Talking to Joanna Zylinska.
The artificial intelligence in artistic creation

An interview with Joanna Zylinska by José Vertedor

Joanna Zylinska. La inteligencia artificial en la creación artística.
Entrevista a Joanna Zylinska por José Vertedor

Joanna Zylinska

Department of Digital Humanities, King’s College, London, UK.

José Vertedor (Ed.)

Universidad de Málaga, Málaga, Spain.

Ignacio López (Trans.)

Universidad de Granada, Granada, Spain.


We interviewed Joanna Zylinska, Professor of Media Philosophy and Critical Digital Practice at King’s College London, who is a leading researcher in digital media and technology studies, focusing on the intersections between culture, technology, ethics and art. Her work explores how technology influences our perception of the world and our understanding of contemporary art. The interviewee offers a critical and reflective perspective on the relationship between artificial intelligence (AI) and art, establishing herself as a leading voice in the field.

The interview delves into the intersection between Art and AI, questioning its impact on creativity and artistic ethics. The key points addressed include Zylinska’s early critique of AI art, which she deems as technical exhibits lacking artistic depth, and the maturity and evolution of the artistic field. The author acknowledges the progress and maturity of AI art and highlights the importance of questioning who creates art, for whom, and for what purpose.

The controversy surrounding the term ‘AI Art’ is addressed, and alternatives such as ‘computational creativity’ are proposed. The article explores how AI challenges pre-existing cultural and financial notions of art. Ethical reflections are also presented, challenging the notion of ‘ethics in AI’ and arguing for a deeper ethical engagement in artistic practice involving AI.

Additionally, the interview features ethical reflections, questioning the notion of ‘ethics in AI’ and advocating for a deeper ethical engagement in artistic practices involving AI.

Key Words: AI Art, Ai-driven art, Machine Visions, Machine-Learning, Anthropocene, Creativity and AI.


Conversamos con Joanna Zylinska, catedrática de Filosofía de los Medios y Práctica Crítica Digital en el King’s College de Londres. Como destacada investigadora en estudios sobre medios digitales y tecnología, su trabajo examina las intersecciones entre cultura, tecnología, ética y arte, y cómo la tecnología moldea nuestra percepción del mundo y nuestra comprensión del arte contemporáneo. Su perspectiva crítica y reflexiva sobre la relación entre la inteligencia artificial (IA) y el arte la consagra como una voz referente en este ámbito.

Durante la entrevista, exploramos la intersección entre el arte y la IA, analizando su impacto en la creatividad y la ética artística. Zylinska compartió sus primeras críticas al arte con IA, que consideraba meras exhibiciones técnicas carentes de profundidad artística, y analizó la evolución y madurez del campo artístico. De igual forma, reconocía los progresos realizados en el arte con IA y subrayaba la importancia de cuestionar quién crea arte, para quién y con qué propósito.

También se abordaba la controversia en torno al término “arte con IA” y se proponían alternativas como “creatividad computacional”. El artículo ahondaba en cómo la IA desafía las nociones culturales y financieras preexistentes del arte

Además, la entrevista se cierra con una reflexión sobre las implicaciones éticas asociadas, cuestionando la noción de “ética en la IA” y abogando por un compromiso ético más profundo en las prácticas artísticas que implican IA..

Palabras Clave: Arte de la IA, Arte impulsado por la IA, Visiones de la máquina, Aprendizaje de la máquina, Antropoceno, Creatividad e IA.

cómo citar este trabajo / how to cite this paper

Zylinska, J. (2023). Talking to Joanna Zylinska. The artificial intelligence in artistic creation. An interview with Joanna Zylinska by José Vertedor (J. Vertedor, Ed.; I. López, Trans.). Umática. Revista sobre Creación y Análisis de la Imagen, 6.



José Vertedor

Joanna Zylinska


Sin financiación

Received: 15.12.2023
Accepted: 20.12.2023

The use of artificial intelligence in artistic creation

In your book AI Art: Machine Visions and Warped Dreams, you explore the relationship between AI and art. How do you think AI is changing how we create and experience art?

