beyond policy lab 5
The Canvas that Gazes Back: A Live Lab for Artists on AI and the Intelligence of Making and Creating
The group began with a round of introductions by sharing their names, the artforms they work with, their training and learning journeys, and one thing art taught them about power and responsibility.
Riya Nagendra works with illustrations and comic art, and traditionally works on paper with water colours, and more recently with digital art because of ease of collaboration and editing. Mostly self-taught, Riya grew up reading a lot of books and comics and was encouraged to draw as a child. Discovering that she was good at it, Riya kept at it over the years. Making and doing art has taught her that stories carry a lot of weight, particularly around how we tell stories, who we tell stories about, the language we use, and who we are focusing on.
Ankita Chauhan picked up painting as a visual arts medium about a decade ago, as an activity that she thought she could do without overthinking. It wasn’t planned but emerged alongside her corporate career. Ankita works with oil paints and acrylics. Her learning has happened by symbiosis and observation. Art helps her reflect on deeper questions around sustainability and climate change, and what it means to be alive and to express oneself while carrying a carbon footprint for each of those things.
Piyusha Vir works with mixed media and fluid art, and learned from Rhiti Bose, Anuja Agarwal, and Ritesh Singla, while most other aspects of her learning emerged from a self-taught practice. She learned a range of different art styles and now works with acrylic and alcohol ink and resin art. Rhiti’s teaching led Piyusha to recognize the importance of making art without taking space from other artists, especially those for whom art is embedded in their livelihoods and historical realities. She thus refrains from making Indigenous artforms for sale and makes it a point to recommend Indigenous artists instead. Her journey into art also moved her to reflect on making art eco-friendly, such as using UV-stabilized resin that does not yellow very easily and lasts long.
Rhiti Chatterjee Bose works with folk and tribal art forms and uses natural ingredients or the safest options closest to natural ingredients. She also works with water colours, acrylics, and gouache. Rhiti learned from renowned artists Dr Shankar Chatterjee (also her father and first guru) and Sri Sunil Guha. She has also self-taught other art forms. Rhiti’s learning from her journey so far is to be mindful as urban artists, to not take away or appropriate from those whose livelihoods depend on artforms, and to give oneself time to grow as an artist through practice, learning, and exposure. Rhiti works with artists who are looking at finding a second career after a break, and guides rural artists in showcasing their work to get better sales. She is also an art teacher and guides diverse groups.
Geetha Srinivasan remembers always having been imaginative as a child and enjoyed reading. A self-taught artist, art was a way for her to express the imagination that reading and life unlocked, but what she expresses on paper or canvas is about 10% of what her mind imagines, and she keeps at it. She admires art in every form she encounters and finds art very healing. She tends to get more out of a painting as it helps her understand and reflect in a way that feels right for her.
Priyanka Kotian works with water colours. Growing up, though she was good at drawing, she didn’t pursue it further because she didn’t have the guidance she would have liked to know how to proceed. Her learning has largely come about because of access to tutorials online, on Pinterest, YouTube, and SkillShare. She notes that if it wasn’t for artists who put out their work online in the way they did, she would not have had a place to start. She finds that if we did not have the constraints of work, time limitations, and other commitments, we would all indulge in art. Looking at the oldest form of art, cave paintings, she notes that it started because they wanted to record something. She is part of a group of Urban Sketchers who go around town, sketching what they see. To her, art is very inherent in humankind and has evolved with time. She does not believe in gatekeeping art and believes that while every community has its own art, a person taking an artform and creating something that incorporates their own personal contexts and symbolism, is also welcome. She feels sad for how artists who make a living out of art are affected by the advent of AI, which in her view, lacks soul because it does not have a human element. She considers AI art to be a mix of things, and more content than art.
Naming and framing the problem
The group began with some scene setting that looked at the way AI is built and operates. Trained on the data including art from all over the world, AI produces content after being trained on data, where it learns from patterns to make probabilistic outputs. The group noted that this journey of accessing data for training has also included multiple iterations of data extraction without the consent of data owners. This has included the works of artists across space and time. On the use side, the group reflected on how it has become easy for people to ask an AI tool to produce images and videos that mosaic multiple artforms, that the lingering question is whether one is an artist when what they are actually doing is prompting. They also noted how the images generated flatten complex realities, histories, emotions, experiences, and personal perspectives that come with human art production.
On the other hand, the group also made space for the truth that AI has been helpful in enabling certain groups with making, creating, and preserving their artforms, such as Kite by Dr. Suzanne Kite, Michael Running Wolf's First Nation Languages AI, and PolArctic by Leslie Canavera, Te Hiku Media of Aotearoa, and Adi Vaani. They also noted how AI has helped people with disabilities produce art in ways that traditional tools have not: Such as M Eifler, who created Prosthetic Memory, a personalized homemade AI system designed to act as a computational extension of their brain; Pete Aaberg who uses Midjourney to create art after a major cerebellum stroke paralyzed the left side of his body; Sarah Ezekiel (diagnosed with ALS) and Becky Tyler (diagnosed with quadriplegic cerebral palsy) who use Tobii Dynavox Eye-Gaze trackers, predictive AI algorithms, and digital drawing pads to make art; and Tony Quan, whose entire body has been paralyzed by ALS, uses The EyeWriter to produce art.
