Usage of AI in Art and Music
Artificial Intelligence, also referred to as AI, has transformed the creative fields of art and music by becoming a consistently-improving, highly sophisticated tool for creating and manipulating works. Among its countless purposes ranging from medical to architectural, artificial intelligence algorithms have been used as tools for creative expression by numerous artists and musicians, generating original compositions and creating brilliant and collaborative artworks. With the usage of deep-learning algorithms, AI is made able to possess similar abilities that humans naturally possess, have allowed machines to analyze vast amounts of data, learning from it and enabling them to create paintings, music, and even poetry that is indistinguishable from that produced by human artists. In the future, artificial intelligence can potentially increase access to the arts by massive margins, allowing people with little to no artistic training to create works original to them. It could become an integral tool for artists and musicians, enabling them to explore creative possibilities and generate new ideas that they might not have considered otherwise. However, there are concerns that AI could replace human creativity altogether, leading to a loss of emotional investment in the arts and what they truly represent or are viewed as. Despite these concerns, the intersection of AI and art and music is an exciting frontier with the potential for incredible new developments in the years ahead.
The sheer power of artificial intelligence for anything is difficult to quantify simply due to its ever-growing capabilities and perpetual learning. Similarly, art has followed an ever-changing path throughout human history and even pre-history, and throughout this path, its methods and forms have evolved in dramatic ways, some of which would be utterly unrecognizable to people as ‘art’ only one to two centuries ago. That said, the evolution of artificial intelligence, along with its potential uses, poses significant questions with art in mind: Will artificial intelligence replace humans in the visual arts field? If so, will the existence and creation of art remain one of the key complex concepts that separate humans and other organisms?
Many famous art pieces took lengthy amounts of time to complete. The Mona Lisa, for instance, was worked on by Leonardo for 4 years before he began other projects.1 The keywords here are ‘worked on’ – it was not even completed until 1519, 15 years after he started it. However, artificial intelligence can come up with numerous versions of the Mona Lisa, The Last Supper, or any other notable artwork, all within minutes or even seconds.2 All that said, artificial intelligence is a term which has not yet been explained within this paper, which is necessary to do before expanding any further. “Intelligence” originates from Latin and means “collection, collection, collection, and selection,” and is attributed to humans due to our nature of “expressing ourselves through mental work in activities that recognize and transform the world.”3 What differentiates human intelligence from this artificial intelligence is the ability of humans to dream, a subconscious ability. Therefore, “artificial intelligence” would suggest that it is intelligence belonging to something not capable of dreaming, something inorganic or perhaps not alive in the sense that humans are alive.
With this paper’s understanding of artificial intelligence now stated, we move on to the changes that are actively occurring in art with the usage of this concept. In his article, “Research on Application and Breakthrough of Artificial Intelligence in Art Design in the New Era,” Yonghui Lin explains how the “emotional expression” aspect of human art is nonexistent within art produced by artificial intelligence.4 Rather, machine algorithms and numbers are what drive the creation of this art. While still capable of coming up with beautiful works, the implementation of artificial intelligence within this field now erases the meaning behind the art. However, there is a catch: as of now, artificial intelligence takes inspiration from whatever images and other concepts it has access to, whether in local storage or on the internet. Lin acknowledges this and concludes his article by pointing out that human design, and by extension, creativity and the emotion behind it, can still play a role in the production of AI art. Contrary to Lin’s notion that artificial intelligence is comprised of algorithmic intelligence and access to recorded information rather than the creative and subconscious abilities of humans, the authors of the arXiv article “CAN: Creative Adversarial Networks, Generating ‘Art’ by Learning About Styles and Deviating from Style Norms” claim otherwise, arguing that because of its ability to produce art as well as show creative problem solving skills, it is in fact “intelligent.”5 The authors of this article state that artificial intelligence producing art should utilize human artworks for its own learning, and as a result, are more interested in developing algorithms that may be used for this rather than having the human art remain the inspiration. However, one must think about if artificial intelligence “learns” from human art and produces its own, it may not necessarily be truly unique. To some degree, it is still based upon the work of mankind, whether it be styles of shading, brushstrokes, or what is depicted. While humans may base art upon observations and other humans’ work, we also possess the imagination that does not revolve around algorithms and deep learning, analyzing what we perceive through our senses. We have dreams, we have creative thinking; artificial intelligence has memory and coded behaviours. This seems to be the point at which Lin and the arXiv article authors disagree on with regard to the capabilities and understanding of artificial intelligence. Both parties support the usage of AI with art production, but there is no consensus on the originality behind it and what traits this would mean the artificial intelligence possesses.
