Where AI meets creative practice at the Royal College of Art
AI sits at the center of hype and backlash. The more useful conversation is happening in studios and crit rooms: what can AI actually do when artists and designers set the terms.
At the Royal College of Art, two creatives are proving that the best AI work starts with harder questions, not shinier tools. One mines a personal archive to explore memory. The other builds a public framework for tech that serves communities, not just metrics.
A film built on forgetting
Gregor Petrikovič is a Slovak-British artist and filmmaker who completed an MA in Photography at the RCA as a Burberry Design Scholar. His film, Sincerely, Victor Pike, was selected for New Contemporaries and won the Colección SOLO AI Award 2024.
The project began as function, not concept: an audio archive recorded since 2016 to manage chronic memory loss. Transcribing hundreds of hours with AI turned casual conversations into raw scripts-intimate, funny, profound moments that memory had blended together.
Gregor then leaned into early generative video for its glitches and dream logic-the way memory actually feels. He fed voiceovers into AI and ran into a wall: the system missed metaphors, poetry and nuance. That friction became the point.
His stance is clear: use AI without worshipping it. Keep the human voice, the imperfect recall, and the specific emotion in front. Connection might not be why these tools were built, but it can be the reason we use them.
- Start with what you already have. If you're sitting on voice notes and interviews, convert them fast with Speech-To-Text.
- Let constraint set your aesthetic. Embrace artifacts and "errors" if they serve the idea of memory and loss.
- Treat AI as assistant, not author. If the model can't read metaphor, keep that meaning-making firmly human.
Rewriting the narrative
Ramla Anshur studies part-time on the MDes Design Futures programme while working at Accenture as an experience designer. Her core question: what happens when communities that are usually subject to AI get to design and govern it?
That question came from practice. On a citizen-facing voice bot, her team hit blind spots-accents, speech impediments, accessibility needs. Without deliberate care, those gaps would harm the very people the tool was supposed to help.
She co-authored a 2025 paper that pushes back on "AI is inevitable." It lays out four moves any creative team can use:
- Resist: Name harms and challenge deployments that normalize them.
- Refuse: Set boundaries-opt out of low-consent data and misaligned uses.
- Reclaim: Redirect tools to serve community values, culture and needs.
- Reimagine: Prototype alternatives governed with those affected.
She's blunt about 2026: the "get on board or be left behind" pitch is loud, but audiences aren't buying it wholesale. Many artists are labeling work as "Human-made," and demand for human craft remains high. Her advice: organize, stay visible, and don't concede the future. Build futures that honor culture and planetary limits.
The RCA effect
Here's the twist: Gregor didn't touch AI at the RCA. He journaled, shot 16mm, and experimented with pre-cinema toys. Those hands-on, imperfect methods primed him to later see the beauty in generative glitches-and taught him to share messy drafts in crits and focus on ideas over polish.
For Ramla, the part-time route gave her depth. A standout module, Envisioning Futures, saw her team collaborate with the Design Museum's Futures Observatory on more-than-human scenarios, like a land council balancing the needs of all inhabitants. The outcome: a practice grounded in justice and regeneration, using tech only where it's truly needed.
Royal College of Art creates space for this kind of critical, practice-led work. For futures methods and public programs, see the Design Museum's Futures Observatory.
Action steps for creatives
- Audit your archives. Transcribe and cluster themes with Speech-To-Text tools.
- Pick a constraint and make it a rule. Memory gaps, dialects, access-let limits drive form.
- Prototype with imperfect models first; refine only after the concept lands. Explore workflows under Generative Video.
- Test for exclusion. Include varied accents, impediments and access needs before launch.
- Choose a stance from the 4 R's and put it in the brief so every decision ladders up.
- Share work-in-progress early. Treat crits and community feedback as part of the method.
About the creatives
Gregor Petrikovič studied MA Photography. He was a Burberry Design Scholar and is an alumnus of the International Studio & Curatorial Program (ISCP) in New York. His film Sincerely, Victor Pike won the Colección SOLO AI Award 2024, has been selected for New Contemporaries, and has exhibited at IDFA DocLab, Late at Tate Britain, and Sónar+D Barcelona. He is currently an FLAMIN Fellow at Film London.
Ramla Anshur is currently a part-time student in the MDes Design Futures programme. She works at Accenture as an Experience Designer and co-authored the 2025 research paper 'Resisting, Refusing, Reclaiming, Reimagining: Charting Challenges to Narratives of AI Inevitability' with We and AI. Ramla was selected for Catalyst's Kindling programme which supports tech justice projects and includes an in-person retreat.
Bottom line
AI is only as interesting as the question you bring to it. At the RCA, the questions come first.
Your membership also unlocks:
