Exploring Creativity at the crossroads of AI and Art

12-minute read
cryptography
aes-ecb • aes-cbc

In today’s rapidly evolving technological landscape, AI tools are being developed and released and seeks to create a rich dialogue around the intersection of AI and art, ultimately fostering at an unprecedented pace. This has sparked a widespread interest in understanding the impact of these advancements on various industries, particularly the art world. Artists, both traditional and digital, are exploring the potential of AI to create novel and innovative forms of expression. This project delves into the thoughts, experiences, and perspectives of artists who have embraced the fusion of art and technology, particularly AI, as an integral part of their creative process.

The aim is to unveil the opportunities and challenges faced by these artists, as well as to investigate the relationship between AI-generated and traditional art forms. By doing so, this project deeper understanding of the role technology plays in shaping the future of creativity.

To achieve this, the project will address a series of research questions, focusing on artists’ journeys into the world of AI, their creative processes, the value of AI-generated artworks, the potential synergies between AI and traditional art forms, and the ethical considerations that come into play when working with AI in the art industry. Through a series of engaging podcast interviews with artists at the forefront of art and technology, this project aims to offer valuable insights into the rapidly changing landscape of art, as well as the potential opportunities and challenges that lie ahead for artists and the art industry as a whole.

Research questions to be addressed include:

  1. The artists’ journey into merging art with technology, specifically AI.
  2. Their creative process when using AI and new media tools.
  3. Unique challenges and benefits of incorporating AI into their artistic practice.
  4. The potential of AI to complement or enhance traditional art forms.
  5. The emotional and intellectual value of AI-generated artworks compared to human-created art.
  6. The evolving relationship between AI-generated art and traditional art.
  7. Ethical considerations for artists and the art industry when working with AI.
  8. Surprising or unexpected results encountered while creating art with AI.
  9. The future role of artists in the context of AI.
  10. Inspiring or exciting projects and collaborations in the AI and art space.

Methodology: A series of 15-20 minute podcast interviews will be conducted with artists working at the forefront of art and technology. The conversation will flow organically, with potential amendments to the questions based on the discussion.

Artists to be interviewed include:

  1. Shiny Parashar
  2. Hasan (@curiouswala)
  3. Computational Mama
  4. Anushka Tridevi(@Ophelia Game)
  5. Samarth
  6. Rasagy Sharma

The first podcast is recorded with Shiny Parashar, currently doing her masters in art.

The tools used:

  1. Jitsi Meet for recording the meeting room. About: Jitsi is a collection of free and open-source multiplatform voice, video conferencing and instant messaging applications for the Web platform, Windows, Linux, macOS, iOS and Android.
  2. Adobe Premiere Pro for video editing

Link: Second Podcast I was unhappy with the production quality of the first podcast. So I halted and researched for better tools. I also improved on my editing significantly. The second podcast is recorded with Hasan: https://curiouswala.com/, Tools:

  1. OBS Free and open source software for video recording and live streaming.

  2. VDO.ninja : to get full resolution video footage of the 2 speakers and also get us on separate audio tracks which made editing easier. About VDO.NINJA: 100% free; no downloads; no personal data collection; no sign-in, Peer-to-Peer

  3. Adobe Premiere Pro for Video Editing.

  4. Essay

In this essay I explore the concept of AI-assisted art and its implications in the realm of creativity and synthesize my evolving thoughts on the subject, drawing from the discussions held during the podcast series as well as relevant existing literature. I delve into the question of whether AI-generated art qualifies as true art and examines the symbiotic relationship between artists and AI technology. Then I discuss the need for inclusive AI art tools, challenging power dynamics and representation in the art world. I talk about the evolving statefulness between artists and their tools, considering the balance between efficiency and artistic expression. Finally, I look into the impact of AI on transitioning between artistic mediums and reevaluate the significance of McLuhan’s “the medium is the message” in the AI era.

Let’s dive into the realm of artistic elephants, where their presence cannot be ignored, nor their graceful dance overlooked[8]. The question of whether AI-generated art qualifies as true art immediately raises a significant debate. To address this issue, it becomes necessary to establish a universally accepted definition of art.

“What is art”

Shiny believes it’s anything that has its own unique style, for Hasan it’s a container that he can pour himself into. Gombrich on the other hand claims there is no such thing as art. There Are Only Artists. Perhaps it’s not fair to look at art without an artist?

“Not AI art, but AI assisted art.”

