Innovation in the Age of AI: Why the Next Big Idea Looks Like Failure First

This is the ninth article in a series on the human capabilities that AI cannot replicate.

Thomas Edison didn’t describe the thousand unsuccessful experiments that preceded the lightbulb as failures. He described them as steps. A thousand-step process, he said — not a thousand defeats.

Most organizations have built their culture around the opposite assumption.

They want results. They want ROI. They want to know, before the investment has even been made, whether the idea is going to pay off. The pressure is understandable. The quarterly report doesn’t wait. Thin margins can be unforgiving.  The board doesn’t invest in patience. And so, systematically, without anyone deciding to do it, organizations build cultures hostile to the very thing they say they want most.

This is the innovation cost no one puts on the ledger. Not the absence of creativity. The impatience with it.

Why Organizations Kill Their Best Ideas

Teresa Amabile at Harvard Business School spent years studying creativity in the workplace. She analyzed more than 9,000 daily diary entries from people working on projects that required creative output. Her conclusion is one of the most important findings in organizational research: When creativity is under the gun, it usually ends up getting killed. Time pressure doesn’t sharpen innovation. It narrows it. Severe deadlines push people toward the safe, the familiar and the immediately defensible — and away from the genuinely new.

Clayton Christensen identified the same dynamic from a different angle. In The Innovator’s Dilemma, he showed why well-run, profitable companies consistently fail to capitalize on disruptive innovations: They are too good at the short term. Driven by quarterly results, they pour resources into the products and markets that already exist, rationally concluding that the emerging innovation isn’t large enough or certain enough to justify the investment. By the time it is large enough and certain enough, someone else has built it.

The cost of this pattern is invisible — until it isn't. When an organization kills an idea early, no one sees what was lost. The ledger records it as fiscal discipline. But it doesn’t record the revenue stream that never materialized or the competitor who built it from the blueprint you shelved.

Eastman Kodak is the cautionary case. In 1975, a Kodak engineer named Steven Sasson built the first digital camera. Management shelved it to protect the company’s lucrative film business. By the time digital photography became unavoidable, competitors had built the market Kodak had invented. The company that created the future didn’t get to live in it.

The Adhesive That Didn't Work

In 1968, a chemist at 3M named Spencer Silver invented an adhesive that nobody knew what to do with.

It was, by every conventional measure, a failure. The adhesive bonded lightly to surfaces rather than permanently. Silver had done everything right. The science was sound, the manufacturing process was clean, the product was consistent. The problem was that it didn’t answer the question everyone was asking: How do we make a stronger adhesive?

For six years, the adhesive sat in 3M’s research archive as an experiment with no application, a solution without a problem. In 1974, a colleague named Art Fry had a small frustration: The paper bookmark in his hymnal kept falling out. He remembered Silver’s adhesive. A bookmark that would stick lightly to a page and come away cleanly — that was the question the adhesive had been waiting for.

3M brought the product to market in 1977. Post-it Notes became one of the company’s most enduring and profitable product lines. They are still on desks around the world today.

What changed between 1968 and 1974 was not the adhesive. What changed was the question and the context. Someone finally asked how a low-tack adhesive could actually be used. In that context, the innovation became visible.

This is the reframe. And it is the engine of every genuine innovation.

The Patience Dividend

The model I use with leaders who want to build genuinely innovative cultures is the Patience Dividend — the return on organizational investment that comes from creating conditions where ideas are given time to find their questions.

The Patience Dividend is recognition that the distance between discovery and application is almost always longer than the quarterly calendar allows. And organizations that refuse to span that distance will consistently hand their best ideas to competitors willing to do so.

Two things are required to collect the Patience Dividend. The first is a reframe of what failure means. Edison did not say the experiments didn’t matter. He said they mattered in a specific way — as evidence, as elimination, as progress toward a defined destination. Organizations that cultivate this framing ask what failed experiments taught them and what they eliminated. Iteration is a step (or multiple steps) in the process of innovation.

The second is a structural commitment to protect the time and space required for that learning to accumulate. 3M’s culture, in Spencer Silver's era, gave researchers time outside their project mandates to explore without immediate commercial application. That protected space is where Post-it Notes came from. 

Three Practices That Cultivate the Patience Dividend

1. Reframing failure as iteration

The single most important thing a leader can do for innovation is change the organizational vocabulary around what an unsuccessful experiment means.

In most organizations, failure ends conversations, closes budgets and hurts the credibility of the person who proposed the idea. Future risk-taking becomes less likely. In innovative organizations, failure is one step in a multi-step process. It narrows the problem, even when it doesn’t solve it.

The shift requires leaders to model the language explicitly. For example: “We learned that approach doesn't solve the problem — which means we now know more about what does.” It requires rituals that celebrate learning alongside results. And it requires the discipline, in the presence of an unsuccessful experiment, to ask the Edison question: What step is this in the process, and what does it tell us about the next one?

2. Protecting the incubation window

The Patience Dividend requires protection of the space between discovery and application, the window in which an idea is real enough to pursue but not yet proven enough to justify on a balance sheet. Most organizations collapse this window under pressure. The result is what Amabile's research documents: the absence of conditions in which creative thinking can take root. Embers need oxygen and fuel to become a fire.

Protecting the incubation window means building organizational structures that give ideas room: dedicated time outside core project mandates, budget that isn’t tied to immediate ROI, and leadership language that explicitly grants permission to explore without a predetermined outcome. 

3. Questioning the question before solving the problem

The Post-it Notes story is the model. Before a team moves to solutions, leaders who build innovative cultures ask: Are we solving the right problem? What assumptions are embedded in how we’ve framed this? If we reframe the problem entirely, what becomes possible?

In our Culture Catalyst program, this is among the most disruptive practices we introduce. It interrupts the momentum toward speed and demands a moment of genuine inquiry before the race to the answer begins.

Why Innovation Is a Human Capability AI Cannot Replicate

AI is extraordinarily good at solving defined problems. Given a clear question, it can generate, evaluate and optimize solutions with a speed and breadth no human team can match.

What it cannot do is question the question.

AI, given Silver's adhesive and its properties, might generate a hundred applications. But it can’t make a spontaneous connection between a lived frustration and a years-old discovery. It cannot feel the mild irritation of a bookmark falling out of a hymnal and make the creative leap that connects a years-old archive to an unmet daily need.

The reframe from “this failed” to “what is this actually for?” requires human cognition. It requires the capacity to hold a discovery and a question simultaneously, to let them sit in relationship without forcing a resolution, and to recognize the connection when it finally appears. It requires patience and perception.

AI can accelerate the steps of a thousand-step process. It cannot decide that the process has been asking the wrong question from the start. That is the work of human leaders. And in an age when answers are cheaper and faster than ever, the leaders who know how to ask better questions will be the ones building the next thing worth building.

In my next article, I’ll explore adaptability — the human capacity to evolve continuously when the environment demands it, and why resilience in the face of change is a cultural skill.

This article is part of a series on the human capabilities that AI cannot replicate. Previous installments have covered connection, trust, accountability, curiosity, discernment, integration and delegation.

If your culture has genuinely creative people but keeps killing what they bring, I'd like to help you think through solutions. Book a 30-minute insight call with me and let’s talk about what’s happening and what to do about it.