<<Previous – Up – Next>>
If Computer Art is truly a new departure in the visual arts, then the computer must contribute specific elements to a work of art. At the conference “Beyond Art”, Robin Baker (Director of Ravensbourne College), posed four questions to assess the computer’s contribution:
1) Can computational ideas help us to understand human creativity?
2) Can computers assist creativity?
3) Can computers recognise creative material?
4) Could they eventually become creative?[1]
The first question is problematic because any computational idea can only be applied to human creativity by analogy. In the same way that calling the human brain a “computer” is at best a simplistic metaphor, so calling its processes “computational” simply means they are sequences of procedures. However, if by “computational ideas” Baker is referring to the operation of programs that simulate creativity or invention, then his question has already been answered by neural nets and recent software that can seemingly innovate independently. In one case, a program evolved more efficient circuit designs than those made by human designers. Moreover, Harold Cohen’s AARON program creates pictures using artistic knowledge encoded by Cohen over the past 25 years.
Yet as James Bailey points out in After Thought, the computer is such a different “thinker” that one cannot really base any assumptions about human creativity on the performance of computer software. Computers may take on tasks that require intelligence when humans perform them, yet “the capabilities they develop may bear no relationship to the “intelligence” we have developed.” [2] Bailey believes the result will not be “artificial intelligence” per se, but some new definition of intelligence, constrained by its development on a computer.
He argues that the computer’s form of intelligence is quite different, alien even, and it thinks in ways that bear no resemblance to a human mind; any form of apparent artificial intelligence arises in entirely new ways. However, whatever its underlying “intelligence”, it is still a salutary experience to see a process-driven machine apparently innovating and making art of its own accord. Just as Edward Ihnatowicz’s massive sculpture, the Senster, exhibited apparently life-like movements even though it reacted to sounds through a relatively simple program, so complex emergent behaviour can arise from relatively simple starting points. [Cf "Maggoty Intelligence" by Ihnatowicz]
Baker’s first question on human creativity is closely connected with his fourth: whether computers could ultimately become “creative”. It very much depends on what criteria one attaches to creativity, and whether one is willing to make allowances for the programmer’s input into the parameters of the software. Both questions treat the computer as an entity, which is the impression one gets when one sees it operating independently of any human input. Its status as an entity is increased by the physical presence of the hardware (especially the old monolithic mainframes), and any quirks on the part of the software which usually serve to frustrate the user.
However, in the making of Computer Art the computer only exists to perform its calculations and in a sense only exists through this task. It may have sufficient intelligence to recognise or generate pictorial elements, but it lacks the artist’s accumulated contextual knowledge: the artist’s universe includes the computer, but the computer cannot conceive of the artist. Thus the artist will always surpass the computer as creator (or author). Yet the computer can often manifest seeming independence: the “wily computergeist”, as theorist Tim Binkley termed its idiosyncratic responsiveness.[3] This “computergeist” seemingly animates the machine; it is seen when unexpected results appear through random processes; or when the artist is afflicted by the computer’s wilful quirks, such as crashes. Binkley deems the computer a partner to the artist:
The computer is a creative partner – a protean machine intelligence we have few precedents in comprehending. Artists are just beginning to learn how to work with these enigmatic creatures and their far-reaching ramifications for culture in general will be an open question for some time to come.[4]
This semi-sentient activity may result in apparent computer “creativity” through the use of Artificial Intelligence; or processes that generate and evolve new images. It is a moot point whether this amounts to truly independent creativity. Rather, the computer produces unexpected and unpredictable images in which a human viewer might observe the results of a creative process; its flexibility is a testament to the programmer’s skills. This may address Baker’s fourth question: perhaps the computer will never become creative in a human sense – able to innovate entirely new base material; but instead it may perform endless innovatory variations on material it is supplied with.
At the most basic level, the computer acts as a conduit for visual information because all images created on are reducible to sequences of code. (Analogue computers create images by using different waveforms to control a cathode-ray tube, but as a class of computer they are obsolete.) In this sense the computer is neither a tool nor a medium; its imaging tools and visual space are themselves based on code and are not intrinsic or irreducible features. The artist initiates the processes and the computer becomes a pipeline which channels artistic ideas into visual forms. Time becomes non-linear, since actions can be reversed and different timelines can be easily created.