We mostly agree on the fact that artwork created using tools like Krita, Adobe suite like Adobe Illustrator, Adobe Photoshop, or art drawn on a Wacom tablet is widely accepted as legitimate forms of art. This raises the question of whether or not art generated by artificial intelligence (AI) should also be considered as such.

Hasan offers a compelling perspective on this debate. Instead of viewing his work as purely AI-generated art, he describes it as “AI-assisted art,” emphasizing the collaboration between the artist and the technology.

Throughout his artistic practice, the artist utilizes AI at various stages of his work. He often revisits his pieces multiple times during the creation process, refining them using MidJourney or by himself drawing with a stylus. After several iterations, each driven by the artist’s deliberate intentions, he achieves a final piece that satisfies him. By considering the role of AI as an assistant rather than a generator, this approach highlights the symbiotic relationship between human creativity and technological innovation.

Shiny, shares her experience of using AI as a valuable tool for curating and marketing her art. Shiny, a young artist, expresses her thoughts and emotions through brushstrokes on canvas, but selling her art requires a different skill set. She often struggles to find the right words to curate her work for potential buyers. Tools like ChatGPT, have helped her develop compelling narratives for her pieces, enabling her to better market and curate her work.

In both of these instances, AI serves as a partner of sorts, enhancing the artists’ creative processes and facilitating their success. This collaborative approach demonstrates that AI’s role in the world of art is not to replace human creativity, but rather to complement and amplify it.

“Challenging the Center-Periphery Divide: The Need for Inclusive AI Art Tools”

MidJourney, Style Transfer, and other AI tools perform well when tasked with borrowing stylistic influences from Impressionism or Cubism but struggle to replicate non-Western modernist styles, such as Kalighat. This limitation raises concerns about AI’s representation of non-metropolitan art and the potential exacerbation of the center-periphery relationship.

The French philosopher Jean-Paul Sartre once stated that Surrealism was stolen from the Europeans by ‘a Black (the poet Aimé Césaire) who used it brilliantly as a tool of Universal Revolution.’[1] The English art historian W. G. Archer wrote an influential account of Indian modernism. His analysis of the painting of Gaganendranath Tagore, one of the first Indian modernists, consisted almost entirely of tracing Picasso’s putative influence on him. Unsurprisingly, Archer drew the conclusion that Gaganendranath was “un cubiste manqué”; in other words, his derivative works, based on a cultural misunderstanding, were simply bad imitations of Picasso[2] Colonial mentality asserts cultural transmissions to be a one-way process flowing from the Occident. When artists from peripheral regions borrow from the Western canon, it is seen as a sign of inferiority. Conversely, when European artists borrow from other cultures, it is often described positively as “affinities” or disregarded as inconsequential[3] The AI tools’ inability to effectively engage with non-Western art styles may inadvertently reinforce these power dynamics. Since many AI algorithms are primarily trained on Western art styles, artists from peripheral regions may find themselves at a disadvantage when using AI to explore their own artistic heritage. This disparity risks perpetuating the notion that non-Western artists are secondary to their Western counterparts, echoing the past injustices of colonial mentality.

To challenge and counteract this, it is essential to ensure that AI tools are developed with a more inclusive and diverse range of artistic styles, representing the richness and plurality of artistic expressions from around the world. By doing so, AI-generated art can become a platform for deconstructing the power hierarchies that have long permeated the art world and foster a more inclusive global artistic landscape.

“Embracing Statefulness: The Double-Edged Sword of AI-Assisted Art”

Until now, the artist’s work, in some sense, has been stateless. That is, the artist and the canvas do not have a shared state. The canvas does not know how old the artist is, or what gender they are, or if they even believe in gender as a concept. For every piece, the artist has the chance to start fresh. A blank canvas. He does not have to paint in the same style today. And there is not one influencing him as such. The canvas is definitely not putting out a “suggestions for you to draw” reel, prompting, rather nudging him into a direction. He can paint anything in this world, or not in this world! For all one knows, he can be reborn today. But what if the canvas started giving paintbrush, or form, or line suggestions?

What if instead of a week now he only requires a day to produce his work because the canvas now has taken the load of making a million intentional decisions off of him.

Similarly, the photographer’s camera does not share knowledge of his owner’s mood. It does not know that today he is romanticizing a college campus by basking in the mess lawns and reading nietzsche. To capture this feeling into a photograph the photographer has to find the right frame, the right exposure, ISO, shutter speed balance, edit the image to the appropriate warmness, and do a gazillion things. But what if the camera now knows about his owner’s mood, and can predict some of his intentions, and make his job easy?

