Artificial Intelligence must rescue the World of Arts, and this is how.

How art is vital for society

It has not gone unnoticed that Artificial Intelligence (AI) has been taking over the world by storm. The past decade has been especially transformative for AI. From autonomous weaponry, fraud detection methods, early-stage cancer diagnosing to identifying pandemics and helping alleviate starvation, AI is a truly innovative technology with far-reaching effects in all aspects of today’s modern society. However, when it comes to AI infiltrating all aspects of our society, one crucial field remains underexposed: the world of arts. Art is a broad concept and therefore hard to capture in one comprehensive definition. In this article, we particularly refer to visual art which encompasses diverse media such as painting, sculpture, drawing, photography, printmaking and decorative art. But why is art so crucial to society? To start with, economically speaking, art attracts tourists, businesses, talents and generates new employment opportunities. To illustrate, as of 2017, the art industry was worth 92 billion pounds and accounted for 14.2% of UK’s Gross Value Added (GVA) . Secondly, art has an undeniable great psychological impact. It promotes expression, creativity and the development of soft skills. In fact, participation in arts has been shown to increase the likelihood of young adults to be recognized for academic achievement by four times. Also, research supports the idea that engagement in art is linked to higher levels of mental and physical wellbeing as well as improvements in cognitive functioning. Lastly, culturally speaking, art provides historical context, supports thoughtful cultural discussions and embeds cultural-dependent symbolic meanings within. Thus, art benefits all facets of society. Nevertheless, art engagement has declined significantly over the past decades. The average number of attended art-events dropped from 6.1 to 4.8 within 10 years. Also, audiences have been steadily greying, implying that the field of art may be threatened with extinction. Thus, the shrinking field of art forms an undeniable threat to our society and a solution is therefore desperately needed. We strongly believe that AI serves as the overarching solution to the eclipsing world of arts. 

Overcoming barriers

There are numerous reasons why people find it hard to get engaged or stay engaged in arts. There has been extensive research in the enduring boundaries one faces regarding art engagement. Both a study from the Art Council England as well as an extensive report on how young people could get more engaged in arts and culture state that striving to increase engagement is a three-folded task. Which means, we have to overcome the following three barriers to engagement: an attitudinal, functional and practical barrier.

Figure 1: The the barriers as described by Art Council England

As stated before, we believe that the implementation of AI in the field of art will serve as the grand solution in overcoming those barriers as a result of which art engagement is eventually enhanced. This enhancement, in turn, will rescue the world of arts inducing far-reaching positive effects for our society. 

AI overcomes the attitudinal barrier

The attitudinal barrier refers to a thorough feeling that art is not something for people like you. You simply feel like you don’t belong. As can be seen in Figure 1, overcoming the deep-seated attitudinal barrier will likely benefit the whole of society in the long-term, making this the most critical barrier to overcome.  The feeling of not belonging in the world of arts is rooted in an overall sense of not getting involved. This is how we believe that AI overcomes the attitudinal barrier.

To start with, the possibility of AI to create endless amounts of unique art pieces will bridge the scarcity of art as a result of which art is no longer something prestigious only available for the elites of our society.

Figure 2: Edmond de Belamy

The 25th of October in 2018 marks a special day for AI generated art, as the first AI generated art piece went off for auction at Christie’s.  It was estimated that the piece, named Edmond De Belamy (see Figure 2),  would sell for around $10,000, while in fact it was sold for a whopping $432,500. Generative Adversarial Neural Networks (GAN) is the prime technology behind AI generated art and was created by Ian Goodfellow in 2014 . GAN is a very sophisticated machine learning (ML) technique whose general idea comes down to a pair of Neural Networks (NN) that fight against each other. Github, a website that hosts coding projects, returns over 4,925 open-source implementations of GAN’s. GANs tend to generate an image comparable to its training set. While creating Edmond De Belamy, a training set made out of 15000 portraits spanning the 14th to the 19th century was being used. The fact that Edmond De Belamy was sold for $432,500 was simply by reason of it being the first AI generated art piece that went for auction. Despite its price, this example illustrates the feasibility of generating unique art pieces by the use of AI. Whilst real paintings (and Edmond de Belamy) are mainly reserved for the very rich, GAN generated art is accessible to everyone in society due to its ability to create infinite amounts of art, resolving art’s scarcity. One could additionally argue that more art will lead to more engagement as there becomes more for people to enjoy. The same holds for any social media platform, when there is little content there will be little engagement, increasing the amount content, in turn, generally leads to more engagement. 

