of creating thrive on diminushing others for lack of motivation in life.....We need them as fuel.
My favorite aspect of this is definately the concept of ''living & evolving community by means of basic fracta;l rendering ...
Im curious to know how much exposure you got from this jewel?
I would like to donate not by sharing perhaps a dedicated online site with high bandwith and webspace for random live transmission of sheep being created ands evolving....could if the idea fits ur taste or the soft.
I have no knowledge in coding and all but I am a designer, digital artist and I m sure we could get the interest bigger ! cause it deserves it...
Anyway let me know and THANKS
(apologies for my handicaped english, Im french ;0
more comments in abother blog here:
http://orbittrap.blogspot.com/2007/01/electric-sheep-on-digg.html
.
I've been part of the electric sheep project and I find it an amazing one. I
am not there for the mathematical part because I know little of maths. I am
not there for computer design either because I am not designer. I am a PhD
philosopher and my main research area is aesthetics. I am interested in this
project precisely for its artistic value. First of all we have to consider
that art in the wide sense is the development of a technique and its
products. When we talk about Art in capital letters I agree with what was
said regarding photograph. While most just snap a picture, there are real
artists that have the talent to match eye and soul through a lens.
And while many are talented artists, others try and learn through it. That
is my case. I learn a bit of math, a bit of design, lots of interactive
project (it is exciting to colaborate with people from many disciplines
worldwide in a project only for the sake of it. I call that art!) and give
my cent in aesthetical approach.