Over the last decades, Hito Steyerl has been challenging the political boundaries between art and technology through her films, lectures and essays. One of the most celebrated experts in digital cultures, Artificial Intelligence and computer-generated images, in 2017, Steyerl was listed by ArtReview as the most influential person in the contemporary art world. Her works have been exhibited at numerous international biennials, including São Paulo, Istanbul, Shanghai, Taipei, Gwangju, Venice, documenta and Manifesta, and she has held solo exhibitions in prestigious venues such as the ICA in London (2014), MOCA in Los Angeles (2016), the Art Institute of Chicago (2021), Stedelijk Museum in Amsterdam (2022), and Fondazione Prada in Milan (2025). In 2021, Steyerl was awarded the Federal Cross of Merit by the German President Frank-Walter Steinmeier for her contributions to art, which she declined in protest at the German government’s handling of the Covid-19 crisis. She was recently appointed professor of Current Digital Media at the Academy of Fine Arts in Munich, and some of her most important books include The Wretched of the Screen (2012), about the contemporary circulation of images, as well as Duty Free Art: Art in the Age of Planetary Civil War (2017), about the role of art in the age of digital globalisation. In this conversation with Electra, Hito Steyerl addresses her new publication, Medium Hot: Images in the Age of Heat (2025), a collection of essays exploring how the use of large language models (LLMs) and the algorithmic creation of images is transforming our understanding of the world, raising serious questions about culture, environment, politics and history.
Hito Steyerl: What Can an Image Do?
In this conversation with Electra, renowned German filmmaker and writer Hito Steyerl addresses the complexities of the constant evolution between art and technology in the age of Artificial Intelligence, as well as the present convergence between the rise in computer-generated images and the emergence of authoritarian regimes worldwide.
Hito Steyerl, How Not to Be Seen: A Fucking Didactic Educational .MOV File, 2013. © Courtesy of the artist, Andrew Kreps Gallery, New York and Esther Schipper, Berlin, Paris, Seoul Stills © Hito Steyerl
AFONSO DIAS RAMOS You were always reluctant to describe yourself as an artist. Can you tell us how you went from documentary filmmaking to contemporary art spaces? And how did writing play into this?
HITO STEYERL I studied film both in Tokyo and in Munich with the idea of becoming a documentary filmmaker. However, the industry crashed at that moment, and so that never happened. But somehow, I found myself increasingly showing things in the art world. I ended up there by coincidence, or by mistake. When it came to writing, I always worked wherever I could earn money, including as a journalist. But I have also been working as a teacher for close to 30 years, so all of this came quite naturally. In fact, for a while there, I didn’t know whether I was going to become an academic or something else… This is how it turned out.
ADR Are all those activities complimentary?
HS Writing definitely involves a different set of crafts than the realm of video essays or art installations. But there is also film-making, architecture, a lot of technology, and all sorts of different things. Writing is its own separate craft. But I would say that it is now pretty much endangered, or at least called into question by Artificial Intelligence, or the so-called ‘Artificial Intelligence’. And I’m not really sure about how this will continue.
ADR You have reflected a lot on the notion of ‘artificial stupidity’. Is this more apt to describe what is going on?
HS Not anymore. Things have gotten much more complex in the meantime. LLMs are now very competent and can fully write on my behalf. I just have to deal with it and find new ways.
ADR In 2009, you put out an influential essay on the ‘poor image’, arguing for the subversive potential of low-quality images circulating online. But now you talk about the ‘power image’ instead. What does this shift mean?
HS I started writing the text about the ‘poor image’ at a time when YouTube had only been in existence for maybe two years. The idea that moving images specifically, could circulate around the internet was pretty new at that point. This is definitely no longer the case today. At that moment, a lot of works came into circulation which I hadn’t seen for many years because they were never shown anywhere in cinemas, and they were never shown on television. They just disappeared from the surface of the Earth and then reappeared on the internet. And then, of course, the circulation of images also accelerated and expanded a lot, and even their resolution became much higher in general. At that point in time, the resolution was PAL, featuring 576 pixels. But now the standard is HD or 4K, or even more than that. So, in that sense, the ‘poor image’ as I described it, no longer exists. The resolution of the image no longer really points to its status in the world of images, because now it could just be any resolution. Nowadays, the new parameter which needs to be taken into account is: how much power is the image backed by? And by this, I mean power in terms of energy consumption and render time, which is to say, the energy needed to store it and to retrieve it, but also other forms of power which make it visible in comparison to other images.
"Now we are rather talking about a network of images wherein each of them does its own thing in the world, communicating with one another."
ADR Harun Farocki once introduced the idea of ‘operational images’ to describe a new visual regime that was coming into being. Is it still useful today?
HS Absolutely. But I think that we have now even moved one step further, into the realm of agentic images. ‘Operational images’ are images that are programmable, images that have a function. They do something in the world. But now we are rather talking about a network of images wherein each of them does its own thing in the world, communicating with one another rather than with a viewer, or something else tangible. It also stops making sense to call them images, as they are not visual for human beings.
ADR How would your proposed term of an ‘inoperative’ image work in this context?
HS In 1986, the French philosopher Jean-Luc Nancy imagined that term in relation to a community. He said that many communities are defined by work, or somehow have to work together. But he was not convinced by that. He thought that any community that is defined on these terms of belonging always risks collapsing into reactionary modes, into trying to get rid of people that are not considered part of their community. So, he came up with the idea of the ‘inoperative community’ to talk about a community that does not work on many layers. I think that an ‘inoperative image’ or an ‘inoperative artwork’ could also be made in order not to work, so as to refuse the labour that it is supposed to undertake, not to produce any sort of profit, surplus or effect, and to just exist for its own sake instead.
ADR You also proposed the provocative idea of a ‘degenerative art’ in opposition to the generative images of AI. What would that look like?
HS I’m not really sure, you know? Obviously, I came to think about it because this term was invented at the art academy in Munich where I now work. In the 1930s, they came up with the term ‘degenerate art’ [entartete Kunst] to scapegoat any kind of art that they didn’t like, such as modern art. But then I thought about the fact that we currently have generative art and generative image-making which, in many cases, are also tied to the new, emerging authoritarian regimes around the world. So how could we basically undo that? Can we repurpose the idea of degenerative art in order to undo this right-wing generative effort that is happening now?
[...]
Read the full article in Issue 32 - Spring 2026
Share article