So far, artificially intelligent systems have successfully taken many tasks previously done by humans and with each new development, they continue to outstrip our best human efforts. While AI has proved superior at complex calculations and predictions, creativity seemed to represent the last truly human domain.
This is why the new advances in "creative AI", such as the proliferation of AI artists and AI art, is making us feel uncomfortable - and for a reason, too. As artificial intelligence begins to generate stunning visuals, profound poetry, transcendent music, and even movie scripts, the nature of art and the role of human creativity in future societies start to feel uncertain.
The first-ever original work of art created using artificial intelligence to come to auction, Portrait of Edmond de Belamy (2018), smashed all expectations at Christie's New York last year when it fetched $432,500, soaring past its highest estimate of $10,000. The sale came after an intense competition from bidders over the phone which lasted for more than six minutes.
The work is one of a group of portraits of the fictional Belamy family generated by an AI trained by Obvious, a Paris-based collective. As its members explained, the artwork creation involved a process of feeding the system with a data set of 15,000 portraits painted between the 14th century to the 20th. As the program is composed of the Generator and the Discriminator, it is the Disciminator's task to determine whether there is a difference between the artwork created by the Generator and a human-made image.
Although this auction marked the first time that a major auction house paid attention to the field, the artwork and the process which generated it stirred many controversies in the art and AI world, especially when it comes to the attribution and originality issues.
Artists have been experimenting with AI for the past 50 years. The most prominent early example of this was in 1974 when Harold Cohen presented AARON, a computer program that can paint plants and people by itself. Cohen has continuously improved the system, teaching it more and more rules about perspective, human anatomy and the optical foundations of plants, but also how to use colors. A computer graphics artist and researcher, Karl Sims is best known for using particle systems and artificial life in computer animation. In 1994, he applied evolutionary computations techniques to interactively evolved images in his Genetic Images system.
However, this new wave of AI art differs from these examples where artists generated art by writing detailed code that specifies the rules for the desired aesthetics. These developments incorporate AI and machine learning technologies to allow the computer more autonomy in producing images. In contrast, these algorithms are set up by the artists to “learn” the aesthetics by looking at many images using machine-learning technology. Artificial intelligence then generates new images in adherence to the aesthetics it has learned. While an artist continues to be an important part of this process, whether in pre-production or in the tweaking of the algorithm when needed to achieve the desired outputs, the results which AI can create can be surprising even for the artists themselves.
However, in these surprising results, what is missing is an artist's intent. On the other hand, some argue that this is simply a piece of process-driven Conceptual art, where art is not in the outcome or final image, but in the process.
There is an increasing number of contemporary artists exploring the creative potential of artificial intelligence.
An artist using a neural network, code, and algorithms as his preferred tools, Mario Klingemann explores machine learning in the hope of understanding, questioning and subverting the inner workings of systems of all kinds. He aims to illuminate the subjects of human perception and aesthetic theory by creating algorithms that show almost autonomous creative behavior.
A New Zealand-based artist specializing in “creative coding”, Tim White was included in the first mainstream gallery show dedicated to art made using artificial intelligence last summer. White works with the neural network called Convolutional Neural Networks, which are used in today’s computer vision applications to give modern machine-learning systems the ability to perceive the world through vision. Through his work, he investigates the perceptual abilities of these systems by finding abstract forms that are meaningful to them.
An artist and programmer, Gene Kogan is interested in how generative systems, computer science, neural network and software can be used creatively for self-expression. He trains neural networks, a popular type of machine-learning software, on images, audio, and text. in order to develop generative models, or to help teach the software to create fresh, varied artwork based on inputs.
Earlier this year, an AI artist called AICAN and its creator Ahmed Elgammal took over a New York gallery with two series of canvas works portraying harrowing, dream-like faceless portraits, a result of a collaboration between a human and machine. As with Portrait of Edmond de Belamy, the machine was fed with 100,000 photos of Western art from over five centuries, allowing it to learn the aesthetics of art via machine learning and create a new work of art without human intervention.
We also saw the announcement of Ai-Da, the first ultra-realistic drawing robot artist who will have its first solo exhibition, Unsecured Futures, at Oxford University in July, 2019. The combination of her mechanical abilities and AI-based algorithms allow here to draw, paint and sculpt. Although lacking true consciousness, thoughts, or feelings, the exhibition organizers believe that the robot will serve as a stimulant for engaging with critical questions about what kind of future we ought to create via such technologies.
Many are skeptical of AI art. Pulitzer Prize-winning art critic Jerry Saltz said he finds the work produced by AI artists boring and dull. When reviewing works of art created by AI, the critic explained he was looking for "humanity, dignity, horror, originality.” His conclusion was that even the best effort of today’s AI technology “doesn’t come close to art.”
If art is defined as a process by which human beings express some idea or emotion, filter it through personal experience and set it against a broader cultural context, it might be argued that what AI generates is not art nor is it creative. Also, all the AI images so far have actually had a significant amount of human input.
However, while AI art is still in its infancy, the potential of the technology seems huge, demonstrating its ability to learn and mimic any distribution of data. And these developments raise some difficult questions about the future role of humans in an increasingly automated world. What does this mean for art and artists? Does it threaten them? It is certainly blurring the definition of an artist.
Marcus du Sautoy, an Oxford mathematician, argues that we should see the relationship between AI and humans not as adversarial, but as collaborative. He believes AI can get creative humans out of their usual ruts and spur them to think in new directions. This has been demonstrated by a range of artists experimenting with AI today. As all technological achievements served to push ourselves beyond our biological limitations so far, it is possible that these developments with AI creativity would only expand on our human creativity and push the boundaries of art.
The award-winning author of The Music of the Primes explores the future of creativity and how machine learning will disrupt, enrich, and transform our understanding of what it means to be human. While most recent books on AI focus on the future of work, The Creativity Code moves us to the forefront of creative new technologies and offers a more positive and unexpected vision of our future cohabitation with machines. It challenges us to reconsider what it means to be human―and to crack the creativity code.
Featured image: Ai-Da Robot with a painting created by her response to an oak tree Ai-Da Robot with a painting created by her response to an oak tree. Nicky Johnston.
Read Other Interesting Stories
While two art venues in Geneva and London are hosting solo exhibitions by Trevor Paglen, Fondazione Prada is hosting a show curated by the artist.
In an exclusive interview, Carrie Able talks about creating art in VR, the technology itself, its advantages and challenges, potential and much more.
In his latest exhibition at LACMA, Christian Marclay absorbed the material of digital media into a work that engages visitors interactively.