As we move towards a future where creators and their tools share states, we might see that artists no longer think through a medium, but rather think with the medium. We can already see this happening in text-based creations, with autocomplete, word recommendations, and AI-generated email responses in platforms like Gmail. The once stateless relationship between the artist and their canvas is evolving, allowing for a more intuitive creative process where tools can better understand and adapt to the artist’s intentions. While this transition brings undeniable benefits, it also raises concerns about monotony and the loss of distinctiveness in creative works. Platforms like Instagram, which once fostered discovery and diversity, now push users into specific directions, leading to a sea of repetitive content. Similarly, the modern web seems to be filled with websites that share the same design elements, typography, and color palettes.

The question we must ask ourselves is whether the efficiency gains brought by AI-assisted tools are worth the potential loss of individuality and diversity in art. As we embrace statefulness and the ever-evolving relationship between creators and their tools, it is crucial to strike a balance between efficiency and artistic expression to prevent a descent into monotonous, uninspired creations.

Transitioning Between Mediums and Revisiting McLuhan’s “The Medium is the Message”

A natural extension of stateful tools, seems to be the ability to change mediums more efficiently. Imagine a world where changing mediums becomes much more seamless. Transforming a photography campaign into a hand-drawn or poster campaign would take significantly less time and effort.

For example, a photographer who, after capturing a series of stunning images, decides to transform their campaign into a hand-drawn or poster campaign. The AI technology would analyze the original photographs and apply the artist’s distinct characteristics to generate hand-drawn or poster-style visuals, add catchy headlines, while maintaining the essence and vision of the original work .With AI-assisted tools that understand and adapt to the artist’s unique style, this once labor-intensive process could be accomplished with significantly reduced time and effort. AI tools could assist in this process, predicting the desired effect and guiding the artist in choosing the right techniques, colors, and textures to achieve their vision in the new medium. A month-long process requiring a staff of 50, could now be potentially achieved in 2 weeks with 5-10 people. The need to reevaluate Marshall McLuhan’s famous theory that “the medium is the message” becomes crucial. McLuhan’s theory posits that the medium through which a message is communicated exerts a profound influence on how the message is perceived, often overshadowing the content of the message itself[7]. However, with the increasing influence of AI tools on creative processes, we may see a shift in this dynamic.

In the era of AI, the lines between artistic mediums are blurring, and the tools used for creation start playing a more prominent role in shaping the message. Consequently, we might observe that the message is increasingly influenced by the tool itself (e.g., a large language model, MidJourney) rather than the medium in which it is conveyed. As a result, the tools become essential factors in the perception and interpretation of messages, significantly impacting the way audiences receive and understand them.

This shift in focus towards the tools might lead to echo chamber-like interactions among users and consumers of various creative forms, as similar tooling across different mediums can result in homogeneous content. AI-assisted tools could create a feedback loop, where creators and audiences become locked into a cycle of producing and consuming content that reflects the biases and limitations of the tools themselves(much like instagram reels). This phenomenon challenges the traditional understanding of McLuhan’s theory, as the tools now become an essential part of the message.

References:

  1. J.-P. Sartre, Black Orpheus, trans. S. W. Allen (Paris, 1951), p. 39, quoted in R. Linley, ‘Wifredo Lam: Painter of Negritude’, Art History, ii/4 (December 1988), p. 533. See L. S. Sims, Wifredo Lam and the International Avant-Garde, 1923–1982 (Austin, tx, 2002)
  2. W. G. Archer, India and Modern Art (London, 1959)
  3. The Triumph Of Modernism In India: India’s artists and the Avant-Garde (1922-1947)
  4. Chat GPT. I would like to highlight that this entire essay has gone through several iterations with chatGPT. While it has been mostly for proofreading and editing purposes, I admit chatGPT’s abilities of academic articulation are far better than mine. Hence, there could be instances of plagiarism, it is not intentional and I am not claiming the author’s words as my own.
  5. Podcast 1 with Shiny Parashar.
  6. Podcast 2 with Hasan.
  7. McLuhan, Marshall (1964). Understanding Media: The Extensions of Man. ISBN 81-14-67535-7.
  8. Midjourney: https://www.midjourney.com/home/?callbackUrl=%2Fapp%2F
  9. Chat GTPs creativity for “elephant in the room”

Acknowledgements:

I would like to thank professor Debayan for taking this wonderfully relevant and engaging course. To Shiny for bearing and sitting patiently with me for the first podcast while I was figuring out how to record and overcoming technical issues. To Hasan for his time and the conversation. To Bhanu, my little brother, who was instrumental in producing the second podcast.