A rather interesting question followed by a rather epic first post.
I wrote this in university many moons ago. I like it’s randomness, and it’s applicability to Graphic Design. After all, the majority of graphic design is now designed using computers, although perhaps not by computers.
Whilst planning this essay I was listening to music, this morning I chose Cat Power as my background noise and I decided to play her on my computer using Windows Media Player. It was then that I noticed the beautiful graphics that Media Player has, the hypnotic swirling colours that move so cleverly in time with the music. I had noticed them before, but never really considered them. I decided to try and copy a stationary image to use as my desktop background, the only way I could think of to achieve this was to use ‘print screen’. But as I copied the image what I got was not a colourful, crisp image similar to those produced by media player, but a distorted, out of focus image, beautiful in its own right, but not the same. The computer could not keep up with the graphics whilst copying it and so had changed the image.
Who was the artist of this image? Cat Power certainly never had this in mind, but the graphics were moving in time to her music, it was her music that made this particular graphic appear. The people that designed Windows Media Player programmed the graphics to randomly appear when particular rhythms are played, they designed the capability for this picture to be made, but they’ve never seen it. Can they be the artists? I saw a graphic that I liked and I pressed the print screen button, but when I pressed it the graphic didn’t look like this, it was colourful, moving and it had a sharper definition. When I copied the image the computer changed it. Not on purpose, it didn’t think about it in the way that a human might, it was following instructions given by many different people and by following them to the best of it’s capability it came out with an original piece of art. Not dreamed up or imagined by any of the above, so is the computer an artist?
In the past it has been common to think of a computer as nothing more than a tool when it comes to producing images (e.g. Benjamin, 1992), but when does a computer become more than just a tool? Can a computer ever be considered an artist in its own right? What attributes does a computer have to have in order to be an artist? What attributes does a human have to have? Do we have to replicate the features that make humans artists in order to be able to label the computer and not the manufacturer, programmer or even operator of the computer the artist?
It is only in the past 20 or so years that the issue of attributing art to anything other than a human has come about. In today’s world artificial intelligence is no longer just a figment of the imagination of film producers and novelists; it’s a reality.
Before considering if computers can be artists it may be wise to address the question, what is art? The reason being that if art is nothing more that an eye pleasing image then it may be easier for computers to produce ‘works of art’, however if it is more than that then maybe we have some way to go before AI can claim to be artistic.
Both expression and cognitive theories of art hold that art communicates feelings and emotions, or thoughts and ideas (Freeland, 2001). Leo Tolstoy (1898), a Russian Novelist believed that art communicates something in the realm of feelings, he believed in Expression Theory. His book written in the early twentieth century: “What is Art?” conveys this. John Dewey (1934) explains cognitive theory in his book: “Art as an Experience”. He argues that art is a source of knowledge, it can tell us how to perceive the world around us, or at least how others perceive the world. Both of these lead into complex problems when considering whether or not computers can produce art. How can a computer communicate feeling when it has none? A programmer gave the computer options, told it what things should go where and with what to make it believable. The same applies with cognitive theory. A computer producing an image is not producing its understanding of the world, it’s producing an image of someone else’s understanding; therefore, if we are saying that art is all about the understanding of the world, the artist must be the person who’s understanding the image is portraying. In this case the artist must be the programmer.
Harold Cohen was originally a successful painter who set out to incorporate his own rules of composition into an artificial intelligence program called Aaron (King, 1997). According to Burton (1997) Cohen saw the level of feedback within the computer program as defining whether or not the computer was more than just a tool. If the only feedback on the image being produced is from a human then the computer is nothing more than a tool, however, if the computer generates it’s own feedback, in effect deciding itself where the next mark, colour or line should go based on what it has done before, then the computer has transcended the role of tool. The images it has produced are its own, but is the computer the artist?
Aaron produced both of the above images on my computer (the programme is now downloadable). The answer as to whether Aaron is the artist is a difficult one, for if we look at the above definitions of art the computer cannot be the artist. It’s not showing its own feelings or understandings of the world… or is it? True, Aaron hasn’t learnt about the world in the way that a human might, but it has still learnt about the world. It just so happened that its method of learning was through the experiences of another: Harold Cohen. Of course it could just be said that because the computer is using the experiences of Harold Cohen and that Cohen is the artist. However similar to the first image that I produced on my computer, using Windows Media player, the programmer of the program that produced the art has never even dreamed of, let alone seen these images. So how can they be the artists? As Harold Cohen him self puts it:
“AARON exists; it generates objects that hold their own more than adequately, in human terms, in any gathering of similar, but human-produced, objects, and it does so with a stylistic consistency that reveals an identity as clearly as any human artist’s does. It does these things, moreover, without my own intervention… It constitutes an existence proof of the power of machines to do some of the things we had assumed required thought, and which we still suppose would require thought, and creativity, and self-awareness, of a human being.
