I design environments where human capability emerges through play, not instruction.

AI is being built for two games. Efficiency — work solved in the background, agents in competitive headless dark forest underwood markets, tasks as commodities. And entertainment — software as content, discovered, forked, memed. Viral applications with attention as primary currency and short shelf life.

Both valid. Neither produces progress alone.

Radioactivity. Relativity. Washing hands to prevent infection. All discovered by capable minds being playful with ideas and therefore ending up changing trajectories. They manifested wormholes through informed intentional serious play.

We are currently experiencing a three body problem. Play as search. Play as entertainment. Play as competition. But entertainment and competition captured almost everything. The search mode is loosing its gravity. The forces are out of balance.

But every game in existence is a result of free play. Games emerge from play. Without play, no new games. In our drive to win or be entertained, we forgot how.

Learning, discovery and by extension progress, innovation and change are outcomes of free play. Autonomous agentic dark markets and software as viral content are already here. What's missing is a way to collaborate with AI natively that's focused on search, not research. Emergence, not convergence.

AI tools today collapse everything to an input field. State your intent. AI does the rest. Do everything for me.

But clear intent rarely exists at the starting line. Like taste. Like intuition. It's an emerging property, not starting condition.

My belief is that design, learning, and innovation is a wave function collapse. Ideas, concepts, prototypes — all in superposition until the process resolves them. How you play with constraints early on enables different forms of collapse.

Good design is logical. Great design jumps through a wormhole though.

Most current tools are designed for an immediate collapse. Not enough time and resources are spent yet on designing for the superposition — the space where capability, concept, and ideas actually develops.

Early in my career I got my childhood dream job and designed physical LEGO sets. What stuck: LEGO's genius is that the measurement system is a protocol. Not a constraint — a protocol. Guarded religiously. A brick from 50 years ago fits today. A 5-year-old and 36-year-old me use the same bricks. Both make something real. Intentional sticky protocols don't limit possibility. They make it infinite.

Play-testing taught me the most. Kids follow instructions not to finish but to learn enough to stop following them. The moment they go off-script is the moment the opportunity space starts to expand. That's not onboarding. That's capability design.

Working at LEGO meant access to any brick in existence. Surrounded by the best builders in the world. Learning by access and proximity. It rewires how you approach your own process. Ideas out early. Iterate in parallel. Follow loose thoughts. Share often. Don't discard too soon. Make the real thing. This is also why Midjourney sparked so well through Discord — prompts and other people's outputs acted as building instructions. The browsing was the learning. A pattern in AI we see rarely.

Composability, modularity, interoperability. Not concepts I read. Plastic in my hands. The body (intuition) knows before the mind. Physical systems with immediate feedback loops — you build, you see, you adjust. That's how capability grows.

I didn't have language for any of this until later when I started working in decentralized, distributed and interoperable systems and realized I'd already learned the principles — in Billund, on a table covered in bricks.

I spent five years at IDEO across San Francisco, Munich, and Shanghai. With most of my work under strict NDA — the kind that sparks internal change but never gets acknowledged publicly.

What stuck: you can't understand the future from a deck. You have to map the systems — the networks, markets, incentives, dependencies — and then build something tangible inside it. Prototypes aren't just proof of concept. They're how you communicate what a strategy deck never can. Narratives and vision made tangible.

The range was the education: five-year flagship vision for a top-three gaming company. Virtual theme parks. Industrial IoT. In-car AI. Interactive streaming. Mozilla's innovation approach to the surveillance economy. What I learned — map the system, learn from similarity in analogous or opposite spaces, find the node with the most leverage, prototype there, watch how the system responds.

Overall the work in and around the gaming industry turned out to be the best school for interaction design — not clicks and flows, but engagement loops. Magical circles. Not gamification, but play. Research as design. Design as research.

Zoom in. Zoom out. Find the node. Prototype. Watch the system light up. IDEO had a saying — never go to a meeting without a prototype. With AI the time to prototype collapsed. Now: never leave an idea unbuilt and experienced.

I co-founded WE3 as a design collective and venture fund. Flat. No titles. Collectively owned. Self-organizing. Bottom up. Half compensation as equity in the portfolio we co-create as collective.

What stuck: most industries are the brink of hyperfinancialization and turn more and more towards dark forests. The most adversarial design environment that exists. Every mechanism designed gets gamed. Every beneficial incentive exploited. You ship a product and watch actors / agents / bots stress-test every assumption within hours. Game theory not as concept — as daily practice.

This is where and when LEGO clicked.

Smart contracts are protocols. Agents are multiplayer by default. The internet is more like a quantum possibility space. Composable systems creating network effects through interoperability, not control. This is where I started seeing the future of AI — headless agents operating in observable, verifiable markets. The infrastructure already exists. Just as usual not equally distributed.

Together with WE3 designers I worked end to end. Zero-to-one product design for project enabling capital and social formation. Systems design and exploration around novel reputation systems. Created Lens' decentralized social graph innovation lab, to explore the future of open social experiences. And had the opportunity for six month to research the Ethereum ecosystem with Optimism, Espresso Systems, and the Ethereum Foundation and highlight its dynamics. And through my role as a design partner at IDEO Colab ventures get a first seat where innovation and capital are converging.

What I'm taking from five years so far at the edge: we obsess over systems designed to be gamed, not played. That creates a game for efficiency, but kills innovation and discovery. And in parallel it produces specialist roles that become less relevant every month as AI outperforms them. The game is eating itself. So I can’t stop wondering what if all there is left is play.

Right now the game is eating itself. So I stopped gaming. To focus on play. Searching, learning, embracing ambiguity, discovering through emergence.

A self-funded practice. Not resting — prototyping. 50+ experiments since November. Building new communities. Expanding my networks. Learning how AI transforms what I'm capable of by living it every day.

Two years ago I was copy-pasting ChatGPT into Replit because I struggled often with things obvious to any regular dev. Now — Rust. Go. Python. Typescript. Even though I am not proficient in any of those languages at all. I have a thing for physics-based (multiplicative play) apps for emergent thinking. I was always good at mapping architecture, translating insights and ideas into stories and narratives. But I had to outsource the building. That has changed forever. Not gradually. All at once.

One question drives me now: how do we make genuinely new things happen? Discover the unknown unknowns. Build intuition instead of repeating what's already in the training data. Not faster. Not more efficiently. Navigate ambiguity and wormhole into new territory.

I stand that play is wrongly labeled as something for kids. We need to relearn it collectively — with our new alien technology. The collapse and execution are increasingly solved. What's not solved: how we play with ideas before the collapse. Environments where you branch, observe multiple realities, change the physics of the reality you play in. Where intent emerges instead of being demanded. You don't find superpowers by outsourcing work. You find them by discovering.