Entangled Others, Artificial Remnants, 2019-2021, 3D GAN, 3D-Style Transfer, inflatable sculpture. Exhibition view at Aiiiii, 2021

 

Entangled Others, Artificial Remnants, 2019-2021. Exhibition view at Aiiiii, 2021

 

Birgit Mersmann: The emergence of AI image and video generators has rapidly transformed the production of art, questioned existing display, exhibition and marketing practices, and challenged existing notions of art. When we ask the artificial intelligence chatbot ChatGPT about the definition, impact and consequences of AI art, it concludes that “AI art challenges traditional notions of creativity, authorship and originality, raising debates about its role in the art world.” The positions within these debates are extremely controversial. “Artificial Intelligence can now make better art than most humans,” was the sub-headline of an article by Kevin Kelly published 2022 in Wired. Others analysts complain that due to the process of automatized image generation, AI art lacks human intentionality, emotion, and cultural context. And still others call into question whether AI art is truly creative, and whether it is legitimate at all to speak of AI art as a valuable art form due to the lowering of the creative barriers through image prompting. At which points does AI art challenge existing notions of art?

Xi Li: Artificial intelligence, as a novel medium distinct from those of the past, has the potential to redefine art. Since Duchamp’s urinal was exhibited in an art gallery, the definition of art has been a topic of ongoing debate. Many argue that what distinguishes an artist is their creativity, which is often seen as an inexplicable genius or inspiration. Today, machines are also engaging with this elusive element. Consequently, AI art challenges our current understanding and compels us to seek new definitions of human experience. Roland Barthes posited the “death of the author” and emphasized the inherent randomness in human discourse. Perhaps it is time to consider redefining art from the perspective of the viewer, where the audience determines what constitutes art. Much like nature, which exists objectively, its beauty is perceived subjectively by each individual.

Hauke Ohls: What is the creative potential of AI art?

XL: As a form of digital media art, the distinctive characteristic of artificial intelligence art lies in its data. There are many intriguing aspects within these datasets; they may originate from analog techniques, be derived from digital data since the advent of computers, or consist of personal, external, or public data. Moreover, the labeling of the data may reflect various human biases or even be illicit. The data can be vast in scale or, conversely, quite limited. They can pertain to bodily movements, coral, and much more. I believe these factors represent one of the creative potentials unique to artificial intelligence art.

HO: What are the limits of AI art, and what are the threats for the existing art system still mainly operating on the basis of the principles of the pre-AI era?

XL: AI poses challenges to our ability to discern the original works of artists, particularly for those of us who rely heavily on the internet. While we may not have seen all the works of an artist in person, the internet allows us easy access to them. However, due to the extraordinary mimetic capabilities of artificial intelligence, our ability to distinguish between original works and AI-generated imitations is compromised. In the face of numerous images of supposed original works online, we lack the criteria to make accurate judgments.

BM: What is the state and status of AI-generated art in contemporary Chinese art? Can you name leading Chinese artists within the art scene of mainland China that form the avant-garde of the AI art (r)evolution? What are their interests, topics, and aesthetic practices?

XL: Several Chinese art institutions are actively engaging in this practice. Notable projects include those at the Taikang Art Museum in Beijing, the How Art Museum and the Chronus Art Center in Shanghai.[1] Concurrently, several renowned Chinese artists are leveraging their established reputations to explore artificial intelligence research in alignment with their academic pursuits, such as Xu Bing with the Artificial Intelligence Infinite Film (AI-IF) Project,[2] or Cai Guo-Qiang with AI Model cAI™[3]. Xu Bing collaborates with scientists to employ various artificial intelligence models, enabling audiences to actively engage in the creation of their own film narratives. On the other hand, Cai Guo-Qiang has developed an inspiration library or dataset through the process of digital twinning, drawing upon his past creative database to foster growth; the cAI™ project evolves in tandem with the artist's development. Both artists possess a profound understanding of image generation models and natural language models, integrating these concepts into their respective creative practices. Additionally, emerging artist Sun Yuqian presents 1001 nights,[4] a game that reimagines the Arabian folktale collection in a rebellious manner where players collaborate with AI to write stories. Her research primarily focuses on language models. 1001 Nights encourages audiences to co-author stories with the king in the narrative, using the power of language to ensure the wife’s survival, presenting a work with feminist undertones. This practice is particularly intriguing as it diverges from the typical focus on generated imagery in AI art.

BM: In his book on Art and Cosmotechnics, philosopher Yuk Hui discusses the overtaking of art through recursive machines and the repositioning of truth in the artificial (ly generated). What concepts in Chinese philosophy can contribute to redefine art and aesthetics in the AI technosphere?