In my book I was quite critical about many of the actual outputs produced during the first phase of AI hype, in the early 2020s: garish, GAN-generated artefacts, situated somewhere between a screensaver and a kitsch visualisation for a meditation app. The main goal of those works was more often than not to showcase the technical prowess of their makers –and their sponsors from the Big Tech. However, I do recognise that the field is now maturing and that there are a lot of exciting developments around AI art. We also need to note that the term “AI art” is itself contested. It would perhaps be more accurate to speak about “art enabled by machine-learning technology”, although terms such as “creative AI”, “computational creativity” or simply “media art” are also in use. But let’s stick with “AI art” as shorthand.

One of the most interesting things for me that AI is doing right now is raising the question, anew, of what art is –and what it is for. Equally importantly, who is it for? AI art makes us look again at the cultural and financial values we attach to art, at the class-based definition of culture, at ideas or property and propriety. The outpouring of AI-generated artefacts by the likes of Midjourney or DALL-E 2 is raising a lot of anxiety today, because it puts our human notion of creativity under a spotlight. It makes us think: if a machine can do that, then perhaps what I thought of as my unique human characteristics, an expression of my true soul or self, is not that unique at all. By saying this I’m not trying to ignore justifiable concerns with regard to the future of creative labour and creative education, or get the AI companies that have plundered human and natural resources, without care or remuneration, off the hook. But we do need to situate AI-generated art in the longer context of art produced with nonhuman agents: impulses, viruses, drugs, as well as all sorts of networks –from mycelium through to the Internet. I am hopeful AI art will get us to think again about what kinds of artistic and cultural outputs we value, and about how we can sustain the institutions and infrastructures that cultivate an artistic mindset, or a creative way of being. I don’t think we will stop producing art only because Stable Diffusion can spit out infinite pixel mashups at enormous speed. But we will need to take things in a new direction, in the same way that painting had to rethink itself as a medium after the invention
of photography.

Fig. 01. Cubierta del libro AI Art: Machine Visions and Warped Dreams.

You have written extensively about nonhuman photography. For people who are not yet familiar with your work, which we encourage them to do, could you briefly explain what “nonhuman photography” means to you?

The concept of nonhuman photography refers to photographs that are not of, by, or for the human: (1) depopulated landscapes (e.g. works by Robert Capa, Andreas Gursky or Trevor Paglen), (2) images taken by satellite cameras or endoscopic equipment, and (3) QR codes or photos of human faces used in automated security systems, helping those systems “decide” who to let in and who to keep out. By proposing this concept I’m not saying that there are no humans involved in the production of photography. I just want to point to an entanglement of human and nonhuman elements in the constitution of the world. My goal is also to draw attention to some inhumane uses of photography in our socio-political lives. The concept of nonhuman photography has arguably become even more potent in the era of generative AI, where text-to-image models trained on vast datasets of photographs are used to create photorealistic images. There are multiple processes of agency instantiated here, which need exploring –and also, frequently, contesting.

To sum up, the concept of nonhuman photography allows us to situate our human history in the longer trajectory of nonhuman processes, whereby tanning or fossilisation could be read as forms of proto-photography. With this, we are recalibrating the human to a different scale, with a view to helping us move away from the species-specific narcissism through which we have awarded ourselves a dominant position in the world –and a misguided power to be its shaper and destroyer.

You have produced extensive work on nonhuman photography and how we should understand the Anthropocene. If we focus on these two concepts, we would like you to tell us how to combine these two principles in your work.