If what AARON is making is not art, what is it exactly, and in what ways, other than its origin, does it differ from the “real thing?” If it is not thinking, what exactly is it doing?”
(Harold Cohen, 1995)
However, Cohen doesn’t claim that Aaron is an artist distinct from himself. He has stated that in order for Aaron to become an artist in its own right Aaron must draw something that it wouldn’t have been able to draw without his programming i.e. it must have taught itself something. It must have been creative. Creativity is, in the eyes of some, the key to producing art.
What, then, is creativity? The Concise, Oxford dictionary of English gives the definition of creative as:
“involving the use of the imagination or original ideas in order to create something.”
According to Margaret Boden (1998), Creativity is a fundamental feature of human intelligence. She goes on to discuss what she terms three different types of creativity i.e. ways of generating novel ideas, and how they can be replicated by artificial intelligence. The first type involves novel (improbable) combinations of familiar ideas. It is easy to see how computers could easily replicate this type of creativity, however this type of creativity is not one that can be readily applied to creating art forms. The second and third types are closely linked. They are “exploratory” and “transformational” creativity. The former involves the generation of novel ideas by the exploration of structured conceptual spaces. This often results in ideas that are not only novel, but also unexpected. The latter involves the transformation of some (one or more) dimension of the space, so that new structures can be generated which could not have arisen before. The more fundamental the dimension concerned, and the more powerful the transformation, the more surprising the new ideas will be (Boden, 1998). These two types of creativity are involved in creating art and, according to Boden (1998), have already been successfully replicated by artificial intelligence in order to produce art. She States that Aaron is in fact creative, as it creates novel images by exploring the rules given to it by Cohen, i.e. it generates novel ideas by exploring a particular space of knowledge. Other examples of transformational creative artificial intelligence are AM and EURISKO, and certain programs based on genetic algorithms. Some of these have produced valued structures that the human experts say they could never have produced unaided. Evolutional electronic art is a branch of algorithmic art that uses the concepts of Darwinian evolution to generate family trees of images; artists that have used this technology include William Latham and Karl Simms (King, 1997). However, both of these artists claim the art as their own, they do not attribute any of the finished product to a computer program. If the computer is the one that has created the image how can they get away with what appears to be blatant plagiarism.
The answer is that even though the computer program created the art that the artist used, it also created many other pieces of art, but the computer doesn’t have the intelligence to know what is good and what isn’t. It needs human intervention to select the ‘good’ art. So what does a computer need in order to produce art? The answer according to King (1997) is consciousness. He claims that both creativity and intelligence are by-products of consciousness. Consciousness is something that humans have without thinking about, and it is this that makes programming consciousness very difficult. Something that comes so naturally, without thought, is one of the hardest things to teach. It has, however, been done. MAGMUS is a computer created by Igor Aleksander this is designed to be conscious in the sense that MAGMUS can tell us what it is like to be MAGMUS, it is self-aware. As of yet there has been no attempt to make MAGMUS creative, but if it is possible to replicate certain parts of consciousness then it is almost certain that at some point in the future we will be able to create computers that can not only be creative, but that can distinguish their good creations from their bad ones.
In conclusion, if a computer produces an image who is the artist? One thing that this essay has shown is that the answer very much depends on who answers the question. One way of viewing the question is from the point of view of looking at what art is and then seeing if it’s possible for a computer to fulfill these requirements. There is very little literature on this particular subject; from my own point of view I feel that in its current state no technology can fulfill these requirements. This is not to say that I believe that a computer cannot produce art, I just believe that it cannot fulfill these particular views of what art is (i.e. expression and cognitive theories).
As to whether computers can be creative or not, I believe that they can. From this point of view I believe that computers can be, at least in part, artists. I say ‘at least in part’, for as I have shown, though in its current state computers can be creative, they have no idea as to whether what they produce is of any quality. Humans must currently be involved in the production of computer art even if that intervention is only small. Once consciousness has been incorporated into the programming of art producing computer programs I believe that there is no reason at as to why the art that those computers produce should not be attributed to the computer that produced it.
Reference list:
Benjamin W. (1992) Illuminations, London: Fontana
Boden M.A. (1998) Creativity and artificial intelligence, Artificial Intelligence, 103, 347-356
Burton E. (1997) “Representing representation: artificial intelligence and drawing” In S. Mealing (ed) Computers and Art, Intellect books, Exeter.
Cohen H. (1995) the further exploits of AARON, Painter, Stanford Humanities Review, 4,
Concise Oxford English Dictionary, Tenth Edition, Oxford: Oxford University Press
Freeland C. (2001) But is it Art? Oxford: Oxford University Press
John Dewey (1934) Art as an Experience, New York: Pedigree books
King M. (1997) “Artificial consciousness – artificial art ” In S. Mealing (ed) Computers and Art, Intellect books, Exeter.
Tolstoy L. (1898, translation published in 1995) What is Art? London: Penguin
Gemma Graphics 