Think critically
AI is changing the way we make, experience and understand creative work. It can open doors, make ambitious projects possible. But what does that mean for artists? Is AI simply another tool or is it changing the creative process itself?
In our recent webinar, Make it or fake it: Artists on AI, creativity and what’s at stake, artists and creative technologists Alan Warburton, Shehani Fernando, An-Ting and Ian Gallagher explored the opportunities, challenges and ethical questions emerging as AI becomes part of creative practice.
The conversation was wide-ranging, many acknowledging the uncertainty surrounding the technology, with headlines about job displacement and the absence of protections for artists’ voices and likenesses. Some artists are retreating from the technology wholeheartedly, but some are using it to experiment, question and provoke, making things that weren’t previously possible. Their work is asking us to think critically about the systems and technologies shaping our lives.
Using art to question AI
Artist, animator and video essayist Alan Warburton has worked with synthetic images for around 20 years. His short film Image Empire explores the relationship between computer graphics, computer vision and the technologies behind today’s synthetic media.
Describing himself as someone who has been working with “the tech that everyone loves to hate”, Alan deliberately launched the film on LinkedIn – a platform where people developing AI and those whose work is being disrupted by it sit alongside each other.
The film itself is entirely human-made, using visual storytelling to explore the technology rather than relying on it.
For Alan, artists have an important role to play in helping us understand technological change – not necessarily by providing answers, but by asking better questions.
“Journalists, artists and researchers with an interest in digital media have a responsibility to understand AI and there’s good and bad ways of experimenting with it.”

When AI becomes personal
For immersive producer and director Shehani Fernando, the question of what AI might mean for us personally was at the heart of Vocalize, created with Francesca Panetta and Halsey Burgund.
The immersive experience invited visitors at Hay Festival to train as voice actors with Stephen Fry. There was a twist: after completing a series of exercises, visitors heard their own cloned voices being used to endorse a series of deliberately dubious products.
The experience was playful, but also unsettling. As Shehani explained, the aim was to get people beyond an abstract debate about AI and allow them to experience what voice cloning might actually feel like.
Around 1,000 people took part, many leaving messages and wanting to discuss what they had experienced.
Shehani described the project as an attempt to use the arts to “move the needle in how we think about this” rather than reducing the conversation to a simple for-or-against debate.
AI as an enabler
An-Ting and Ian Gallagher, the creative duo behind Lost Communications, are exploring communication between humans and the natural world.
Their current project, Underwater, combines live music, immersive technology, an online game and underwater recordings to explore the sounds of marine life. Audiences can enter the virtual world as fish, while a live performance combines music with a submarine travelling through an underwater environment.
AI has been used throughout the development process – from researching spatial audio and designing equipment to generating 3D coral and helping develop code.
For Ian, the technology isn’t simply about doing existing work more quickly. It has made it possible to attempt things that previously felt out of reach.
“There are things that we made for this project that I just would never have tried to make in the first place,” he explained, because they would previously have taken months to develop.
For An-Ting, AI has also opened up technical knowledge. As a musician rather than a coder, she has used AI to help her understand programming and create tools to support her creative practice.

Opportunity – but who has the power?
The discussion also explored an important tension: AI can democratise access to knowledge and technology, but much of the infrastructure is controlled by large companies.
An-Ting and Ian encouraged artists to look beyond the biggest commercial AI platforms and explore open-source tools, research papers and communities. They described how freely available code and research can give small creative teams access to sophisticated technologies.
But greater access doesn’t remove the need to understand what sits behind the tools.
For artists, that means thinking about where models come from, what data they have been trained on, what happens to material uploaded to them and what choices are being made on the artist’s behalf.
Creativity, consent and responsibility
The speakers repeatedly returned to the importance of creative judgement.
AI can generate ideas, images, code and possibilities, but it doesn’t know what an artist is trying to say or why they want to say it.
An-Ting put it simply: “What is your thing which can touch me?”
That question points towards something central to the conversation. AI may change the labour involved in making art, but the human experience, intention and judgement behind creative work remain important.
The same applies to audiences. In Vocalize, the team thought carefully about what information they actually needed from participants. They decided against collecting unnecessary personal information simply to make the experience more shocking.
Transparency matters too. Audiences should understand what is being captured, how it will be used and what choices they have.
