Electric Dreaming
Exploring the role of AI in contemporary artists' practice
By Simon Wilkinson – BRiGHTBLACK
In the Autumn of 2024, BRiGHTBLACK’s Electric Dreaming AI artist residency, funded by Arts Council England, brought together six artists to explore how artificial intelligence (AI) can address key challenges in artistic practice. The residency was run by myself and Myra Appannah as co-directors of BRiGHTBLACK, in close collaboration with academics from Cambridge University’s Jesus College, Leverhulme Centre for the Future of Intelligence (LCFI) and Liverpool John Moores University (LJMU).
Back in 2017, I myself was a resident artist at LCFI, digging into the creative possibilities of AI at a time when the tech was already outpacing expectations. The academics I worked with were starting to grasp a stark reality—artificial intelligence wasn’t just evolving; it was accelerating, tearing through its predicted timelines at an eye watering pace. We weren’t just studying AI—we were scrambling to keep up with it.
In 2016, AlphaGo changed the game—literally. Built by Google’s DeepMind, this AI wasn’t just another number-crunching machine; it was something different, something smarter. And when it went up against Lee Sedol, one of the greatest Go players in history, it didn’t just win—it redefined how the game could be played.
This was bigger than what happened in 1997 when Deep Blue narrowly defeated Garry Kasparov at chess. Back then, victory came down to sheer computational brute force—Deep Blue could analyse more moves, faster, than any human ever could. But AlphaGo’s win wasn’t about power; it was about intelligence. Using deep learning and neural networks, it played moves no human had ever seen before, breaking centuries of strategic convention and leaving Sedol visibly stunned.
The match was more than just a milestone for AI—it was a turning point. AlphaGo didn’t just outthink a grandmaster; it demonstrated a kind of creativity, an ability to see the game in a way that no human ever had. And that was the real shock. This wasn’t just a machine following instructions—it was a machine discovering new possibilities. AI had entered a new era, and the world was just beginning to understand what that meant.
Back in 2017, AI was still mostly a conversation about potential—exciting, sure, but always framed as something just beyond the horizon. Fast forward to 2024, and that horizon isn’t just here—it’s in the rear-view mirror. That’s why we launched the Electric Dreaming Residency. The game had changed.
To put it bluntly, if —like us— you wanted to generate an AI image in early 2021, you had to wrestle with VQGAN—an arcane, glitchy beast that lived in Google Colab Notebooks. You’d feed it a text prompt, wait an eternity, and be rewarded with something that looked like it had been beamed in from a malfunctioning alien brain trying (and failing) to grasp human culture. And yet, those imperfections were the magic. AI art wasn’t predictable; it was strange, warped, and often beautifully broken.
Then came 2022. Practically overnight, DALL·E and its ilk flipped the script. Suddenly, you could type in a prompt and get an image that actually looked like what you asked for—stunning, polished, and eerily precise. By the time the mainstream caught on, generative AI had exploded, sending shockwaves through the art world. Illustrators, designers, and visual artists weren’t just experimenting with it anymore—they were bracing for impact. The future hadn’t just arrived. It had bulldozed its way in.
For a company built on staying five years ahead of the tech adoption curve, AI has been the wildest ride yet for BRiGHTBLACK—faster, stranger, and infinitely less predictable.
Take VR, for example. When we first dived into it in 2010, progress was steady. Year by year, the tech got sharper, smoother, more immersive. By 2019, it had more or less plateaued. XR—virtual, augmented, mixed reality—settled into a slow refinement cycle, with no real leaps forward, just iterative polishing of hardware, software and strategies for using them.
AI, on the other hand, appears to be changing on a daily basis and certainly at a speed that no individual artist can hope to keep up with. This isn’t a medium you master—it’s a force you engage with. And that’s why AI - from our perspective - requires a collective response, a network of artists, thinkers, and technologists figuring it out together in real time. Because if you try to go it alone, you’re always going to be playing catch-up.
Hence, Electric Dreaming brought ourselves together with a diverse group of creatives, engaging with an impressive collection of AI researchers, academics, and tools to address four pressing challenges for the arts: meeting contemporary audience expectations, rebuilding arts communities, achieving net-zero sustainability and overcoming career barriers.
