There is this strange thing I’ve noticed about the empty box.

The blank prompt that appears in almost every interface of a GenAI tool. The open field where you can type anything. It looks like a question-answer machine, even when what it’s really offering is closer to a studio. A place to form ideas and imagine how to bring things to life. This mismatch of interpretation matters.

Because a lot of people do not look at the empty box and feel possibility. They feel something closer to performance anxiety, like they are about to be graded on whether they ask the “right” thing. I see it everywhere with clients, coworkers, friends and even my own kids when they come to me for advice. It’s not strictly about fear of the future or lacking tech intelligence. It’s also about simply not having a starting point.

What do I want to create? Why do I want to create it? What’s even possible to create?

We underestimate how much we have trained people away from this kind of moment. Most of us were raised on machines that answer questions. Google taught us that if we type into a box, we will get a curated list back. You scan, you pick, you go. That pattern shaped a generation of work habits: ask, retrieve, move on.

Imagination tools have always had texture

There’s another reason an empty box is difficult. Imagination tools have always come with material. Paper resists pen. Clay pushes back. A whiteboard fills up. A canvas has edges. Even a journal has a binding that closes around what you have written.

The text field has none of that. No friction, nothing to push against. And we have been trained by every text field we ever met that typing is for utility. Search bars, address bars, login fields, email subject lines. Of course it doesn’t feel like a place to make something. We’ve not been invited to make something there before.

Generative AI is a different thing, and it requires different habits. It’s not really an answer machine. It’s a space where you can construct something. Where an idea can become a draft, and a decision can become a completely new way of working.

So when leaders say, “We need everyone to adopt AI,” I understand the logic behind the request. In their heads, adoption leads to innovation. Adoption leads to efficiency. Adoption leads to growth. Adoption leads to a better business.

Then you talk to their people and you hear:

“It’s helping me write better emails.”
“It’s helping me make slides faster.”
“It’s making me quicker at a few things.”

Important, yes. But it’s not what leaders are imagining when they require adoption.

That’s the imagination gap.

The imagination gap is the distance between using AI for “10% better” and “10x different”

SYPartners colleague Mark Newhouse has been writing about Radical Imagination as a leadership competency in the AI era. He describes it as:

“Detecting rules and conventions so entrenched they are largely invisible, exposing and challenging them, and filling the expanded spaces in their wake with previously inconceivable ideas.”

What I’m writing about is the people-side companion to that. The leader can imagine 10x, but if everyone else is still asking 10% questions, the org only moves 10%. Radical Imagination at the top doesn’t compound unless the people across the organization stop treating empty boxes as a search bar.

And right now the gap is filled with emotion. Fear. Intimidation. A kind of low-level anxiety you can hear in even casual conversation. Am I really behind? Am I using the right tools? Is this replacing me? What’s safe to use? What is true and real? This moment gets compared to past technology leaps. To me it feels like a much bigger one, and the advantage goes to the people who already spend their time imagining what does not exist yet.

What shifts when the box stops being scary?

Recently, I saw this gap close in a very clear way in a workshop we hosted. The room was full of people who already identified as creative: architects, interior designers and craftspeople. And they walked in holding their emotional stance on AI. Fears around security, being misled, quality and authenticity in their client relationships. They worried that taste and discernment would be outsourced and relationships they’d built with clients over the years would be flattened by the tool.

This fear kept them stuck on starting. What shifted them was a better way to relate to AI. We asked them to take any part of their work and sort it into three buckets:

• Where AI could augment
• Where it could automate
• And where it should never go

That simple triage was a powerful way to present the use of AI differently. AI stopped being a commanding force that took over. It became something they could place in relationship to their craft.

Then we gave them archetypes, because humans think in relationships: imagine AI as an assistant, a coach, a connector across your business content, etc... Once they had those frames, they stopped staring at the empty box and started talking about use cases. Not abstract ones. Real ones. Someone would say, “I want it to solve my procurement problems,” or “I want to respond to clients better and to write in my voice.”

That’s when they lit up. Because the tool stopped being a threat and became a partner in expanding their thinking.

The behaviors that show up when people aren’t afraid

Once people feel that shift, once they experience that the empty box can widen their imagination, not just speed up their output, you can start to talk about adoption in a different way.

Three behaviors start to really matter:

1. Curiosity: Not “What can this app do?” but the deeper “What can I build now that I couldn’t build before?”
2. Rigor: Not being seduced by the initial output because it sounds good and is convincing. Rigor asks, “Is there a better way of doing this? What’s missing? What needs to be checked?”
3. Taste: It’s discernment. It’s voice. It’s knowing what “good” looks like in an AI moment, the difference between something that’s merely fine and something that sounds like us. It’s the ability to look at 10 plausible options and choose the one that reflects the vision.

Curiosity gets you moving. Rigor keeps you from stopping too soon. Taste makes the result worth sharing.

These are the people-side expressions of what leaders are being asked to model: Radical Imagination, Proactive Discernment and the taste it takes to know which AI output is actually worth shipping. What Mark describes for leaders has to show up at every desk too, or it doesn’t really show up at all.

And those three behaviors don’t appear because someone issued a mandate. They show up when the environment makes it possible for them to practice. As one CEO said to me, “Everyone’s talking about efficiency. Fine. But where’s the cocaine?” In other words: What’s the real value—what’s the pull—beyond making the familiar faster?

The conditions that make the difference

In my experience, these conditions make the difference:

1. Remove the barriers that freeze imagination. People need clarity about what’s safe to try, what data the tool can access, who sees what they type and whether their experiments will be held against them. When that clarity is missing, people keep their use small and safe. Guardrails done well don’t restrict imagination. They create the spaces where imagination is allowed.
2. Give people focus, not overwhelm. Most people cannot keep up with everything that’s changing. They need a clear why, a sense of what matters and guidance for what’s in bounds. Otherwise, they default to the nearest familiar task, i.e. the better email.
3. Make learning visible. This is the people-level expression of what the paper calls Continuous Experimentation. In practice, it looks like normalizing failure notes, sharing prompt experiments openly and rewarding the person who tried something risky and came back with something useful. Without that, the most valuable learning stays trapped in one person’s chat history.

Closing the gap is shared work

Closing the imagination gap isn’t the work of any single role. Leaders have to imagine radically. People have to be given a place to imagine at all. Both have to move together.

When only the leader imagines, the vision arrives as a mandate and people learn to comply. When everyone else is still retrieving instead of inventing, the experiments stay small and the organization optimizes for a better email.

That’s the work. Not teaching everyone to prompt. Not demanding adoption. Creating the conditions where people can look at the empty box and finally know how to begin.

‍