The art of unlearning to get away from AI
Insight - In Her Own Words
If a machine can make something beautiful without feeling anything, does it matter? I asked myself that for years while I worked like a machine: efficient, productive, technically strong. On paper, it looked right. In practice, I felt empty enough to quit design for a while.
What brought me back wasn't a new tool or framework. It was listening to how creativity felt again. I stopped chasing correctness and started paying attention to what I could offer as a human. Creativity without emotion isn't creativity; it's output.
I'm a creative director, educator, and artist from Italy, now based in London. My work questions how we make things, then breaks and rebuilds those assumptions. The goal isn't just better work, but a more human industry-one where soul shows up on the page.
Unlearning beats optimization
As a kid, I took things apart to see what was inside. That was raw curiosity. Somewhere along the way, rules, deadlines, and "best practice" replaced it. I learned how I was supposed to create, not how I naturally did.
Unlearning is a return. It strips away the filters and the fear. It prioritizes aliveness over approval. It asks: what do I feel, and what happens if I follow that?
AI isn't the enemy-but it isn't a compass
In Milan, Berlin, and London, I became a competent, replaceable designer. That's why AI feels familiar: it's great at learning patterns and producing results. Useful, but emotionally absent.
I see AI as a supercharger. It can get you somewhere fast. But you still have to choose the destination. It doesn't dream, wonder, or care. It remixes what exists; it doesn't originate what doesn't.
When AI makes mistakes, we call them hallucinations. Funny word, because that's close to how human creativity works. You stumble, something breaks, and something new appears. Art loves ambiguity. Code needs clarity. That messy middle-instinct before understanding-still belongs to us.
The womb, not the machine
Pregnancy made this visceral. It's the most literal form of creativity: intuitive, vulnerable, transforming. It made me ask whether creativity needs a womb-protection, time, and softness-more than a machine.
What difference does being alive make?
Even if we could simulate the universe perfectly, would we know why it exists? AI is brilliant at the how. The why-meaning, empathy, intention-stays with us.
When I unlearned the rules I'd absorbed, my voice came through. My work stopped being correct and started being felt. It became harder to copy. So did I.
How to unlearn (without blowing up your process)
- Feel-first briefs: Before opening any tool, write three sentences on what you want others to feel. Design to that, not to trends.
- No-undo sessions: One hour with analog tools or a single layer. Constraints force instinct. Mess invites ideas you can't plan.
- Hallucination drills: Prompt an AI to make five "wrong" options. Don't fix them-steal the accident and push it by hand.
- One-rule unlearn: Pick a rule you cling to (grid, palette, hierarchy) and break it for one project. Notice what wakes up.
- Emotion checks: Midway through, ask: what do I feel? If the answer is "nothing," remove one safety net and keep going.
- Client honesty: Share the messy version first. Set expectations that surprise is part of the process, not a detour.
Work with AI, keep the soul
- Use AI for breadth, keep depth human: Let it draft, collect references, or test variations. You choose meaning and finish by hand.
- Protect your signature: Maintain a living library of your textures, marks, phrases, and color biases. Feed the machine, but don't let it define you.
- Slow the final 10%: The last polish decides if it's alive or generic. Spend time here. That's where your fingerprint shows.
A simple challenge
Unlearn one thing this week. The next time you reach for the tool that makes life easier, pause. Try the messy, human way first. See what shows up.
I began by asking if it matters that a machine can create without feeling. Let me flip it: if you create something, will it be different because you felt something while making it? That's the part the machine still can't learn.
Photo credits: Lorenza Ragno
If you want structured ways to work with AI while staying human-led, explore these practical AI courses by job.
Your membership also unlocks:
