Can computers make art?


Is AI as great as the glossy brochures promise? Or is it only the bosses of the computer programmers who want art, and the result is only art or meta-art? The question touches on the question of being human. It stirs up fears and hopes. Fears that the machines will take over. Hopes that perhaps computers could liberate contemporary art practice from the zeitgeist. How beautiful would operas be without politically correct direction? How beautiful would it be if the last late medieval chant were not transformed into a ballet? How beautiful would it be if everyone saw opera exactly as he or she liked it? Big Data makes personalisation possible, but also the creation of works that one could innocently mistake for works of art. The fact that these sometimes seem cheap and shabby is irrelevant. Because art can have different qualities, even if it is made by people and there is no doubt that it is art. The question is rather whether and under what conditions we want to regard works created by computers as art. For this, it seems useful to me to take a look at the past.

A look back at early history

Art has remained something archaic for a long time. It did not integrate into culture, but remained the individual expression of the individual: the externalisation of personal thoughts, sensations, dream fantasies. Culture, on the other hand, was practice without individual origin. Art reflected culture, offered culture an object foundation, but it did not derive its meaning from culture. Consequently, there was no real progress other than the acquisition of new tools and the further development of these tools. A number of objections can be raised against this view, for example, the tradition of art’s self-reflection. The unique characteristics of art cannot be formulated with mathematical clarity either. Above all, however, the demarcations mentioned here are beginning to dissolve, for example through the practice of conceptual art and through some forms of post-dramatic theatre. (Note: what is meant here is not post-traumatic, but post-dramatic.) But also through zeitgeist forms of participation or the theory behind relational artworks such as those by Stellarc. Nevertheless, the origin of artistic practice some hundred thousand years ago (which remained largely determinative until recently) is crucial when it comes to the question of whether computers can create art. Art was one of the few first forms of externalisation of humans – and, as we have recently come to know, also of Neanderthals, whose genetic material we also carry within us. One expanded one’s own physical abilities by building tools (instead of relying on natural abilities or – this is also a significant developmental step of mankind – training these abilities better) In the process, one increasingly developed a preference for individualising, in the long run even a great passion, which time and again helped to survive great catastrophes. People used jewellery in a symbolic way, even as grave goods, to represent possessions to the outside world. And people drew images of animals and processes in paint on the cave walls. Similarly, as an externalisation of thoughts, narratives emerged, which in turn – it can be assumed – developed conscious cognitive thinking. Claude Levi-Strauss, for example, has shown how decisive the structure of stories is: if they are transferred to other societies, they undermine similarity transformations, which, from our contemporary perspective, change the meaning, but not the structure. The structure, however, shapes the dialectic of our thinking.

Externalisation without an inner self is not possible

When we speak of externalisation – or in German, very aptly, “Äusserung” – we assume that there is an inside, or a self. This means that it is not an animal driven by food addiction and sexual drive that expresses itself, but an ego capable of reflection, which in particular possesses a consciousness and is thus also capable of dreaming. Whereby the role of dreaming in this context is a key topic in itself and will not be dealt with here. In any case, it is the self that expresses itself and creates art. A self that is controlled by drives, but is more than these drives. A self that is related to other selves, but not like them (or actually: not like them). A self for which creating art, like consuming and internalising art, is an expression of autonomy. To this day, art has remained an individual expression. It is true that occasionally large production or performance teams are at work, and it is not uncommon for the creativity of all team members to be required, but the focus is nevertheless on both the externalisation of an artist’s inner world and the synchronisation of a viewer’s own inner world with the perceived externalised work. Synchronisation is not to be understood as agreement. Synchronisation can fail. Moreover, the result of a strong self-commitment of the audience to what is presented is usually only a simulation and not a commitment. When Shakespeare’s Scottish play is performed on stage, almost no one appropriates recurrent delusion, though many do experience it at the moment of watching.

Computers are only consumers, not creators

It follows from all these considerations that computers are capable of perceiving art, but not yet of producing art in the original sense. Computers can synchronise their neural networks with series of artworks, or even with individual artworks such as plays, of which many documented performances exist. They can even pack these synchronisations into “sandboxes” so that they can selectively retrieve after any. And they can overlay the synchronisations in such a way as to create an artificial synthesis of the synchronisations, not unlike what happens in humans. On the other hand, as far as externalising is concerned, this comes naturally to computers and the form is the result of UX, user experience design. But this self-evidence is reproductive and not a creative act. And as long as computers do not have a consciousness of their own, it will make little sense to speak of the artistic creation of computers. They lack a self that could express itself. Their substitute for a self is a digital memory that is only capable of reproducing itself. Computer art therefore remains art by people with computers. And it deserves the name if and only if the people behind the computers want to express themselves. Mathematical algorithms cannot even create computer art on their own. They may stand in the tradition of hundreds of thousands of years of human externalisation – and re-internalisation of their own externalisation – but they are only tools with philosophical origins and not creators of art. We should therefore free ourselves from fear AND hope that computers create art. We must – unfortunately or fortunately – continue to live with purely man-made art.

Postscript for long-suffering opera fans

If you have just recently had to experience another one of those particularly benign opera productions that denied you half of the arias because of Mozart’s moral depravity. Please do not be horrified! There is at least half a hope: directors can be replaced by robots – and we are working on it with the conductors! Only one thing, one thing we should not want: Personalised art. Art may have been created in caves, but it only acquires its essence when it leaves these caves. Art is – also – the strange, the disturbing, the crazy and the wrong. And art – as part of culture – is an origin of the social, a form of sharing and a foundation for public space. This by no means rules out visiting an art museum equipped with artificial intelligence, knowledge archives and augmented reality. But we should remain aware that the point is to personally synchronise with the foreign, not to adapt the foreign so that it fits well into our high-cultural wish-box. To avoid misunderstandings: If we replace directors with robots, it is only because they are constantly defusing art and explaining what cannot be explained among adults. And because in their self-important self-presentation they are indistinguishable from the glossy sales brochures of the AI industry!

Creative Commons Licence

AUTHOR: Reinhard Riedl

Prof. Dr Reinhard Riedl is a lecturer at the Institute of Digital Technology Management at BFH Wirtschaft. He is involved in many organisations and is, among other things, Vice-President of the Swiss E-Government Symposium and a member of the steering committee of TA-Swiss. He is also a board member of, Praevenire - Verein zur Optimierung der solidarischen Gesundheitsversorgung (Austria) and, among others.

Create PDF

Related Posts

None found

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *