Reflections on AI Image Generators

Jess MacCormack: Reflections on AI Image Generators (writing)

“Political and military conflict as well as racially motivated migration barriers are important tools in creating this disenfranchised labour force. Perhaps bias is not a bug, but an important feature of a mean system of production. Bias is not only productive on the level of representation, by visually demeaning people. Its supposed elimination is equally productive in helping to consolidate class hierarchies buttressed by wars, energy conflicts and racist border systems, and can be taken advantage of within a mean system of production.” – Hito Steyerl, Mean Images

Millions of people grow up wishing they could be more artistic, better at drawing, capable of creative output but do not have the time to pursue such dreams due to economics, class, normative expectations, etc. Then suddenly a tool becomes accessible that allows them some creative potential they never had, and social media offers a place to share their inventions. Or maybe some artists and directors have been making art or film for many years and are intrigued by the potential of these new tools. What I have seen as a trend in this new wave of “AI artists” is that the excitement of these new tools has led to addictive behaviours that reflect the underlying conditions of mental illness and the isolation and alienation born by capitalism. The desire for connection, recognition and virality within social media has led many people to performative extremes, including many documented deaths while individuals push the boundaries for sensational content.

AI image and video generators provide immediate satisfaction to our compulsivity, which underlies the lack of community and resources in our tech-oligarch mediated societies. Being siloed off into echo chambers online has in some ways reduced our capacity to exist with others, difference and conflict. The dissociation required to participate in living our lives online has opened up our psyches to ongoing manipulation. We must be prettier, smarter, more likeable, more political and more capable of navigating hate to appease the algorithms. We must spend all our time online to be seen, and of course we never truly are. The panopticon has successfully trapped so many of us in a catch-22, our lives must look beautiful or exceptional AND be displayed for others to judge. This climate sometimes engenders social movements and empathy, and other times breeds competition, insecurities and alienation. Our lives online are now also part of our CVs, we need to exist there to find employment or buyers, to promote our businesses and upcoming events – in many ways we need to be online to survive contemporary life. This is currently seen as Palestinians livestream a genocide on Instagram in hopes of getting international help while mainstream media outlets look the other way. Every day we see children dying.

Life online can be both extremely empowering and simultaneously disempowering. Relationships are often reduced to superficialities, misunderstandings, financial exchange and social capital, but there are also oppressed communities who finally get to represent themselves and see and meet like-minded folks, to organize and validate, to meet up and… along comes AI, disrupting everything we understand about the real.

I think many people have found themselves lost in the gaps of society, which is amplified by social media and contemporary politics – and AI feels like the perfect outlet (and distraction) to participate in processing our inner worlds and offering it to others. Using AI tools can make us feel part of the whole again somehow. The recycled residue and wealth of information of the internet is available for direct dialogue. We can push our hands through the digital matter, and it takes shape like other art materials, but with some sort of instant mastery, like a well-trained yet alien collaborator. The magic of text prompts being translated into unique images that quickly pop up before us, proposes an uncanny reflection of our ideas and world. For many, the experimentation with text prompts formed a new practice. How can we accentuate the flaws of AI images as a feature?

Film and photography have been used as a reflection of our lived realities for over a century. Also weaponized as a tool of colonization, social control and propaganda, these mediums have expanded and contracted our world and understanding of what it means to be human. AI generators are not just accessible tools, they are also tools that innately disfigure representation and capture the essence of a contemporary digital surrealism. They amplify our global industrial problems, theft and exploitation, while also offering a balm of creative possibilities. It feels like those of us struggling with mental health problems, and that is many of us, are finding solace in the wild representations of our broken humanity.

AI image generators feel like we are playing a game as addictive as gambling. The gift of instant imagery of whatever our hearts desire or of something beyond our imaginations is gratifying. Unfortunately, there is no reality check to the immense consequences of AI use. Many people rallied against AI upon its slow arrival and then sudden omnipresence. The initial outcries felt reactionary, “AI will steal our jobs!” and somewhat misplaced, “Authorship is the only way there is value in art!”. There has always been fear of new forms of media and societal change, but now it is happening at an incomprehensible rate. Ethical questions were an afterthought: What if AI images just reinforce oppressive biases? How will we sustainably power AI?

