The art of the future / The future of art.

Next weekend I will be attending Inmersiones 2019, an independent project in the city of Vitoria-Gasteiz, to talk about Art and Artificial Intelligence.

This is an event created to meet baque artistic and cultural productions that require new spaces for broadcast. Here you can read a preview of what I will be presenting…



> INMERSIONES 2019, The art of the future / The future of art:  
“A critique of contemporary art made with / by AI”, Mikel Arbiza.

Edmond de Belamy” is the title of a painting that was sold in 2018 at the Christie’s auction house in New York for 432500$, and that was made by the Obvio French collective using a type of Artificial Intelligence algorithms called Generative Adversarial Networks (a subclass of Deep Learning algorithms).

This historic event – announced as the auction of the “first work of art made by an AI” – raises 3 interesting questions about authority, originality and artistic practices as a space for scientific research.

  1. Who is the author of a work of art in a mechanized process?
    First of all, we must emphasize that Deep Learning systems, because they are not human, create results that fall outside our classic classifications of creativity. However, we must begin to accept these new artistic manifestations since their computational architecture is clearly motivated by the way in which the brain of homo sapiens works. These other minds can serve as simulations that will give us very important clues about the neurological bases of different processes that take place in our brains.
  2. To what extent is originality a relevant value when talking about creation in contemporary digital societies?
    The members of Obvious collective when making the portrait of “Edmond de Belamy” used a library of Open Source algorithms created by a young man named Robbie Barrat that had been freely shared on the Github website and that were based on years of research in Artificial Intelligence carried out by different people. Therefore, in the opinion of many people this work of art is not original, a term still socially associated with the genius of the “artist”, and many question the ethics behind the commercialization of works created with algorithms that have been written by others. However, it should be stressed that the worlds of artistic creativity are full of examples of works done through recombination: from sound samples used by hip-hop music, to the impact of african art on Picasso’s painting style.
  3. Can creation and artistic reflection be a fertile space for technoscientific research (and vice versa)?
    The transformation of art, aesthetic theory and the ways of representing the world that pushed photographic innovation in the nineteenth century is equal to the current revolution in Artificial Intelligence that also promises to democratize the means of production. While many artists seem to laugh at the idea that an AI can be creative, certainly these synthetic minds are capable of creating “things” although from our anthropocentric perspective they lack intention and are unable to distinguish what is relevant. Existing computer programs lack too many relevant causal connections to exhibit intentionality while manipulating symbols and learning from them, but this does not mean that perhaps in the near future a new type of “embodied” Artificial Intelligence capable of interacting with their environment can give meaning to his artistic work.

In the opinion of the influential mathematician Alan Turing, and paraphrasing the philosopher Theodore Adorno, we could say that artists follow an intuitive logic, a process like any other activity linked to rules. Harnold Cohen himself, an pioneer in art made by AI, thought when designing his expert AARON system that “it should be possible to devise a set of rules and then, almost without thinking, to do the painting following the rules”. This approach is characteristic of certain types of artists such as Piet Mondrian and his classic classical abstractions of the 1920s and 1930s that were made according to a set of self-imposed regulations: only straight lines were allowed, which could only be found at right angles and they could be represented only in a palette of red, blue and yellow (more black and white).

However, although most people admire the craftsmanship and skills behind representational works, contemporary art has not been exclusively about creating images for a long time, and works done with Generative Adversarial Networks have the problem that they are essentially mimetic. Undoubtedly these artifacts produce interesting variations, but for now an Artificial Neural Network does not know how to distance itself from the world of data in which it operates, so it is possible that they will never be able to produce an image that reflects the historical context and shakes our notion of what constitutes a work of art.

While some people consider that the implementation and use of Machine Learning algorithms that we are seeing in the contemporary art world are the “Gesamtkunstwerk” or the total work of art, other views express that we simply need this type of cultural artifacts as critical of the industrialized use of Artificial Intelligence. Therefore, it is urgently necessary to discuss the criteria to be followed when analyzing and criticizing this type of new works of art made with Machine Learning techniques, since perhaps we are facing a historical moment that is defining the future of artistic practice.

bellamy-banner


*(Picture: screen-shot of my Artificial Intelligence agent's interface & Edmond de Belamy's painting).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.