Thus the computer can make abstract ideas concrete or, conversely, make physical forms insubstantially digital. When imaging tools and software functions are provided, they afford the alteration of the digital image, and as they are made of the same “substance”, they are as malleable as the image itself. As Timothy Binkley said: “the concretion of a digital image is arbitrary in a way media are not.” This means that none of the forms in which the image is output, from the monitor to the print, can claim primacy as “being the “true” realization of the virtual image”.[5]
While stored as digital information, the image is still in flux because the surface remains malleable and the work of Computer Art – its inherent artistic idea – is also subject to change. Yet these images can be made physical through printing in two or three dimensions. “Computer Art” may express its reactive and interactive potentials through a physical medium, or remain entirely on the computer. The art-making process must engage the computer and artist in a dialogue to fulfil my definition of Computer Art.
If a piece of art results from a truly interactive process taking place in an unconstrained data-specified medium, this should be recognised as “Computer Art” even if the resulting trace is a print-out. Of course, this print is but a snapshot of the artistic process – it does not fully realise the potential of computer artforms.
Insofar as the “computer” is the platform on which the software runs, it supports the visual space which the software creates. It does not really “contain” this space, because the space invoked by software is an operational space, which only exists through processes and never achieves independence from the platform on which it runs. To say that this space exists “inside” the computer is misleading. However, common usage has conflated the computer as platform with the software it runs, so it is more convenient to address the space accessed through the computer as being “within” it. In practice, this is true because the computer is the only gateway through which we can experience this non-material space. The computational medium, insofar as it exists, can be treated as a synergy between software processes, the hardware platform and the visual results by which these processes can be comprehended.
Through simulation, the visual appearances and dynamic processes of the physical world can be recreated with appropriate software. Computer-created objects can be subjected to simulated temporal and physical forces, and even made to “evolve” according to certain rules. Alternatively, abstract and mathematical images may be created with no physical limitations on their form. Describing Virtual Reality systems, Sheldon Brown touches on a factor which is true of all computer-based media, insofar as they can contain the practices and appearances of previous artforms:
VR can contain cinema, it can contain video, it can contain sculpture and architecture, but it doesn’t simply replay them. It re-stages them in a metamedium, which is experienced as the interplay of representations. [6]
The computer also introduces interactive and reactive elements into art, even at the level of image creation, which can be made independent of the artist to some degree. As a consequence it allows the artist to relinquish some control over the image to generative processes. This makes the art continuously dynamic and in some cases it effectively has no “final” form. The viewer can also have an important impact on the art, because it can react to their presence.
The computer (more precisely, the software) removes physical constraints normally encountered by artists. Previously, the artist always aimed to master their chosen materials to the extent of these substances’ inherent physical qualities. These material limitations placed the ultimate boundary on the form of their work.
Computer graphics overcomes physicality by simulating objects in a virtual space. One might consider in what sense these onscreen objects are “real”: they hover somewhere between the conceptual creations of the artist’s imagination and physically realised external objects. For the purposes of Computer Art, they are treated as externally-existing objects described by the data and located somewhere “inside” the space presented by the computer screen.
In general, the computer is controlled via standardised physical interfaces which allow the artist to use expanding and changing virtual [software] constraints. The major constraint is the lack of specialised physical interfaces; by contrast, computer musicians have access to a wide variety of interfaces which have been developed for specific tasks.
It is important to distinguish between the inherent constraints of particular computer hardware and software, and the artist’s choice to place limitations on them. Artists can choose to limit themselves to the dimension or aspect of virtual space they wish to develop, selecting or developing tools for this area. However, they also face the technical constraints of their approach, which are inherent to it and impose limits on the overall artwork. For instance, the graphical user interface is the current constraint on many computer artists, but this will feasibly break down over time. Ultimately, the computer might be controlled with gestures or even by reading brainwaves.
Art communicates and evokes emotional responses, in addition to more concrete ideas. It may be considered in terms of satisfaction – aesthetic or otherwise – and as an exercise of “uncommon skill”. The emotional element may become ever more important as computers develop interfaces to recognise and reflect emotional responses. As a recent Wired article explains, artists are already generating computer images by converting emotional states into electrical impulses.[7]
[1] From the conference “Beyond Art”, Oxford, May 1999.
[2] James Bailey, After Thought, p11
[3] Schwartz, Lillian & Laurens F., The Computer artist’s Handbook, (New York, 1992), introduction by Binkley, p xiv
[4] The Computer artist’s Handbook,Lillian Schwartz, Introduction pxiv
[5] “The Quickening of Galatea: Virtual Creation without Tools or Media” by Timothy Binkley
[6] “Real Art and Virtual Reality” by Sheldon Brown, ACM SIGGRAPH Computer Graphics, Volume 4 Number 4, November 1997. Web Ref
[7] http://www.wired.com/news/culture/0,1284,52940,00.html [expand]
