An orthopedic surgeon once described to me how he operates on shoulders using microsurgical tools. The problem isn’t the precision of the instruments or the complexity of the procedure. The problem is where he has to look.

To operate, he must watch a monitor above the patient rather than looking directly at the shoulder he’s working on. His hands move in one space while his eyes track another. For someone who spent years training to work by touch and direct sight, this is profoundly counterintuitive. Every instinct fights against it.

The problem exists in plain sight. Surgeons experience it daily. Yet it remains unsolved not because the technology doesn’t exist—augmented reality could project the surgical field directly onto the patient—but because the people who understand the problem have never collided with the people who could solve it.

The software engineer who could apply AR doesn’t know this problem exists. The surgeon doesn’t know AR could solve it. Put them in a conference room for a “cross-functional innovation workshop,” and you’ll get polite presentations. The surgeon outlines challenges in medical terminology. The engineer showcases AR capabilities in technical specifications. They nod appreciatively, exchange business cards, and leave without collision.

Innovation doesn’t happen in that room.

The Physics of Ideas

Innovation requires collision—not collaboration, not coordination, not the careful orchestration of complementary skills. When fundamentally different mental models meet under the right conditions, energy is released. Ideas that were impossible in isolation suddenly become visible.

Think of electrons colliding in a defined space. You need three things: enough energy to overcome resistance, the right particles encountering each other, and a container that focuses rather than diffuses the impact.

Vinod Khosla didn’t just observe the productive tension between software people and motor people in early electric vehicle development; he banked on it. He recognized that friction between paradigms—algorithms vs. torque, abstractions vs. thermodynamics—wasn’t a liability. It was the engine. That tension produced breakthroughs neither group could have reached alone.

But here’s what we’ve done instead: we’ve professionalized innovation facilitation right out of existence. Innovation labs with whiteboards and post-its. Hackathons with staged progression. Cross-functional workshops with tidy agendas and learning objectives. We’ve made innovation safe, comfortable, and presentable.

We’ve designed collision out of the process.

Movement as Mechanism

Collision requires movement—not metaphorical movement, but actual physical movement. Static presentations produce static thinking. When people sit in fixed positions, power dynamics calcify. The person at the head of the table leads. The person with the projector controls attention. The junior person stays quiet. The expert dominates.

This is where play becomes essential—not as an icebreaker or team-building activity, but as the mechanism that enables collision. Play creates movement. It breaks hierarchies. It grants permission for half-formed ideas. It supports rapid iteration without the weight of formal proposal processes.

When a group is moving—rearranging furniture, building prototypes from whatever materials are at hand, acting out scenarios—the surgeon and engineer stop presenting at each other and start solving together. The surgeon can physically demonstrate the counterintuitive hand-eye coordination problem. The engineer sees it not as a specification but as a human experience to be transformed. They don’t leave with slides—they leave with a crude prototype made of tape and cardboard that actually works.

Movement isn’t incidental to innovation. It’s the condition that makes collision possible.

The Containment Field

Collision sounds violent—impact, friction, resistance. But a nuclear reaction without a containment vessel isn’t a power source; it’s a bomb. And this reveals the paradox of the reactor: to sustain the intellectual force of colliding ideas, you need a containment field of absolute social safety.

This is where improvisation enters—not as entertainment, but as technology for building that containment field. Improvisational theater runs on the rule of “yes, and.” It doesn’t mean agreeing with the idea; innovation requires ruthless challenge. It means agreeing on the reality. Accepting what your partner offers and building on it rather than blocking it.

“Yes, and” is the safety rule that prevents intellectual friction from becoming personal conflict. Improvising together teaches people to separate ego from output. Trust develops not through personality assessments or trust falls, but through the high-wire act of creating something together in real time without a script. The group forms its identity through the collision process—not before it.

They become a reactor, not just people in a room.

Designing the Reactor

I once designed an innovation environment intended to bring together people from radically different domains. Colleagues reviewed the plans—the movable furniture, the reconfigurable space, the absence of fixed presentation areas. Then they noticed one specification that concerned them: one wall would be covered in street graffiti.

Not curated art. Not commissioned murals. Actual graffiti—unauthorized expression, the visual language of claiming space against official order.

They thought it unprofessional. That was precisely the point.

Institutional spaces—government offices, corporate headquarters, universities—quietly enforce institutional behavior: measured speech, careful positioning, deference to hierarchy. The graffiti wall breaks that contract instantly. It’s a visual permission slip. It signals: This space operates under different rules. Unauthorized expression is not only permitted; it’s expected.

The reactor isn’t a metaphor. It’s a design principle. Every element either enables collision or prevents it.

Movable furniture means power structures can shift. No one owns the front of the room. The group reorganizes space as problems evolve, not according to the stages of a prescribed process. And that movement also supports the thinkers who need silence; the room can be configured for deep solitary processing as easily as for loud collaboration. It respects the need to withdraw and reflect after impact.

Endless reconfiguration means the environment adapts to emergent needs. When the surgeon needs to demonstrate the counterintuitive surgical problem, furniture moves. When the engineer needs space to mock up a heads-up display, the room transforms. We don’t need facilitators managing agendas; we need architects who set the constraints and let the reaction run.

Contrast this with typical innovation spaces: conference rooms with fixed tables, labs with permanent whiteboards and branded materials, workshop venues with predetermined “stations.” These spaces prescribe behavior. They may be decorated differently, but they are fundamentally static. They enable presentation, not collision.

What We’ve Designed Out

Most innovation programs fail not because they lack smart people or good intentions, but because they’ve been designed to prevent the very collisions they claim to enable.

We’ve optimized for comfort when we need productive discomfort.
We’ve structured encounters when we need emergence.
We’ve built spaces for collaboration when we need containers for collision.

The surgeon’s problem—operating while looking away from the surgical field—remains invisible not because surgeons and engineers never meet, but because they’ve never been placed in conditions where they could truly collide. We’ve given them presentation slots, not reactors. We’ve facilitated networking, not movement. We’ve encouraged collaboration, not the productive tension of fundamentally different mental models meeting under conditions that force transformation.

The reactor isn’t about adding elements—beanbags, whiteboards, innovation consultants. It’s about removing barriers to movement, trust, and collision. It’s about designing spaces where institutional rules visibly don’t apply, where furniture moves as freely as ideas, where the safety of “yes, and” enables the danger of radical thinking.

Innovation happens when problem-understanders collide with solution-holders. But collision doesn’t happen in conference rooms.

It happens in reactors—spaces designed not for comfort, but for transformation.

The graffiti stays on the wall.