With this in place, the group was invited to bear three overarching, guiding questions in mind in reflecting on the next piece on governance:
What does your practice know about the difference between copying a form and understanding it and what does that tell us about what AI is actually doing?
Your art exists in relationship: to materials, to community, to place, to the sacred, to the living and the dead. What does AI governance look like if it is required to account for those relationships?
Artists have always navigated questions of influence, appropriation, ownership, and originality. What have you already understood about this that the people governing AI haven't thought to ask artists?
Creating a Repository of Wisdom
Each member in the group was invited to share one concept, practice, principle, or embodied understanding from their work with the arts that speaks to governance. The themes that emerged were:
Acknowledging and crediting the intuitive process of creating art based on experience, relationships, and human connection.
Sharing personal experience and emotion to create art
Adding disclaimers and disclosure texts to name whether images were AI generated or human generated
Understanding the history and evolution of the artform
Art does not need to be perfect to be celebrated: The flaws and rawness are what make art beautiful and stand out and human challenge, the joy of creation, and finding peace are more important than chasing perfection
Teaching the arts and encouraging creation
The concept of servicein art, where making is an act of service rather than self-expression.
The living geographies of tribal artforms, where every image is an ecosystem in itself, a whole cosmology and a map of relationships between human, animal, plant, and spirit life.
The role of gaze, both for the creator and the perceiver. certain kinds of seeing are extractive. Looking without relationship, consent, or care is a form of harm.
Collective authorship, where individual virtuosity thrives within and because of a community matrix, which subverts the myth of the lone creator
Art is made to be seen and to move you to do something, either to protect, invoke, heal, or take some action.
Building a Container for Governance
The group began by reflecting on the critical question of AI producing art in itself, and whether it can be considered art. They noted that even as humans, we produce art that is inspired by the works of those who came before us, by observing the worlds we inhabit and navigate, and all that we learn from artists who produce and share their works of art. The group noted the importance of understanding that what we create this way still has a part of us as artists, which is the essence of what an individual contributes through their own experience, understanding, technique, and history. The group thus called for preserving the human element meaningfully.
Acknowledging that art exists in material and non-material relationships to materials, community, place, the sacred, and to the living and the dead, the group noted that AI has an existential crisis. It is taking over spaces where it is not needed. It could well be used to make advances in science and medicine, but instead, it is being used as a tool for capitalism. It is not engaging with materials, people, environments, or lived realities meaningfully, but extracting from all of these. This is mainly because those who build and create these AI tools are wealthy, powerful, and guided by profit as the core motive. Responding to this, the group called for attentiveness to power and control, and how these dynamics shape who makes AI for whom, for what, for how long, in what way, and at whose cost.
The group made a compelling argument for the preservation of authenticity of expression. While innovation is always welcome, there should be strong protections for the authenticity of an artist’s voice. The people behind the technology, the group noted, seldom care for the collective legacies an artist carries. The neoliberal approach of hyper-individualism does not make a value addition here, because this is where the flattening emerges. Instead, assigning collective authorship, as embedded in traditions like Pattachitra and Kalamkari, where individual virtuosity exists within and because of a community matrix. This subverts the myth of the lone inventor and creator that thrives in the space of AI art creation. This would also acknowledge the rich legacies that hold up all forms of art, given that all efforts of learning of art have been made possible because of artists who came before making a choice to put their art out for the world to witness, experience, and learn from. The contribution to one person’s art comes from several other people past and present, and the notion of individualised intellectual property does not offer as much by way of governance. The journey of making art by learning from all the influences around you is humbling and beautiful: Producing pieces that speak to the ethos of those it is meant for is the human touch at work.
Taking this further, some in the group made the case for collaborative engagement with AI, where AI is a tool and nothing more. One called for retaining the human element over AI, where the tool can generate ideas but the human should produce the final art product with some measure of their own wisdom and practice. Another made the case for a consent-based data sharing process where watermarked and signed work by artists are out of bounds as training data, but artists can be commissioned to produce samples in exchange for a fair compensation, with a clear agreement with them on what those samples will be used for and how the AI tool will be used, and what it might mean to consent to training a tool on that artwork.
Drawing from the living geographies of tribal artforms, where every image is an ecosystem in itself, the group called for moving away from flattening real. The living geography of tribal art, such as Gond and Warli, guides us to understand that art is neither linear nor unidimensional. This is also true of human life. By recognizing that every image is home to a wholesome ecosystem, a cosmology, a map of relationships between human, animal, plant, and spirit, it is important to govern a technology that reduces image to a mere pattern through strong guardrails that protect complexity and do not datafy a vast, interconnected social life into a single datapoint. In expanding on this interconnected reality, the group called attention to preserving and protecting theecosystem of art from being erased by AI. In the global rush to act on techno-optimism from the creation and use of AI, we are moving toward a reality where libraries, art galleries, and art studios are being sacrificed while more and more investments are going into data centres. These data centres are also actively erasing nature’s artistry – where trees are being cut, water is being used, and more heat is being generated and affecting other species, all in the name of building this technology.