This article reflects on the findings of our residency, summarizing insights from the collective experiences detailed in our post-residency evaluations. These reflections aim to spark further conversations about how AI can shape the future of artistic practice.
Before we dive in, a huge shoutout to Dr. Aisha Sobey from Jesus College and Dr. Mark Wright from Liverpool John Moores University—absolute powerhouses who went all in during the residency. They cracked open the doors to a razor-sharp team of academics who didn’t just show up but properly got stuck in, sparking deep dives into social AI, AI intimacy, representation and justice in AI, the messy, beautiful dance between human and machine creativity, and the nuts and bolts of machine learning.
Meeting Contemporary Audience Expectations
With traditional arts venues watching their audiences drift away and video games tightening their grip on the cultural pulse, artists are hustling for fresh ways to pull people in. The game has changed—passive spectatorship is out, immersive, interactive experiences are in. So, have the Electric Dreaming artists turned to AI to craft audience-driven experiences that hit differently?
Brad Rumble developed AILA, an interactive installation.
“I utilised copilot, ChatGPT and Phind to help put the code together for developing my award winning piece titled “AILA” in which audiences were able to interact with an AI programmed into multiple roles based on Jungian archetypes. The AI helped with the code, with finding the correct model names for hardware such as the raspberry pi over arduino, selecting the correct gauge wires to use with certain power ratings etc..”Chipo Mapondera
“I have been using ChatGPT to assist in developing the tech pipeline and workflow for creating an immersive audiovisual experience using Raspberry Pi and human motion sensors to trigger content. I have also used ChatGPT to understand how to implement projection mapping, and what to consider to meet the technical requirements”Marcus Joseph
“I began using FYI as a conversational AI aide, which added a new layer to my creative process. This app enabled more interactive, real-time discussions around creative ideas, offering quick feedback that was ideal for brainstorming sessions. By developing ideas around live music and poetry experiences with FYI, I’m able to offer more than just traditional performances; I’m creating a shared, responsive experience that resonates with audiences on a deeper level.”Simon Wilkinson and Myra Appannah - BRiGHTBLACK
“One thing that caught our eye during the residency was Convai, a tool for creating AI responsive virtual characters who can interact with an audience member conversationally. It’s not perfect yet, but it's very impressive and the rapid pace of evolution means it’s certainly one to watch for us. Meanwhile we are using ChatGPT all the time now to help with C# coding, shader coding and making playmaker actions for unity games engine. We are using AI powered tools such as 3D scanners and Mixamo to rig 3D humanoid models. But more than anything we are using it to help us think, to help structure our work and to learn how to do new things that make our work better”.Kexin Liu explains that before using AI to make interactive work she frames its use around the following questions.
“How can I use AI to learn to create with interactive media? How can I use AI to overcome poorly thought-out functionality or design in an interactive piece to provide audiences with a seamless, satisfying experience? How can I use AI to involve my audience in the initiation, production, testing, marketing, and distribution of my work, making their participation meaningful even beyond experiencing the work itself?”
Willow Ritchie uses ChatGPT and other AI tools to simplify complex tasks such as 3D scanning in workshops.
“It’s amazing how much faster and easier the process of 3D scanning and character rigging becomes,” she shared. “it opens up possibilities for people to genuinely get involved in complex creative and technical processes and to see high quality results”.Hwa Young
“This is really hard for me, because I’m trying to get away from screens and tech-mediated interactions. I’m really not sure who ‘contemporary audiences’ are. Personally, meaningful interactions in a work context for me are in person, so I really can’t see how AI would help me”.
Rebuilding Arts Audiences
The last fifteen years have been a battlefield for artists—every aspect of their work has shifted. It’s not just about difficulty attracting funding or technology driven industry shake-ups; the entire cultural ecosystem has mutated. Fifteen years ago, most of the UK hadn’t heard of Netflix, the global video games industry was still comparable in size to cinema and social media was barely a whisper in the background. Fast forward, and audience habits have changed entirely—arts venues have been losing audiences and struggling to fund fees for touring shows, leaving artists to build their own communities from the ground up. So, have the Electric Dreaming artists found AI tools that can help them forge new connections, amplify their reach, reinvent what constitutes a venue or develop new ways of engaging with their communities?
Simon Wilkinson and Myra Appannah - BRiGHTBLACK shared how ChatGPT helped refine their outreach strategy.
“Venues booking and showing our work is a great way to build our community but that has become impossibly difficult over the past few years. In 2019 30% of our overall revenue was from international touring but first the pandemic, then brexit decimated that part of our practice almost completely. However, we recently used ChatGPT to help find new venues to approach. We explained a specific show in detail to ChatGPT and asked it to list all venues worldwide who pay fees and which would particularly suit that show. Whereas previously we would send hundreds and hundreds of emails and get a response rate of 2%, now we target a specific selection of venues and get a higher hit rate. ChatGPT has also helped us develop a social media strategy and a strategy for incorporating new platforms which are more responsive and less annoying than newsletters”.Marcus Joseph used the AI Roadie app as a virtual personal manager.
“AI Roadie, while practical and insightful, is still in its beta stage. When functioning properly, it allowed me to track performance stats, manage schedules, and engage with my audience in a more organized way. Having a virtual assistant for these tasks helped me streamline my workflow, letting me focus more on creative projects and interact with my community meaningfully “.Kexin Liu describes how she uses Large Language Models as part of a process when writing newsletters and other content.
“Usually, I can speed up the process by writing a first draft and then asking LLMs to rewrite it in the style of [insert your favourite blogger, link to their blog posts page, and highlight the best quality of their writing]. I would do this several times using the styles of different bloggers, then compare these re-writes side by side to study their tone and phrasing, which I can then use to edit my draft”
Digital Platforms and Metaverse Spaces
Some of the artists have also experimented with new AI enhanced ways of reaching audiences directly.
Simon Wilkinson and Myra Appannah - BRiGHTBLACK
“Metaverse spaces aren’t just alternatives—they offer new ways to engage with audiences who might never step into a traditional venue. AI helps us to learn new skills and with building these shared and embodied online spaces. Our first metaverse show 1000 Conversations About Death was built in Unity games engine and was effectively a theatre performance. These days when we have Unity open, we also have ChatGPT open, it has become an intrinsic part of the workflow. Where in the past we made work to be shown in venues, these days we have the option of building the venue itself. In the past four years only one real world venue has been able to offer us a fee for touring a show, all other offers involved us taking 100% of the risk. We have, though, had some success getting audiences to pay us directly for shows experienced online or in hybrid spaces. We see the use of AI in building on those successes as a key strategy for rebuilding touring incomes”Brad Rumble
“I feel myself pulled away from creating work that exists solely in digital spaces because I think this is the ordinary every day experience of most of the population, for me creating work that is out there and requires people to leave their home to experience, that is where the most impact can be had
Hwa Young
“ I have come to accept that being unable to reach more people is worth the sacrifice of not being ‘successful’ in the mainstream definition for an artist, as long as I can connect and have fewer but more deeper interactions with people”
AI and Sustainability
In an era of climate urgency, artists are reevaluating their practices to align with net-zero goals. Have any of the Electric Dreaming artists discovered AI tools which help them assess and improve their climate impact?
Simon Wilkinson and Myra Appannah - BRiGHTBLACK
“With regard to measuring potential carbon emissions or kilowatt hours resulting from different aspects of our work, what used to take hours of manual calculation now takes minutes with ChatGPT. This is useful both for funding applications and project evaluations but also for planning. Not only does AI give us the numbers but it will also suggest ways to improve the scale of our impact”Marcus Joseph
“Venues using AI-driven energy management systems like BrainBox AI or Grid Edge can optimize heating, cooling, and lighting, reducing energy waste during rehearsals and performances.”
Simon Wilkinson & Myra Appannah - BRiGHTBLACK
“Every project involves months of research. Back in the day this would involve buying books and doing google searches, but producing and transporting a paperback book equates to 2.6Kg of carbon and google searches are highly inefficient. These days we can get the information we need from a specific book in a few ChatGPT prompts. There are problems with this obviously, many authors have not received any extra reimbursement when their publishers sold the content of their books and articles to the AI companies. It’s in areas like this that we begin to see the danger of labour displacement playing out”.