XL: The Western definition of art is influenced by ancient Greek culture and the evolution of rationalism. In contrast, traditional Chinese culture did not have a direct equivalent for the term ‘art.’ Take Chinese calligraphy as an example; it is considered the pinnacle of Chinese artistic expression. While the characters themselves are universally recognized and convey specific meanings, the artistic value lies in the manner of their execution. Here, aesthetic value takes precedence over content. As Yuk Hui points out, no single cultural perspective is inherently superior or inferior. In today’s context, it is essential to adopt a diverse range of perspectives to broaden the scope of artistic expression and human potential.

HO: Yuk Hui also states that AI is “prone to mutation” and considers this an important factor for AI to open up new dimensions in contemporary art. Apart from the concept of mutation, are there other strategies in AI art that are used productively?  

XL: Yuk Hui emphasizes the individualization of technology. Once artificial intelligence, as a tool, has developed, it becomes uncontrollable due to its relationship with the “associated milieu.” I believe this perspective provides a foundation for discussing the uniqueness of artificial intelligence through the lens of the relationship between technology and humans, as well as the consideration of various contexts. While this viewpoint is inherently grand, it is an essential and primary step for furthering the development of the associated issues.

BM: Parallel to the technological breakthrough and advancement of AI image generation, the institutionalization of AI art was set in motion targeting the promotion of emerging artists, a new, broad, and young-generation art audience and a large market share. The world’s first AI art gallery, named Dead End Gallery, opened in Amsterdam in March 2023, the world’s first AI museum “Dataland”, created by famous AI artist Refik Anadol, opens in 2025 in Los Angeles. Yet, the establishment of the Aiiiii Art Center in Shanghai happened earlier in response to the Covid-19 pandemic; it was founded in 2021 with the aim “to support, promote, as well as incubate both international and domestic artists and projects related to intelligent algorithms”[5] and became a pioneering institution for art and artificial intelligence. As you have been working as the director and co-founder of the Aiiiii Art Center, can you provide insights into the main idea, motivation, function and program of this cutting-edge museum institution?

XL: The artificial intelligence institution was primarily initiated by three professors from Tongji University, with Chinese digital artist Zhang Zhoujie serving as the principal initiator. Preparations for the project commenced in 2020, and it was officially established in 2021 with the support of Tongji University, College of Design and Innovation (D&I), and the Deqiu Art Park. Given that the initiators are all academic professors, an associated laboratory, the Art & Artificial Intelligence Lab at Tongji University, D&I, was also established to support this endeavor. Consequently, our initial positioning was as an academic institution. Simultaneously, because the mission of Aiiiii places a strong emphasis on practical application, we aimed to engage with genuine practitioners of art. Thus, the project was designed to support artists in their explorations of artificial intelligence, facilitating the realization of their ideas and possibilities.

BM: What do the 5 “i”s stand for? Is it an exclamation? A Dadaist lettrist play? Does it include a special meaning in Chinese?

XL: The five “i”s are somewhat arbitrary; the initial attempt to register a domain with three “i”s was unsuccessful, but a domain with five was available. In Chinese, “ai” phonetically resembles the word for “love,” which adds both significance and appeal to the name.

BM: What were the main challenges to exhibit AI art—for which a small screening generally suffices—in the extensive physical space of the Aiiiii Art Center, a redesigned power station?

XL: One of the main challenges in exhibiting AI art in the extensive physical space of the Aiiiii Art Center, a redesigned power station, is the lack of white wall surfaces due to the preservation of the space’s original appearance. This poses significant challenges for exhibitions, as smaller works can become overwhelmed by the vastness of the space. To address this, we collaborated with artists to create three-dimensional and volumetric representations of their works. Additionally, when Aiiiii was first established, many people, particularly artists and audience members, were unfamiliar with artificial intelligence. Therefore, we aimed to ensure that each work could interact with the audience, allowing them to understand the meanings expressed by the artists through these interactions. Audience members could take away virtual animals and stories as part of their experience.

BM: Have you developed particular exhibition strategies for the display and mediation of AI-based artworks?

XL: Yes, as mentioned earlier, we tended to collaborate with artists to create works that emphasize spatial narratives, focusing on the performative aspects of AI-based artworks within the space.

BM: Is it not self-contradictory to exhibit AI art in a physical exhibition space? In your curatorial view, what would be the ideal exhibition site and presentation mode for AI art? The so-called Metaverse? Does AI art require a ParaVerse, i.e. a parallel universe for manifestation? 

