Innovation vs. Innovation Theater: How Leaders Tell the Difference

A hospital system I worked with recently invested $2.3 million in an AI diagnostic system. It was impressive technology. It had been validated at academic medical centers. The vendor had polished sales materials and case studies. The board approved it. The executive team announced it at their annual conference.

Eighteen months later, the system was being used by fewer than 5% of the physicians in the network. The executives had moved on to other priorities. The AI implementation was quietly archived.

When I asked the CTO what happened, he said something that I have heard, in different forms, a hundred times: "It turned out the thing we thought was broken was not really broken. But by the time we figured that out, we had already spent the money and built the political capital around the project."

That is innovation theater. It looks like innovation. It feels like innovation. It got presented as innovation to the board. But it created no improvement that mattered to the physicians, the patients, or the organization's core mission. What mattered was not whether the technology was cutting-edge. What mattered was whether it solved a problem that was actually constraining the organization from serving its mission better.

This is why innovation is the second pillar of the AI Leadership Triad, right after adaptability. And why understanding the difference between innovation and innovation theater is the most important leadership skill in the AI era.

Innovation Theater Is Everywhere in AI

Innovation theater is not new. But AI has created perfect conditions for it to flourish.

First, because the technology is genuinely new and impressive. The capabilities of large language models are remarkable. The ability to process and synthesize information at scale is a real breakthrough. That genuine impressiveness creates a kind of halo effect. If the technology is impressive, leaders assume, the application must be valuable.

Second, because adoption is treated as a competitive signal. Every industry is watching every other industry. Someone reads an article about how OpenAI or Google is using AI to transform a workflow, and suddenly the board is asking: are we doing that? A consultant comes in and shows what competitors are doing with generative AI, and suddenly there is a sense that if you are not building something, you are falling behind.

Third, because it is easier to count a technology investment than to count a problem solved. You can announce to your board: we deployed an AI system. That is a measurable thing. You can show a budget. You can show adoption numbers. What is much harder to announce is: we discovered that the thing we thought needed innovation did not. So we are going to focus on the three things that actually do.

The result is organizations across industries spending enormous amounts of money on AI projects that are technically sound but organizationally pointless. Joel Salinas sees this pattern in nonprofits, healthcare systems, financial services, and corporate training programs. The investment level changes, but the pattern is remarkably consistent.

Why Boring AI Consistently Outperforms Flashy AI

The most successful AI implementations I have seen are almost never the ones that started with "look at what this technology can do." They start with "our mission is being constrained by X. What is the best way to solve X?" Sometimes the answer is AI. Sometimes it is not.

A nonprofit that processes visa applications was drowning in data entry. Every application came in as a PDF. A human had to read it, extract the key information, and enter it into a database. It was slow, error-prone, and preventing them from helping more people. That constraint was real. It was measurable. It was directly limiting their mission.

They implemented an AI system to extract data from the PDFs automatically. It was not flashy. It was not impressive at a tech conference. But it reduced data entry time by 70%, eliminated a major source of errors, and freed up staff to do higher-value work. It did not look like innovation. It looked like someone solving a concrete problem with an appropriate tool.

That is exactly the approach Joel Salinas recommends. Start with the constraint, not the technology. Start with the problem that is preventing you from serving your mission more effectively. Then ask: what is the right tool for this problem? Sometimes it is AI. Sometimes it is a better process. Sometimes it is hiring someone. The technology is not the starting point. The mission is.

A financial services firm I worked with wanted to implement a generative AI chatbot for customer service. It seemed like a natural use case. Every fintech was doing it. The board wanted it. The technology was available.

Before building it, I asked: what problem would this solve? The answer was revealing. The problem was not that customers could not get answers to questions. The problem was that customers did not know what questions to ask. They would call support and say "I need help" without knowing what specific thing they needed. The back-and-forth of clarifying questions was where the real friction lived.

A chatbot would not solve that problem. A chatbot would add another layer of back-and-forth. What would solve it was better onboarding — actually teaching customers what the product could do when they signed up. No AI required. The firm saved millions by not building something that would have looked innovative but would not have addressed the actual constraint.

The Four Questions That Separate Innovation From Theater

Joel Salinas uses a specific test with every executive he coaches. Before approving any AI project, ask yourself these four questions. If you cannot answer all four clearly, you are probably looking at innovation theater.

1. What specific outcome is this supposed to improve?

Not "faster." Not "more efficient." Specific. Measurable. Connected to your mission. "Reduce the time between patient diagnosis and treatment plan by 30%" is specific. "Improve our digital capabilities" is theater. If you cannot name the specific outcome you are trying to improve, you do not yet understand the problem well enough to solve it with any tool, AI or otherwise.

2. Why is this the biggest constraint on that outcome right now?

You probably have ten things that would improve your core metrics. Why this one? What did you measure to determine that this is the bottleneck? If the answer is "the board thinks we should," you are looking at theater. If the answer is "we ran a bottleneck analysis and found that this step is preventing us from X," you have real innovation.

3. What will actually change when this is deployed?

Be specific about what people will do differently. Not "use an AI system." What actual behavior changes? Who does it differently? What stops being done? If the answer is "people will have more time to do higher-value work," the follow-up is: what will they actually do with that time? If you have not thought through the downstream effects of solving this constraint, you are solving a constraint in isolation, which often just creates a different constraint downstream.

4. Who is actually impacted by this not being solved?

If it is mostly the executives who want to seem like they are innovating, you are looking at theater. If it is the people at the front line of your mission — the teachers, the clinicians, the people serving customers, the team members doing the work — then you are looking at innovation. Real innovation solves problems for people who feel them daily. Theater solves problems that executives theorize about.

The Innovation Audit as a Leadership Practice

Every time Joel Salinas works with a new executive, one of the first things they do is an innovation audit. You look at every AI investment currently in flight or approved. You ask those four questions about each one. You measure the gap between what the project team thinks it is solving and what the actual operational problem is.

Almost always, there is a gap. Sometimes a small one. Sometimes a massive one. Sometimes you discover a project that started as real innovation but has drifted into theater as it has progressed.

Here is what I recommend to executives: do this audit yourself, honestly. List every AI project in flight. For each one, write down: the specific outcome it is meant to improve, why this is the biggest constraint, what will actually change, and who is impacted. Show it to the people closest to the work. Ask them: does this match your reality, or are we solving the problem we want to solve rather than the problem that actually exists?

That one exercise often reveals more about your organization's relationship with AI than months of strategy discussion. It shows you where innovation is real and where it is theater. It shows you where your leadership team is aligned and where you are solving different problems.

Why This Is the Second Pillar

Adaptability without innovation produces organizations that change without purpose. Adaptability and innovation together produce organizations that evolve their methods in service of a clear mission.

The AI Leadership Triad works as a system. Adaptability keeps your mission clear while you evolve. Innovation keeps that evolution focused on things that actually matter. Creativity — the third pillar — generates the new possibilities that innovation can pursue.

But of the three, innovation is the one that separates organizations that are genuinely using AI to amplify their mission from organizations that are using AI to convince themselves and their stakeholders that they are modern and forward-thinking.

The question worth asking yourself right now is simple: of the AI projects your organization has launched in the last year, how many of them actually solved the problem you thought you were solving? If you are not sure, that is worth investigating. The executives who are building sustainable AI cultures are the ones who have the discipline to say no to impressive technology in service of no problem, and yes to boring technology in service of a real mission constraint.