Much of the work on the Anthropocene has been developed in the context of trying to understand our human responsibility for our planet. Humanities scholars have been particularly concerned in recent years with how this rethinking of human agency beyond the boundaries of our human body and mind demonstrates what many non-Western ontologies have known for a long time: the fact that we are of the world, that we are co-constituted with it. This relational understanding of our being turns the question of responsibility and ethics into the only viable response to the world and its affairs, be they environmental or socio-political. The situation of the climate crisis, which is a cognate term to the Anthropocene, makes this response particularly urgent. Images, especially technical images known as photographs, play an important role in our understanding of our situatedness in the world.

Importantly, we need to see photographs not only as figures of representation and visualisation but also, more importantly perhaps, as figures of imagination. Looking at their antecedents in fossils, as discussed earlier, we can connect the practice and technology of photography with the working of the Sun, and, through that, with processes of pollution, radiation, fossil fuel depletion and extinction. It is in this sense that the problem of the Anthropocene can be connected with the planetary demand of the present moment. Nonhuman photography can become a thought- and image-device that can help us approach this problem, not just conceptually but also materially, physically –through our fingers, noses and lungs.

The present and the future of artistic creation with digital media

Could you tell us more about your current research into human and machine intelligence and perception?

I have recently published a book titled The Perception Machine: Our Photographic Future Between the Eye and AI. It is a follow-up to Nonhuman Photography, but it takes the problem of the human, of human perception and understanding, as its central axis. I am particularly interested there in how the recent developments in, and articulations of, (supposed) machine intelligence call on us humans to rethink our ideas and values when it comes to how we see ourselves and the world, how we build our knowledge systems, and the role of images in the construction of those systems. The notion of the perception machine I propose is a metaphor for different layers that we can use to describe the current organisation of our society and polity, and of the human and nonhuman entities and organisms that inhabit them. The perception machine can describe a condition whereby we are all constantly seeing and being seen, not just by other humans but also by mechanical eyes, from surveillance cameras through to the operations of machine vision –and their functionaries. I also explore how human perception is changing in and through our relationship with machines. Last but not least, I examine the working of image-making technologies, from cameras as either standalone devices or functions included in smartphones through to image-generation programs and models. The book’s title transposes Paul Virilio’s Vision Machine in an encounter with Vilém Flusser’s work on the future of various media (photography, books), to produce my own take on media philosophy. It also offers a feminist attempt to respond to the current planetary demands, beyond all sorts of bombastic salvationisms, be they of a philosophical or military kind.

How do you think the planetary crisis affects how we imagine the future, and how does this elate to your work on media art?

The situation of the deep crisis our planet finds itself in on multiple levels can induce a sense of paralysis, the inability to not just act but also imagine future possibilities. I think art, and media art in particular due to its explicit technological kinship, can help us become unstuck. It can help us search for new ideas and new articulations, beyond just the mournful celebration of the world in crisis. (And, truth be told, we have had plenty of exhibitions of this latter kind in recent years, wallowing in the pleasure of the ruin –which itself is quite a well-known artistic trope, albeit a rather disabling one.) As well as working philosophically, I have an art practice, which for me is a space to mobilise a different part of the sensorium and a different sensibility, to feel and think otherwise. All in all, I write and think with an awareness of only ever being able to offer partial views and fragmentary responses, but this multimodal way of working is an attempt to present a collage of images and ideas out of these fragments, while making the stitching and the glue visible in both my texts and images.

How do you see the future of art in the era of computation?

I would like to play the devil’s advocate here and suggest that human creativity has perhaps always been algorithmic, that it has relied on the execution of various sets of rules known as programs –even if the most interesting outputs have been those that have involved an algorithmic glitch. Computation as an enactment of algorithmic technology and thinking by differential engines that we know as computers is only the most recent stage in this process. Having said this, computer-enabled art, and in particular art driven by technologies of machine learning, has opened up new horizons and new challenges, which we discussed earlier. I am quite excited about the possibilities and also by the questions raised by this development. But I also believe we need sustainable art education and responsible creative policy on a national level if we are to avoid having the discourse around art hijacked by Big Tech, with art and creativity reduced to the mindless production of computationally-generated artefacts.