Stay curious – and stay critical
The closing message from the artists wasn’t to embrace AI unquestioningly or reject it outright. Instead, they encouraged artists to experiment, understand the systems they are using and develop their own ethical frameworks.
Shehani encouraged people to audit the digital tools they use, look at their terms and conditions and think carefully before uploading valuable creative work.
Alan urged artists, journalists and researchers to understand the technology rather than treating it as something happening somewhere else.
And Harmeet Chagger, The Space’s Head of Innovation, summed up the spirit of the session: “Stay curious and critical.”
For artists, that may be the most useful starting point of all. AI is changing quickly, and there are still plenty of unanswered questions about creativity, ownership, consent and power.
Artists don’t have to have all the answers. But they can experiment, question the technology and help all of us imagine what comes next.
Final words from our artists
An-Ting – AI is a learning tool
For me, AI is an incredible learning tool. It helps me find my way toward the vision I already have in my head. But without a clear goal, it can easily generate a lot of slop. Real creativity and a strong sense of direction are what make it genuinely useful. At the moment, I mainly use the DeepSeek API running as an agent on my own machine. It works well for what I need and is far cheaper than most comparable models. Things are changing fast though, and I’d encourage people to explore beyond the familiar big-brand names and get to know what’s actually available under the hood.
Ian Gallagher’s AI tips
My advice to get started would be to download ComfyUI from https://comfy.org/ and follow some interesting AI people on X:
-
https://x.com/PurzBeats – makes great tutorial videos on ComfyUI
-
https://x.com/peterom – Open Source AI Art advocate and creator of the https://www.banodoco.ai Art community
-
https://x.com/janusch_patas – CEO of https://lichtfeld.io – posts about machine vision and gaussian splatting
-
https://x.com/giffmana – Researcher at Meta – posts about LLMs and coding agents
There are many interesting and exciting papers coming out all of the time. These usually have a project page describing the new model or research and often have code to download or a https://huggingface.co/ demo to try out. For example, in Underwater we have been using https://lukashoel.github.io/video_to_world/ to generate 3D coral from videos
If you want to get started with AI agent coding, pay for the GPT Plus plan, download the Codex app, point it at your project or an empty folder and ask it what to do next!
Alan’s final thoughts….
There’s room for both refusal and various degrees of engagement with AI, but the main thing is to be thoughtful about it. Right now, most public judgement is based on the most visible uses — the low-quality, opportunistic stuff we call AI slop — while artists using the technology carefully and sensitively get lost in the noise, or worse, caught in the crossfire.
But that lack of nuance is exactly why it’s worth staying in the conversation. Those of us who’ve spent decades in the creative industries have a real stake here: our labour fed these models, and I’d argue we have more of a legal and moral claim to them than the people churning out slop. AI belongs to us, in other words — and artists on tiny incomes shouldn’t be expected to be paragons of ethical purity while the people who created this situation fly around in private jets.
My own way of maintaining integrity has been to make work about AI rather than with it: my film Image Empire is a kind of public information film, built on research and applied media theory, and it shows you can interrogate the technology productively without trying to make a quick buck. That said, saying anything useful does require hands-on experience — for me that’s meant working with ComfyUI, running open-weights models locally, which has a lighter environmental footprint and doesn’t scrape what you make.
My PhD looked at how professional creatives have essentially been an outsourced R&D department for software and hardware companies for decades — far from just clicking buttons, we’re the ones who standardise, optimise, bugfix and forge functionality from half-broken ecosystems. So we didn’t just build the AI models with our work; we built the digital tools, the game worlds, the creative suites that got sold back to us. AI isn’t so much the next frontier of industrialisation as the conclusion of that longer story — which means the questions it raises are really political ones, about how artists sustain themselves, and I think ideas like universal basic income belong squarely in that conversation.
Craft and creativity aren’t going anywhere, but arts organisations and artists may need to face the tensions in their work more directly now — and that could actually be clarifying.
As for tips: contrary to the full-on anti-AI crowd, I think journalists, artists and researchers interested in digital media have a moral responsibility to understand this stuff, and there are good ways to do it. Experiment with open-weights or local AI, try not to feed the bubble with subscriptions, and keep an ethnographic hat on — it’s the Wild West out there, and that makes it a fascinating thing to observe as well as a thing to be wary of.
How useful was this resource?