XL: I believe there is no inherent contradiction; rather, this reflects the inherent challenge faced by all digital artworks. Artists strive to allow audiences to experience their works within the physical spaces typical of contemporary art, making efforts to explore narratives beyond digital screens. Whether in virtual or physical spaces, the presentation of any artwork signifies a desire to communicate and engage with the audience. The most critical aspect is the continuous exploration of effective methods for expression and communication.

HO: Unfortunately, the physical institution of the Aiiiii Art Center has ceased to exist. The institutional investment ended in May 2023 due to different interests, positions, and conflicts. However, the promotion of AI art was continued by Aiiiii, a research lab for art and artificial intelligence supported by the College of Design and Innovation at Tongji University, Shanghai. How do you cope with this loss of a physical art space? What is the future agenda of the AI Factory?

XL: In response to the loss of a physical art space, the Aiiiii has maintained its focus on the academic research of AI art. The original framework involved conducting a single exhibition annually to engage with theoretical discourses emerging from conferences; thus, the absence of a physical venue has a relatively limited impact on our research agenda. While this loss is indeed significant, it may also signify a transition into a new phase for the Aiiiii. Previously, the landscape was defined by a lack of familiarity with AI art among both artists and the public, necessitating ongoing efforts to cultivate awareness and understanding of the subject. However, this definition has become increasingly ambiguous as AI-generated art has proliferated. Consequently, our discussions have evolved within the broader context of art and technology, allowing the absence of a physical space to paradoxically afford us greater opportunities for reflection and exploration.

The Aiiiii has consistently prioritized cross-disciplinary collaboration with academic institutions and educational entities to facilitate rigorous and meaningful inquiries into AI art. We are actively developing collaborative projects with various organizations, emphasizing Asian philosophical perspectives and scholarly interpretations of artificial intelligence to investigate a more diverse range of viewpoints. 

BM: For defining the mission and realm of the Aiiiii Art Center, a Da(t)aist manifesto entitled “Git-festo about AI and Art” was published—which can be found on the website https://www.aiiiii.com/. Why put the Aiiiii Art Center into the legacy of the Dadaist avant-garde movement, its manifest-driven revolutionary, destructive, non-sensical impulse of taboo breaking and border crossing between art, technology and life?

XL: The original author, Professor Kostas Terzidis, draws a parallel between artificial intelligence and Dadaism, emphasizing the black box of artificial intelligence that processes diverse datasets to generate unexpected poetic expressions. This creative process is analogous to the method employed by Tristan Tzara, who would cut and rearrange fragments of newspapers to construct new poetic forms.

Along with the manifesto, there is Dürer’s print of the rhinoceros, which was produced based solely on a description, as Dürer had not observed the animal firsthand. Consequently, this print serves as a significant reference point in our examination of AI art.

BM: In the final, 12^th statement of the git-festo, China is described as “the perfect place for exploring new territories. If not us, then who? If not now, then when? If not here, then where”. What conditions, structures, attitudes and policies in China provide a good breeding ground for advancing the research on art and artificial intelligence and for the implementation of novel approaches to exhibition curating in the AI-era?

XL: The rapid development of artificial intelligence in China contrasts sharply with the environment in Europe, where the adoption of new technologies may occur at a different pace. In China, the societal impetus for economic development fosters a heightened receptiveness to technological innovations. This sensitivity is further supported by a relatively open governmental stance toward the application of technology. Additionally, the demographic concentration in megacities generates an unparalleled volume of data and information. Given these unique characteristics, we contend that China represents a particularly fertile context for discussions surrounding artificial intelligence, offering significant potential for exploration and advancement in this field.

BM: The Bucharest Biennale 2024 was the first art biennale curated by artificial intelligence, i.e. the AI JARVIS program. What is your opinion about art curation with AI? Will AI-curated exhibitions make the profession of the curator obsolete?

XL: In my view, the challenges posed by artificial intelligence to the field of art are contingent upon the definition of curation. Curation can encompass both rigorous academic research and the expression of personal opinion. Consequently, the advent of AI is likely to catalyze a transformation in the conceptualization and practice of curation, prompting a reevaluation of its parameters and methodologies.

HO: Artificial life, altered nature, more-than-human world, and wetware seems to be a key focus of your work at the AI Factory, with three events on this topic. How did this focus emerge?

XL: In the course of our research, we observed that many artists base their work on biological data. Additionally, we found that many discussions center around questions of authorship and the comparison between artificial intelligence and human creativity, which do not advance the discourse. Once the definition of humanity is clarified, these issues become unreliable. Therefore, we believe this pertains to a perspective on life, akin to the influences of Darwinian evolution. Just as nature itself may not recognize its own beauty, humans do. Consequently, we aim to explore how artificial intelligence can foster human development from its ontology aspects, rather than merely continuing the trajectory of alienation as described by Marx.