A new paradigm in teaching methodologies in the field of art

Could you explain how you use digital media and artificial intelligence in teaching and creating fine art?

In my own practice I’m interested in exploring what artist David Young has called “little AI”. In other words, I want to draw attention to the singular moment in time when the technology is still quite imperfect, when it reveals problems rather than being completely seamless. This is most evident in the film I made in 2021 titled A Gift of the World (Oedipus on the Jetty). It is a remake of Chris Marker’s famous photofim, La Jetée. To make it, I extracted the still images from Marker’s film and had them remade with the help of a GAN model. I also fed the film’s famous opening line, “This is the story of a man marked by an image from his childhood”, to the GPT-2 language model, which subsequently produced its own version of the script in response to that sentence. Because generative technology, both on an image and text level, was still in its infancy, the outputs were very glitchy, yet also, from my point of view, artistically interesting. The script obtained ended up having a vortex of genders and pronouns, rewriting the very masculine story of salvation present in La Jetée as a gender-fluid polyvocal counter-apocalypse. This rather light-hearted and low-tech approach to art and art-making is also visible in my teaching.

How is this actually reflected in your teaching, and how do you instil these ethical principles in your students as you teach them to use digital media and artificial intelligence in their art?

It goes without saying that, rather than ban students from using AI, I actively encourage them to experiment with the technology. The goal, however, is not any kind of technical perfection but rather the ability to understand the rationale behind AI models as well as their limitations. For example, on the MA course I teach called The Digital Image, students have to produce a practice-based assignment exploring media ecologies, or considering the possibility of us all living in the perception machine, by using a platform and a set of tools of their choice. The work has to be accompanied by a research-based essay, offering a critical reflection on the creative choices while anchoring them in the wider theoretical debates. Questions of ethics are important both in my pedagogy and in the assessment. But I’m referring here to something different from the narrowly conceived ‘AI ethics’, which often takes the form of procedural cleansing to pretend the companies involved are a force for good, rather than encouraging any deeper investigation of the problem of responsibility in a world in which the plural Other –both of the human and nonhuman kind– always precedes me and makes a demand on me (I derive this idea from the philosophy of Emmanuel Levinas, although I rework it through the critical posthumanist perspective). This is why I believe we need serious engagement with ethics, full stop –not with its truncated version called ‘AI ethics’.

In this respect, could you indicate any methodology to apply these technologies to artistic practice and teaching?

My methodology is probably best encapsulated by my job title: I’m currently Professor of Media Philosophy and Critical Digital Practice at King’s. So my approach involves combining thinking and making. This is done in full recognition of the fact that different media have different affordances that allow us to accomplish different things. But I also aim to use practice –especially, in my case, image-based practice, although driven by conceptual concerns– to open up a different way of thinking and sensing. I try to convey this methodology to my students, through the material we read and look at, the format of the classes and the assessment we do.

The ethics of artistic creation with artificial intelligence

In your opinion, must some ethical considerations be taken on board when developing powerful AI models?

Yes, absolutely, although these considerations need to be socio-political as well as ethical. As mentioned before, I am rather suspicious about the developments around AI ethics, or rather about their rationale. As long as tech companies can tick the ethics box by showing they have ‘considered’ issues of bias, representation or hate speech, they are deemed to be on the side of the angels and are allowed to carry on with whatever they are doing, without having to bother with deeper problems of injustice, inequality, racism, sexism, untruth and violence that their products may strengthen. There are therefore some deeper, or we could even say fundamental, questions that are not addressed as part of such an ethical enquiry: if they were, we wouldn’t need to speak about AI ethics but rather about ethics as such. The widespread use of (so-called) AI does of course call for a specific response to this set of developments, but I’d argue it’s on the level of regulation and policy that such responses would be most successfully executed, with benefits to large sections of society. Yet companies that develop dominant AI models and tools tend to loathe regulation, equating it with the big state, the stifling of innovation and curbs on their profits. This is why they come up with toothless, though nice sounding, ethical principles while allowing themselves to carry on with their damaging activities, until someone (the European Union, the US congress, trade unions) tells them to stop.