Visual arts aside, another common way artificial intelligence has been used in recent years has been within music. However, music faces the same issue that art does: it is another form of self-expression and is a manifestation of a person’s creativity and imagination. Therefore, artificial intelligence being used in the production or editing of music not only may not be original to the AI, but also could be attributed, again, to algorithmic learning abilities encoded within them. This goes without saying that the field of music on its own will improve throughout time just as technology does. Due to this, perhaps there is genuine potential for the involvement of artificial intelligence to aid in the improvement and development of music. In their Journal of Sensors article, “Research Based on the Application and Exploration of Artificial Intelligence in the Field of Traditional Music,” researchers Xia Huang and Yi Yang examine the changes experienced by traditional Chinese music.6 A current struggle faced within China is the gradual disappearance of traditional folk music as a result of a lack of successor styles or alterations made to them.7 What is proposed is the integration of artificial intelligence and these traditional styles of music to assist with their preservation and continual teaching, due to the value and place of importance they hold within Chinese culture and to their identity. This would be accomplished through melody recognition by the artificial intelligence, looking at various aspects of the music such as pitch, tones, and rhythm, among others.8
Other applications of artificial intelligence within music actively being investigated are within the educational side of it. Producing music is simply one part of the equation; the other part is learning how to do so effectively. SiYuan Yuan explores this use of AI and provides his thoughts in the article, “Application and Study of Musical Artificial Intelligence in Music Education Field.”9 What Yuan claims is that due to the rapid diffusion and advancement of artificial intelligence in recent years, “AI-related courses” should be implemented into education at the grade school and junior high school levels.10 He argues that it would be especially useful to take AI-related courses when studying music as it would both enable the improvement of AIs’ ability to “evaluate music information” and add to students’ “musical thinking.”11 Within this AI-music education, artificial intelligence would be utilized in a similar manner as Huang and Yang proposed: identifying various aspects of music and retrieving/developing music based on limited pieces such as emotion, lyrics, or tone.12 This would be invaluable within music production and education, as it enables those studying this field to both develop their abilities as musicians and to examine the methods used by the artificial intelligence to continue to produce or predict what pitch or melody a musical piece should use as a result of its various components.
While art and music may not be identical in many ways aside from the purpose in their conception, there does appear to be overlap and similar interests between what researchers state artificial intelligence has the potential to be utilized for within these respective fields. Rather than focusing on writing literature regarding original, unique works created by artificial intelligence, whether they be art pieces or musical compositions, researchers believe that the advancements experienced by AI enables them to be utilized well within teaching these fields, along with the generation of content in tandem with human effort and creativity. With this established, there is a lingering question to be answered: What happens when AI has advanced far enough to have a “consciousness” of sorts, and begins producing art or music on its own?
“Who is the real owner of the art and/or music?” If artificial intelligence continues to advance at its current rate, there will be artificial sentience eventually. To be “sentient” is to be “: responsive to or conscious of sense impressions,” as defined by Merriam-Webster’s dictionary.13 Given that AI does not truly “sense” or perceive information as many living organisms do, nor does it have a consciousness in the same sense, it cannot be sentient. It’s only a matter of time before machine learning reaches that point. Until then, if someone tells an AI to, say, “produce art until I say to stop” or something along those lines, who is the one really responsible for this? There is the option to simply say that whoever wrote the code for the artificial intelligence is technically the owner of this work. There is also the option to say that because the user is simply issuing a command while the artificial intelligence is actually gathering information it can use to accomplish this task, the AI should receive the credit. With a lack of true artificial sentience at this time, the possibility of an AI making the decision to produce artworks or songs or the like is not quite possible. And even if it did, this would bring up the point made by Lin of whether AI has “imagination” or is just using algorithms to collect information and base its work on them.14 And because of this, if an artificial intelligence did make the decision on its own to create an art work on its own, it would have to search through its storage and utilize some form of reference or inspiration. Without such things, the entire concept of machine learning breaks apart, as while it is not a living organism, it still learns through experience, observation, and any other “senses” programmed within it. Even so, it is the work of algorithms designed to take apart and analyze others’ art and music and to recognize patterns or key features within them that it can replicate within its generations of either type.
