By Balambal Suryanarayanan
For ages, art has been a blend of inspiration, imitation, and an artist’s reality, finding form in everything. From the strokes the fiery arms of the sun make over the canvas of the skies at eventide or the emotions that talk through a singer’s voice and an actor’s eyes to the weaving of a silkworm’s rich cocoon, the beauty of art can come from everywhere, its spell unfading across repetitions.
Musing along these lines and backed by the words of the famous surrealist artist Salvador Dali about imitation being an integral part of art raises the question if the term “art” in artificial intelligence was referring to the truth about AI being an art all along. This is because the growth of AI through the proposition of The Imitation Game by Turing was driven by one objective: training machines to imitate learning by experience like a human. But through observations of AI art taking the internet by storm, it is essential to address the future implications of AI’s influence on art. As if reading our minds, on Oct. 23, Carnegie Mellon Faculty Dialogues provided insight from four distinguished faculty experts who are actively involved with art and AI through a virtual webinar titled “AI and The Future of Art,” addressing the right concerns and sharing their insights about what entwining art with AI could mean for artists and shedding light on the controversial question, “Is it still art if AI creates it?”
Professor Vernelle Noel, the Lucian and Rita Caste Assistant Professor in CFA, first took the stage to showcase the dancing sculpture in the carnival celebrations in Port of Spain, Trinidad, highlighting the emotions of creativity, innovation, and joy. Explaining her project’s purpose to use technical interventions like machine learning and artificial intelligence to educate the public about various rich cultural and historical practices, she mentioned its significance in giving a sense of what might happen in a carnival, particularly the ones who might not be able to experience it in person. Training a model with such images, audio, and music, in her opinion, allows AI to serve its purpose for what art is: a life beyond shape and color, letting viewers feel and see what they have to see — the stories that the rich heritage of many different places tell.
Professor Johannes DeYoung, Associate Professor of Art in CFA, presented his next project, called “The Endless Mile,” a video mural and computational artwork that takes shape as a non-repeating and infinitely scrolling shadow play. For every presentation, a new arrangement of visual and audio elements is assembled in unique combinations across two types of performances. For a live presentation, the neural net model that was trained over 1,200 mid-20th century post-war jazz holds elements that respond to audio frequencies, allowing it to be presented live with artists and musicians. The other performance, a passive one, has the machine producing a resemblance of a voice by blending in low-frequency components. In another example, Professor DeYoung elucidated on a project his students undertook in an animated storytelling class called “Myth Revived.” Training a machine learning model with manually tagged traditional Donghua paintings, the students used ComfyUI, a powerful open-source diffusion graphical user interface (GUI) that generates images from the text prompt provided and adjusted the weights of the model to be able to generate an animated video with different panoramas in the style of a traditional Donghua painting from the images they provided, irrespective of their art mastery. With these examples, he shed light on the advantageous use of AI to retrieve art according to the environmental requirements and application use cases.
Finally, Professor Brett Crawford, Associate Teaching Professor of Arts Management in Heinz College of Information Systems and Public Policy, employed her presentation to powerfully elucidate on the use of AI by artists, its significant role as a powerful collaborator and the ethical implications that come with AI use. Using examples of the collaboration between dancer Kaiji Moriyama and Yamaha Corporation, The Royal Shakespeare Company’s production of “The Tempest,” and the Van Gogh experience, she mentions that artists have always entwined AI to give people an immersive experience. It is when big giants employ AI without receiving appropriate permissions from the artists that things go south. Finally, she presented the project “Dali Lives”, constructed by The Dali Museum in St. Petersburg, Florida to express the use of AI as an opportunity to engage with audiences, bringing the beloved and famous surrealist artist, Salvador Dali to life. By training AI to reproduce the likeness, mannerisms, and speech production of Dali, people can gain access to art by learning about the artist from the artist himself. While praising the impact of AI, she concluded with a question to reflect: With AI redefining immortality, how many artists would like to live after death, with their life dictated others’ hands?
The conclusion of the presentations transitioned to conversations with the presenting faculty led by Professor Daragh Bryne of AI and Computational Design, Associate Teaching Professor in the School of Architecture. Weighing on the ethical dilemmas surrounding AI as a sociocultural force and expressive medium, the faculty agreed that in addition to using AI to understand the cultures one identifies with, it is essential to reflect on the values associated with AI use for art, critically examining their need and use. Recognizing the power of art and its impact on people and the way the artists translate the privilege of art to communicate, the faculty also highlighted various reasons why the power of entwining AI with art stays important, stressing the use of AI tools like Photoshop as part of the art creation process. They acknowledged that the interdisciplinary environment of Carnegie Mellon opens doors to entwine AI and art. Finally, the faculty concluded their conversations by taking on the crucial question of the presentation: irrespective of who or what creates art, art is what one wants it to be.