The US government utilized the destabilization of the masses to usher in fascism and crown each tech oligarch as a king. Power has been centralized, and the people have been left behind. Pick a side politics rein, reducing every issue into a binary, so we lose all nuance of the complexity at play in humanity. Does the utilisation of image generators for art further propel this dystopian hellscape while also offering some relief from it? Since Trump got into office he has posted various AI-generated propaganda and deepfakes; a video of himself, Elon Musk and others taking over the Gaza strip to turn it into “Trump Gaza” vacation properties or Donald Trump as the new pope. The creator of the Gaza video intended it as satire, but then Trump posted it on Truth Social, illustrating how meaning making no longer resides in the medium nor the message. The material is both offensive and banal but echoes the internet’s earlier viral AI-generated images of Pope Francis in a swaggy puff coat or All Eyes on Rafa, which was shared 40 million times on Instagram stories. As Hito Steyerl wrote in Mean Images:

“They converge around the average, the median; hallucinated mediocrity. They represent the norm by signalling the mean. They replace likenesses with likelinesses. They may be ‘poor images’ in terms of resolution, but in style and substance they are: mean images.”

The U.S. government is now using AI to monitor online activity to identify Palestine sympathizers and revoke their visa applications.

As an artist on social media…

How could AI be used as a progressive new tool for digital arts? AI image and video generators are complicit in theft and environmental damage. Is there any possibility for the creation of ethical AI art? Corporations plowed ahead in AI’s development without considering any of the political or environmental consequences. Data and artworks were, and continue to be, used as data sets without the artist’s consent or renumeration. Companies continue to compete to create newer and “better” models without finding environmentally friendly options. The underlying motivations for production continue to be defined by neoliberal capitalism, greed at any cost. But whether we opt in or opt out AI generators are still here, and people are using them.

For the past 20 years my art practice has engaged with social media and its relationship to digital art. I’ve been particularly interested in how the digital world impacts our relationships, identity and politics. For the past three years this has included inhabiting and participating within a new subculture of “AI artists” on social media sites X and Instagram. Originally, I was drawn to AI image generators out of fascination for the new technology but soon found myself finding the process addictive, and then the support for my work enticed me further down this rabbit hole. At this time, I was completely unaware of how large language models worked or what the bigger global consequences might be. Many people were originally awed by the wildly surreal images AI models could create. As time passed, I noticed how many people had become ‘artists’ due to the accessibility of AI and communities began to form around this. Some Canadian artists I knew began using AI image generators and posting their work online in 2022, such as interdisciplinary artist Beth Frey. I watched as her Instagram account blew up and suddenly, she had 100,000 followers and so many artistic and financial opportunities opening around her. Later my online friend Wenhui, known as Nice Aunties, account also went super viral, and she transformed her former career as a successful architect in Singapore into an AI artist who sells NFT’s for $20,000 USD an image, and has articles written about her work in the Guardian. I felt torn between the desire to be seen, recognized and having opportunities, and the lack of critical thought in these communities. I kept making AI works in hopes that “something would happen” while the art world I knew and loved began to shun me along with AI use in art.

What ended up happening was October 7th. Having been an activist for the past 20 years on issues ranging from criminalization and abolition to trans rights and the deconstruction of homonationalism, Palestine has always been part of the political landscape I engage with. The amplification of the ongoing occupation and genocide in Palestine stopped me in my tracks. Making weird art with AI for the sake of experimentation and exploring my own mental illness or queerness through short videos no longer felt purposeful. (At this point I was still ignorant of the environmental damage this AI practice causes). My AI videos became more overtly political as I tried to grapple with how I could help raise awareness of the Palestinian cause with my accounts and following. Trying to initiate public discourse with AI artists was near impossible, but my work hit a nerve with other communities.

For a year I felt more connected to other activists, artists and thinkers online while simultaneously being horrified by the ongoing atrocities Israel was inflicting on Gaza. Meanwhile the community of “AI artists” around me chased bigger and better generators, such as Sora, which were capable of more cinematic realism and quality renderings. I felt my work was a part of something bigger than myself, but as time moved on, I felt this activist work I was making utilising unethical tools was problematic. As a politically outspoken person within these AI art communities, I was harassed and shunned. It would be great to see critical debate within communities of AI users but so far, I have found very few that are interested in AI’s global impact and I have spoken to many who believe it will solve these problems for us. To make an ethical, environmentally friendly model will pose the questions: Do we really care about the sustainability of the planet? What does AI have to offer contemporary art and culture? And why weren’t these questions answered before proceeding with AI’s development?

– Jess MacCormack (April 2025)