Overcoming Bottlenecks and Barriers
Many artists face individual barriers, from limited funding to difficulty accessing networks, resources and space . How has AI helped the Electric Dreaming artists to overcome their own bottlenecks and barriers?
Kexin Liu emphasized how AI can help identifying opportunities and organising resources
“For example, if I need an overview of established conferences in media art, I would prompt LLMs to list major media art conferences worldwide. I would request the information in spreadsheet format, organised by the conferences’ credibility and prominence, from the most influential to the lesser-known.,” she said.Simon Wilkinson and Myra Appannah - BRiGHTBLACK
“With ChatGPT, the two of us can operate like a multi-agent company in terms of the administrative side of what we do. We are not core funded and we are not from the kind of wealthy background that means we don’t need to earn an income. We survive project to project, hustling on a daily basis to get by and all the admin comes down to the two of us on top of the creative work. AI is absolutely not the reason we’re able to produce enough work and run enough programs to be financially viable. There is a lot of thinking and strategy that goes into surviving as a small arts company, it takes a hacking mentality to not go under, and use of AI tools absolutely helps us with our thinking. In that sense I would say that AI, for us, has most value as an aid to thinking”.Hwa Young
“I like writing funding applications, and have partners who help with this aspect, so don’t like using AI tools to write. I think outsourcing this process streamlines the time for me to think through projects and get a deeper understanding of what I’m trying to do and why, so I don't want to miss out on the ‘thinking’ part of my artwork - which happens when I’m not in front of a computer”
Chipo Mapondera
“ChatGPT is enabling me to work with new technologies and upskill independently. This is helping me to work comfortably learning and doing complex work within briefs and deadlines”.
Willow Ritchie
“It occurred to us at Level-3 Festival that a lot of artist support organisations, especially those who work with learning disabled and neurodivergent artists, keep those artists in a client cycle. The artists find themselves constantly doing workshops and programs which fund posts in the host organisation, but never themselves get to become paid artists. We decided to change that model by running an AI Fluency program, helping neurodivergent artists to use tools such as ChatGPT to apply for and get their own funding. Several of our artists have achieved success in this way, with one person receiving £60,000 in funding this year”.
Marcus Joseph
“As a conversational AI tool, FYI serves as a valuable ‘creative assistant’ for brainstorming ideas and refining project proposals”
Broader Reflections on AI in the Arts
The residency sparked deep reflections not only on the role of AI in the arts, but also in more general terms. What did the Electric Dreaming Artists feel about the bigger picture?
Chipo Mapondera
“As an African person, it is incredibly important that we represent ourselves and create AI datasets that represent a true perspective of contemporary and historical African narratives. It is crucial that we empower our communities with the knowledge and resources required to do this in collaborative, multidisciplinary ways, so that we keep abreast with global tech developments.
In general the public are very ill-informed about AI, to the extent of them fearing it. This is why I am very passionate about the conversations that I am having through my podcast series, and promoting alternative perspectives about AI. I strongly believe in sharing this information with my community and empowering them to develop AI in new and creative ways.”.Myra Appannah - BRiGHTBLACK
“I am excited for the advent of AI agents with respect to my work - where I can not only use it to strategize and plan, but to deliver those strategies and plans - But what happens when everyone is doing this? In the current system, it will come down to 1) how fast can we adapt 2) how literate are we in new technologies 3) how confident we are to use them 4) how much are we able to be one step ahead of everyone else.
But what if there were a new culture? One where we use the time gained through the use of AI 1) to create more resources and education programs and share these for free 2) where we work in a system of collaboration not just competition 3) where funding bodies use AI gained efficiencies to offer us more funding and opportunities 4) where we use AI to co-design a decolonised arts system that works for all. So a key question at this moment would be: at this critical time - how do we bridge the gap between the massive individual power of say Sam Altman creating the system, and the many at the bottom trying to make it work for them”.