These problems bring the conversation on uses for artificial intelligence as a source of unique content to a standstill, both within art and music. Again, the lack of true sentience prevents it from gathering information from its surroundings or having a consciousness or subconsciousness that provides it with imagination through any form (e.g dreaming). It cannot be ignored that this is not agreed-upon by every researcher, whether it be the ones addressed within this paper or others actively involved with artificial intelligence advancements. Within music, the extent of AI’s “imagination” is to monitor patterns or other indicative information and then predict or write music based on that, but these are pattern recognition algorithms, designed for this purpose. They are not the natural ability of analyzing and creation that is within humankind. Within art, there seems to be a bit more leeway with regards to its production, but even so, if an AI were to produce an artwork on its own, it is uncertain if there would still be the meaning behind it that humans attribute to their own art. This entire issue appears to boil down to the one thing: the sentience in humans, and the lack thereof in artificial intelligence. Perhaps if an artificial intelligence possessed a mind identical to that of a human, they would have the agency and consciousness to produce art or music in the same way, with the same originality and creativity. Perhaps it would be representative of something in the same way humanmade works are. With the rapid improvements and inventions both with and within technology, many activities humans enjoy are being changed to accommodate and exploit these changes. Art and music, being immensely integral to human society throughout all of recorded history, and even pre-history, are no exception.15 Artificial intelligence, a concept which has seen massive spikes in public interest over the years, is a key example of the developments that are being used to their maximum potential within human activities, especially within the fields of art and music.16 Due to the ever-growing agency and ability of these AIs, however, the topic of originality of what they produce and whether they are sentient has also become frequently-debated. As of now, artificial intelligence does not possess true sentience in the way many living organisms do, particularly humans, though they are capable of producing art and music using methods that humans do employ. The way this is accomplished is through coded algorithms enabling these machines to “learn”, though, unlike humans, who generally are born possessing both natural instincts within their genes, as well as an imagination and consciousness. And it is this consciousness and its accompanying subconscious that gives humans the agency to do what we do, to live as we live, and to produce art and music representing emotions and feelings and concepts we imagine. Artificial intelligence, on the other hand, collects information from knowledge already produced by humans, for the most part, and simply analyzes these to learn to produce works in similar forms. It would then be fair to say that, while artificial intelligence is certainly a useful tool to employ in both art and music, due to its potential within teaching both and producing both in tandem with humans, it is not yet capable of freely producing either of its own whim, much less attributing these works to similar feelings felt and embodied within humanmade works.
Bibliography
“Application and Study of Musical Artificial Intelligence in Music Education Field – ProQuest.” Accessed April 3, 2023. https://www.proquest.com/docview/2557272343?parentSessionId=c%2BnnPRMe%2BTsEurnP9fsXpDXyZAJr8nP2Tk6qenQ5tCU%3D&pq-origsite=primo&accountid=10427.
Caramiaux, Baptiste, and Sarah Fdili Alaoui. “‘Explorers of Unknown Planets’: Practices and Politics of Artificial Intelligence in Visual Arts.” Proceedings of the ACM on Human-Computer Interaction 6, no. CSCW2 (November 11, 2022): 477:1-477:24. https://doi.org/10.1145/3555578.
Cetinic, Eva, and James She. “Understanding and Creating Art with AI: Review and Outlook.” ACM Transactions on Multimedia Computing, Communications, and Applications 18, no. 2 (February 16, 2022): 66:1-66:22. https://doi.org/10.1145/3475799.
“Definition of SENTIENCE,” April 13, 2023. https://www.merriam-webster.com/dictionary/sentience.