How do you address ethics in contemporary artistic creation using digital media? How do you balance creative possibilities with ethical concerns?

As shown in my earlier work on ‘minimal ethics’ (Minimal Ethics for the Anthropocene), I am principally interested in non-normative forms of ethics –that is, forms of ethics that don’t offer any upfront catalogues of rules. (The reason for this is because I don’t think ethics on its own, as an articulation of how people should live, works without quickly turning into a form of moralism or control.) Ethics needs a political supplement, a way of working out of the strategies and ways of being in a society in which there are multiple demands, from multiple subjects, and multiple groups of subjects, with power differentials. What politics needs, in turn, is a horizon of justice (to be constantly worked on and out) and a minimal condition of responsibility, which we talked about earlier. I believe these principles can be applied to artistic creation that uses digital media, including AI. Questions need to be raised about justice –although, in the case of labour practices and the unauthorised use of artists’ works to feed the databases of machine learning while trying to devalue art and creative practice as social non-purposeful activities –we are once again in the realm of politics and policy.

Could you share your perspective on how ethics influences your art-making process, especially in digital media and artificial intelligence?

Let me give you an example from a project I developed for the AI Art book. In 2018, in an uncanny anticipation of the world-become-window situation of coronavirus capitalism, I made a photo-film called View from the Window. It involved me hiring 100 workers from Amazon’s Mechanical Turk (MTurk) platform, which is an online marketplace connecting labour suppliers with providers worldwide, to each take a photo of a view from a window of the room they were in. The MTurk platform, informally called ‘artificial artificial intelligence’ by Amazon, puts humans in the role of machines, with workers all over the globe being available for hire to execute simple yet boring tasks such as tagging photographs or doing surveys –tasks that would be too expensive for a company to program a computer to fulfil.

Fig. 02. Cubierta del libro The Perception Machine: Our Photographic Future Between the Eye and AI.

MTurkers are only known by their numerical IDs, thus functioning as a vaporous digital cloud. What I wanted to achieve with my photofilm was rematerialise that cloudy illusion by producing a group portrait of MTurkers’ locations. View from the Window presents a demographic snapshot of the global workforce, looking out. The human intelligence of Amazon’s invisible labour force mobilises digital technology to simulate the work of machines, but it also fractures the shiny image of the machine world by introducing the material traces of human bodies and their locations into the picture. However, using MTurk for this project was not ethically unproblematic and could actually be seen to be contributing to unfair labour conditions by validating Amazon’s platform. I was aware of that, and the exploration of these issues and conditions formed the very fabric of my project. My aim, arising out of an ethico-political injunction to understand the rhetorical and material force of ‘intelligence’ under conditions of global digital capitalism, was to offer a different vantage point for perceiving the relationship between humans and technology at this particular moment in time. More importantly, it was to recognise that there was a vantage point, and that the ‘view from nowhere’, to use Donna Haraway’s phrase, promoted by many AI companies ends up putting a very specific (white, male, ahistorical) human in the picture. So you could say that ethical concerns are the driving force of my artistic projects, they provide a rationale and a fabric, even if the work itself then chooses aesthetics as its primary mode of expression. But aesthetics is for me –as it was for a long line of European philosophers, from Immanuel Kant through to Jean-François Lyotard and Christine Battersby– inextricably connected with ethics.

Works Cited

Zylinska, J. (2020). AI Art: Machine Visions and Warped Dreams. Open Humanities Press.

Zylinska, J. (2014). Minimal Ethics for the Anthropocene. Open Humanities Press.

Zylinska, J. (2017). Nonhuman Photography. MIT Press.

Zylinska, J. (2023). The Percep9on Machine: Our Photographic Future Between the Eye and AI (open access; includes a link to the film, A GiF of the World [Oedipus on the JeKy]). MIT Press.