Kexin Liu
“AI, especially in mass surveillance and data processing, has significant implications for my current work. I am particularly drawn to understanding privacy as a collective issue shaped by AI technologies. For the general public, data is often perceived more individualistically. In regions like the United States and much of Europe, data is typically treated as a personal possession, while in mainland China, it’s often viewed as a public good—a different form of possession, yet possession nonetheless. Even under GDPR, data privacy is primarily framed around individual consent, focusing on personal risk. This approach tends to imply that data impacts only the individual, rather than recognizing data as something inherently relational. Through my work, I hope to draw attention to how the meaning and implications of privacy have fundamentally evolved in the age of AI”.Simon Wilkinson - BRiGHTBLACK
“AI has the potential to empower individuals in many ways, especially those with the flexibility to shift and change their circumstances fluidly. It has the potential to enable those that use it to level up their opportunities, it is an amazing access tool in the broadest of senses. I have never learnt this much this quickly before and I’m excited about the impact that possibility will have on people who find themselves at the bottom of the pile. On the other hand history teaches us that new technologies present new opportunities for powerful people to increase their advantage. As someone who trained as an economist, I worry about the way AI will displace human beings from so many roles in our society. I absolutely don’t buy the trope that other AI related jobs will appear to replace those that are lost. Powerful people have always wanted labour costs to be minimised and historically they have taken brutal actions to secure labour which is cheap or free. And whilst there are prominent and informed voices cautioning the world to take action on AI, there seems little appetite to allow those voices to be heard. So for me it’s a double edged sword. I can see the benefits, but history and academic researchers have shown us definitively that concentrating power into fewer hands has dire consequences for everyone else”.Marcus Joseph
“I believe that just as we adapted to the internet, AI will naturally integrate into our daily lives and become a powerful part of our creative and practical worlds. AI offers exciting potential, and I’m hopeful it will continue to evolve as a tool that democratizes creativity”.Hwa Young
“I think there will be grave mis-uses, and that under the current extractive capitalist system we are enmeshed in, it keeps tipping the wealth inequality to benefit the small number of gazillionaires. I don’t think the general public are informed enough, a lot of it through deliberate obfuscation by the major tech power players. Personally, a major component of my work is co-creation with underserved marginalised people, and through this how to help shift the dominant narrative of how society is structured to one that’s more caring, and responsible to one another, including our non-human companions on this earth”.Brad Rumble
“The potential for it to reduce the agonising tasks of admin are incredible, the long term impact of its use means that medical professionals will experience less burnout and will be able to spend more time interacting with their patients, in turn leading to a healthier society”.
Willow Ritchie
“The negative implications of Ai are worrying, specifically how it reduces the need for people, and the potential futures that phenomena could lead us down is something definitely worth exploring. But, I also think it’s important to explore the non-scary side of Ai, how it can be used by people to help themselves”.
Conclusion: A Blueprint for AI in the Arts
The Electric Dreaming residency opened up a critical conversation about use of AI in the arts—how it can reshape creative practice, pull audiences back in, drive sustainability, and help us push through systemic barriers. But it wasn’t just a tech-lovefest. It also made one thing crystal clear: AI isn’t just about possibility; it’s about responsibility. Ethical, societal, and environmental questions loom large. This isn’t just a shiny new tool—it’s a force that needs to be wielded with care.
For artists, the takeaway is sharp: approach AI with intent, and it can kick open doors to innovation, expand access, and supercharge creativity. But if the arts are going to thrive in the AI era, it’s going to take serious collaboration—between artists, venues, and institutions—to ensure technology isn’t just a gimmick, but a catalyst for real cultural transformation.
I recently checked in with the artists from the residency—turns out, all but one are now using AI daily and reaping the rewards. Personally, I’m seeing new tools pop up on a daily basis, becoming more usable, faster, and often free.
Now, we at BRiGHTBLACK are taking things a step further—translating everything we learned during the residency into Electric Dreaming Lab, a five-day deep dive designed to help artists integrate AI into their practice. The goal? Breaking down barriers, reaching bigger, more diverse audiences, building new arts communities, and pushing towards net-zero.
Want to find out more? Subscribe to BRiGHTBLACK’s mailing list at www.bright-black.org or follow us on Instagram @brightblack_productions.