Elgammal, Ahmed, Bingchen Liu, Mohamed Elhoseiny, and Marian Mazzone. “CAN: Creative Adversarial Networks, Generating ‘Art’ by Learning About Styles and Deviating from Style Norms.” arXiv, June 21, 2017. https://doi.org/10.48550/arXiv.1706.07068.
“From Artificial Neural Networks to Deep Learning for Music Generation: History, Concepts and Trends – ProQuest.” Accessed April 3, 2023. https://www.proquest.com/docview/2480986460?parentSessionId=MvWsj1%2Bank5TniNckV8fJzvEFVgB0YMBkUicsGa2mOc%3D&pq-origsite=primo&accountid=10427.
Google Trends. “Google Trends.” Accessed April 16, 2023. https://trends.google.com/trends/explore?date=today%205-y&geo=US&q=artificial%20intelligence&hl=en.
Hong, Joo-Wha, and Nathaniel Ming Curran. “Artificial Intelligence, Artists, and Art: Attitudes Toward Artwork Produced by Humans vs. Artificial Intelligence.” ACM Transactions on Multimedia Computing, Communications, and Applications 15, no. 2s (July 19, 2019): 58:1-58:16. https://doi.org/10.1145/3326337.
“Last Supper | Definition, Description, Significance, & Art | Britannica,” March 25, 2023. https://www.britannica.com/event/Last-Supper-Christianity.
“Mona Lisa | Painting, Subject, History, Meaning, & Facts | Britannica.” Accessed April 14, 2023. https://www.britannica.com/topic/Mona-Lisa-painting.
Artincontext. “Prehistoric Art – An Exploration of the Prehistoric Art History Timeline.” Artincontext.Org (blog), November 19, 2021. https://artincontext.org/prehistoric-art/.
“Research on Application and Breakthrough of Artificial Intelligence in Art Design in the New Era – ProQuest.” Accessed April 3, 2023. https://www.proquest.com/docview/2571072351?parentSessionId=J%2Bz426KqysX8LMYsS%2FXoRfOY45pCaNd9zxPIK%2BKJ9R8%3D&pq-origsite=primo&accountid=10427.
“Research Based on the Application and Exploration of Artificial Intelligence in the Field of Traditional Music – ProQuest.” Accessed April 3, 2023. https://www.proquest.com/docview/2690831208?parentSessionId=i3uENul9exDoDLfVZunCHXRUQKzDKn5VfDysSvNURcA%3D&pq-origsite=primo&accountid=10427.“Research on the Application of Artificial Intelligence in Music Listening Learning — Taking Intelligent Hearing Training as an Example – ProQuest.” Accessed April 3, 2023. https://www.proquest.com/docview/2535686553?parentSessionId=2FG%2FYvqrGoEbau5hBiCOh0tEIPkTlIlP%2BVv5J8aSmfE%3D&pq-origsite=primo&accountid=10427.
1. Leonardo da Vinci, “Mona Lisa,” 1519
2. Leonardo da Vinci, “Last Supper,” 1498
3. Yonghui Lin, “Research on Application and Breakthrough of Artificial Intelligence in Art Design in the New Era,” 2
4. Ibid, 3
5. Elgammal, Liu, Elhoseiny, Mazzone, “CAN: Creative Adversarial Networks, Generating ‘Art’ by Learning About Styles and Deviating from Style Norms,” 1
6. Huang and Yang, “Research Based on the Application and Exploration of Artificial Intelligence in the Field of Traditional Music,” 1
7. Ibid, 2
8. Ibid
9. Yuan, SiYuan, “Application and Study of Musical Artificial Intelligence in Music Education Field,” 1-7
10. Ibid, 1
11. Ibid, 2
12. Ibid, 3
13. Merriam-Webster, “Definition of SENTIENT,” 1632
14. Yonghui Lin, “Research on Application and Breakthrough of Artificial Intelligence in Art Design in the New Era,” 1-6
15. Artincontext, “Prehistoric Art – An Exploration of the Prehistoric Art History Timeline,” 2023
16. Google Trends, “Artificial Intelligence,” 2023