HO: Can AI contribute to a new understanding of ecology?

XL: I believe that it is essential for artificial intelligence to advance, but this progress necessitates a collective effort. Central to this challenge are issues related to public perception, the dynamics of consumerism, and the evolution of socio-economic systems. The relationship between humans, machines, and nature has long been a subject of discourse; however, advancements in this area have been slow. The emergence of artificial intelligence serves as a stark reminder that a crisis exists, highlighting the urgent need for a more proactive approach to these interconnected concerns.

HO: What is the relationship between AI, synthetic biology and art in relation to artistic practices?

XL: I acknowledge that my understanding may still be incomplete; however, I would like to provide an example to illustrate my point. One key term associated with artificial intelligence is “neural network.”  However, when a cell is extracted to cultivate a neural network in vitro, it raises a question about what constitutes the artificial in this context and how such networks relate to human beings.

HO: How do you assess the separation that is still drawn between the natural and the technical? Does AI art and aesthetics have the potential to bridge this separation?

XL: Both nature and technology possess a formidable capacity for growth, akin to human development. Drawing parallels with the art forms we previously discussed, I believe this resonance will persist. The potential for artificial intelligence in art to expand remains uncertain; however, we are eager to explore this dynamic further.

 

 

Jake Elwes, The Zizi Show, 2020, video installation. Exhibition view at Aiiiii, 2021

 

Obvious, Works from La Famille Belamy, Facets of AGI and Trials and Errors. Exhibition view at Aiiiii, 2021

 

Certain Measures, MTS_003, 2021. Exhibition view at Aiiiii, 2021

 

Exhibition The Book of Sand, 2021. Exhibition view at Aiiiii, 2021. Photograph by Kai.

Curator Xi Li is the art director and co-founder of Aiiiii Art Center (www.aiiiii.com) since 2021. She is also an assistant professor at the School of Future Design, Harbin Institute of Technology (Shenzhen Campus). One of her leading projects is aai International Conference on AI Art (2021, 2022), as well as the exhibition The Book of Sand (2021). She holds an MA in Narrative Environments from Central Saint Martins in London and a BFA in Art Management from the Central Academy of Fine Arts in Beijing.

Birgit Mersmann is Professor of Contemporary Art and Digital Image Cultures at the University of Bonn, Germany. Her interdisciplinary research covers modern and contemporary Western and East Asian art, global art history, migratory aesthetics, museum and exhibition studies, digital art, image and media theory, visual cultures and visual translation, interrelations between script and image, and history and theory of photography. Recent book publications include: Kritik des Neo-Extraktivismus in der Gegenwartskunst (ed. with Hauke Ohls, Lüneburg 2024); Image Controversies. Contemporary Iconoclasm in Art, Media, and Cultural Heritage (ed. with Christiane Kruse and Arnold Bartetzky; Berlin/Boston 2024); Okzidentalismen. Projektionen und Reflexionen des Westens in Kunst, Kultur und Ästhetik (ed. with Hauke Ohls, Bielefeld 2022); Über die Grenzen des Bildes. Kulturelle Differenz und transkulturelle Dynamik im globalen Feld der Kunst (Bielefeld 2021); Bildagenten. Historische und zeitgenössische Bildpraxen in globalen Kulturen (ed. with Christiane Kruse; Paderborn 2021); Handbook of Art and Global Migration. Theories, Practices, and Challenges (ed. with Burcu Dogramaci; Berlin/Boston 2019). 

Hauke Ohls is a postdoctoral researcher with the Chair of Contemporary Art and Digital Image Cultures at the University of Bonn, Germany. His research focuses on theoretical, sociological, and philosophical questions of modern and contemporary art with particular emphasis on eco-relational art and ecological aesthetics,neo-extractivism, ecofeminist, pluriversal, posthuman theory, the discourse on objects, materiality and images, as well as the relationship between art, economy, and neoliberalism. Additional areas of interest include the intersection of art and music, artists’ writings, media art, and transcultural art history. Recent book publications include: Kritik des Neo-Extraktivismus in der Gegenwartskunst (ed. with Birgit Mersmann, Lüneburg 2024); Many-Valued Aesthetics. Interconnections in the Work of Mary Bauermeister (Bielefeld 2024); Okzidentalismen. Projektionen und Reflexionen des Westens in Kunst, Kultur und Ästhetik (ed. with Birgit Mersmann, Bielefeld 2022); Objektorientierte Kunsttheorie. Graham Harmans spekulative Philosophie im Kontext einer (nicht-)relationalen Ästhetik (Hamburg 2019).