However, experts in the world of arts argue that AI generated art can not be classified as real art. Philosopher Sean Dorrance Kelly claims that AI art is not innovating, as it needs a dataset of artworks made by actual humans in order to create something new. Therefore, it cannot be seen as something innovatively new. Accordingly, AI generated art should thus, according to Sean, not be classified as art. However, the definition of what makes something a piece of art has always been quiet ambiguous and debatable. A widely adopted definition of art, describes how it is the artist who makes a work, and the audience which turns that work into actual art. According to this definition anything, both human created as well as computer generated, can be seen as art as long as there is an actual audience engaging in it. As mentioned before, the possibility of generating infinite pieces of unique art through AI, will increase the amount of art in circulation. More art in circulation, in turn, enhances engagement whereby AI art is being turned into real art.

Another way in which AI enhances involvement, is the capability of AI to tailor art to one’s individual preferences. As mentioned before, GANs tend to generate an art piece similar to one in the training set. Implying that all you have to do is download an open-source GAN algorithm from Github and an artwork dataset that is inline with your preference, and the GAN will create a free and unique art piece which is tailored to your individual taste. Moreover, museums seem to be at the forefront of adapting AI technology as part of their exhibition strategy. To illustrate, the Metropolitan Museum of Art in New York collaborates with Microsoft to pull open data from visitors for delivering an artwork that is believed to resonate with the visitors preferences. By matching art to one’s individual preferences, one may feel personally addressed and increasingly involved in the world of arts.

However, the Knight Foundation and the American Alliance of Museums hosted a series of convenings about the use of AI in museums. They point out that there are hidden pitfalls within the naïve enthusiasm to enhance art involvement with AI and visitor’s  data. They mention the racial biases of most AI systems and address the issues of privacy for visitors’ data. They fear that the use of AI in striving for involvement in the world of arts, will likely cause the opposite effect of involvement, namely exclusion. 

Although privacy and racial biases are decidedly important issues to address in today’s digital revolution, the concerns of the Knight Foundation and the American Alliance of Museums are not particularly justified. As of 2018, the European Union has forced museums to consider the issues of data privacy as they launched the General Data Protection Regulation (GDPR) law. The GDPR is the most rigid privacy and security law in the world as it imposes strict regulations regarding privacy and security onto any organization so long as they collect private data. Those who violate the law will face penalties over tens of millions of euros. We can thus safely state that the privacy of visitor’s data is being guaranteed.

Now that we have shown you how AI overcomes the attitudinal barriers by breaking through the ever existing scarcity of art and by matching art to one’s individual preferences, we would now like to illustrate how AI can additionally overcome the functional barrier

 

AI overcomes the functional barrier

The functional barrier refers to not being able to take part in the world of arts, regardless of how much you might want to. The opportunity to take part is just not there. As can be seen in Figure 1, the functional barrier is situated in between the deep-seated attitudinal barrier and the day-to-day practical barrier. Overcoming the functional barrier will likely benefit arts and cultural organisations in the long-term. The most common reason for the functional barrier to exist, is the lack of information and education about art provided by cultural organisations and educational institutes. There are numerous ways in which AI could augment art education. 

First of all, on an educational institute level, AI is being implemented to teach and engage students in the world of arts. For example, artist and teacher Stephanie Dinkins, recently taught her students how to use a type of neural network called Deep Dream to create self-portraits and t-shirt designs. Aside from implementing AI to let students create art, there are endless possibilities for AI to teach students about art. In fact, Virtual Reality (VR) is already a widely adapted technology in arts education. A study concerning the application of AI in art teaching found that it:

“presents art knowledge to students in a more intuitive way, creates a better learning atmosphere, exhibits design works that combined art and technology, helps students enter the creation world of the artists, makes students truly immerse in the masterworks of great artists, breaks the limits of time and place, and enables viewers to observe the details of art works more intuitively”  

Fanwen Kong

Despite those promising effects that show how AI could positively contribute to better art education, sceptics argue that this art teaching mode has ignored the sense of experience and integrity emphasized in art teaching. They particularly argue that art is to a great extent composed of human emotion and can therefore only be taught through the one-on-one form between two humans. An artificial intelligence system would not be able to transfer the sentiment of art.

However, those concerns are proven to be unfounded and outdated. As the field of AI continues to develop at an excessive pace, AI algorithms have already been developed that proved to accurately recognize emotions and sentiment in art. Engineering professor Leonidas Guibas trained neural speakers that allow the AI system to generate emotional responses to art and justify those emotions in language. Moreover, the system is not solely able to capture the broad emotional experience of an art piece, it is shown to be fully capable of analyzing different emotions within one art piece. 

AI supported art education is not exclusively for educational institutes. Museums have shown to be leading the way in the usage of  AI to educate their visitors about their current exhibitions. An example is the Museu do Amanha located in Rio de Janeiro, Brazil. This museum uses an AI system named Iris+ that was particularly designed to help users better understand and contextualize their exhibitions. As Iris+ uses a conversational interface to interact with their visitors and engage in open-ended questions, Iris+ is capable of personalizing visitor’s experience in the museum by showing how their thoughts, concerns, experiences and preferences relate to the exhibition. 

AI overcomes the practical barrier

The last barrier to overcome is the practical barrier. As can be seen in Figure 1, the practical barrier is the most surface-levelled barrier. This barrier means practically not being able to come along because of an inconvenient time or location.  Also, you simply might not know that something is happening due to a lack of information. Overcoming those practical difficulties will likely benefit individual young people in the short-term. We believe that AI powered immersive art experiences help overcome this barrier. 

Immersive art is a form of art that strives to fully immerse the audience in the artwork. Through light and sound installations, it creates the feeling as if one is totally absorbed to stand within and not before the artwork. These installations come in many different forms striving for the same goal; making art come to life. A wonderful example of how immersive art could help overcome the geographical barrier is the adoption of Virtual Reality (VR). The London Collective brought together more than 20 of London’s art galleries to host their exhibitions in VR. As of 2020, when the world was forced into physical and social isolation due to the COVID-19 pandemic, the demand for virtual exhibitions grew exponentially as museums remained closed. Whereas VR immersive art demands for the possession of a VR headset, virtual museum tours only demand for an internet connection. The British Museum located in London, allows their visitors to take a 3D virtual tour through the entire museum with help from Google Arts & Culture. These VR experiences and virtual museums tours allow all people to engage in the world of arts, even if geographically not around. 

Figure 3: A snapshot from Anne Frank’s bedroom in VR

Critics however argue that the technologization of art is not in line with what the original artist once intended. Also, most virtual museum tours tend to neglect the architectural context of an art piece. Little attention is paid to why art objects are where they are, something that museums work hard for to highlight. They argue that experiencing art is not simply about passively looking at it. It encompasses the entire context in which an art piece is being absorbed. As Jorrit Britschgi, director of the Rubin Museum of Art, once said; “there is nothing that can replace in-person encounters with art. ”We however strongly believe that there is a great opportunity for AI to strengthen and deepen the experience of art as a whole. Even when it comes to the architectural context of art if one is not physically present. The Anne Frank House, for instance, virtually navigates its visitors through spaces of ‘The Secret Annex’ paired with descriptions of her diary. This constructs an experience in which a visitor is able to see and experience the space through Anne’s eyes. For example, sitting at the attic window, the only window at which Anne could see the outside world, visitors are presented with the audio: “As long as this still exists, I cannot be sad.”

It is the audience which makes art art

With this article we aimed to display the endless possibilities of AI to rescue the world of arts. As art engagement has been declining over the past decades, a threat for the field of art itself but also for the society as a whole is lurking. Art has been proven to augment the economy, physical- and mental wellbeing and the cultural sector. It is therefore crucial that AI now infiltrates the field of art and that its endless possibilities are used to benefit this shrinking field. 

Whereas most sceptics argue that technology would never be capable of replacing in-person encounters with art and that AI generated art should not be classified as art since it is not created by human artists, we argue that it must be that those sceptics have not immersed themselves thoroughly enough into the endless possibilities of AI. Because if they did, they would have found out how art is being classified through its audience and not through the humanity of the creator. They would have found out how AI is not simply replacing in-person encounters with art but how AI is reinforcing symbiotic relationships between technology and art experiences. They would have found out how AI can support art education which eventually enhances art engagement of visitors from all layers of our society. They would have found out how AI creates the ability to reach people to engage in art who would normally not be reached because of practical difficulties. And, most importantly, they would have found that we all strive to achieve the same goal; not letting the beautiful field of art extinct.

Finally, we would like to invite you to attend art exhibitions, either offline or online. Because in the end, it is the audience that makes art art. Or try generating a piece of AI art yourself with one of the 41 tools suggested on this website.

Leave a Reply

Your email address will not be published.

Human & Machine

AI and health: A perfect match?

The machine plus clinician is better than the clinician, and it’s also better than machine alone Dr. Eric Topol Technological changes are often initially treated with scepticism and it can take some time before the benefits become clear and the progress is accepted. This is of course not a problem specific to medicine, but it […]

Read More
Human & Machine Power & Democracy

Cancelling Robbogeddon – Why AI Won’t Make Human Labor Obsolete

“The changes are so profound that, from the perspective of human history, there has never been a time of greater promise or potential peril” Klaus Schwab, author of The Fourth Industrial Revolution Humanity is experiencing a time of rapid technological development. Advances in Artificial intelligence (AI) and its ability to augment automation are seen as […]

Read More
Human & Machine

Don’t Let the Industry Take Care of AI Ethics

On January 9, 2020, Robert Julian-Borchak Williams was called by the Detroit Police Department with the request to turn himself in. Williams assumed this was a silly joke, but an hour later the police showed up at his house to arrest him in front of his wife and children. After spending a night in jail, […]

Read More