Bruce Lee Broke the Mirrors. Your AI Program Is Still Fighting in the Maze.

Your competitor knows your process better than you do.

Not because they are smarter. Because it is their process too. Same sixteen steps. Same containment targets. Same cost-per-contact scorecard. You have both been running the same maze for fifteen years, and now you are both adding AI to it and calling that a strategy.

There is a scene in the movie "Enter the Dragon" that explains why this never works.

Bruce Lee is losing a fight in a hall of mirrors. His opponent knows every corner of it, so wherever Lee moves, the other man is already there. Lee stops fighting the man and starts breaking the mirrors. Take away the maze and the advantage disappears.

I heard June Yee Felix reach for that same scene on a leadership podcast last week, and it has been sitting in my head since.

June and I are both part of the Florida Executive Forum, a networking group of Florida CEOs and operators. She spent over thirty years running technology and financial services businesses across the US, Europe and Asia, including senior roles at Verifone, IBM, Citibank and Chase Manhattan. She was Group CEO of IG Group, the FTSE 250 online trading firm, and sits on the boards of RELX, Iron Mountain and Hiscox today.

She saw that movie at sixteen. Then she went and did it.

She walked away from the thing everyone told her to defend.

When June took over IG, the company's entire reputation rested on one product. That product was about to be gutted by regulators. Shareholders were bruised. Regulators were skeptical. Most of the employees had never worked anywhere else.

The obvious move was to defend the product. Spend more on it. Fight harder for it.

She did the opposite.

Five years later revenue had doubled, profit had doubled, and the Japan business was five times bigger.

Asked which move mattered most, she picked charting a new course. Her reasoning was blunt. Without it, everyone would still have been pointing in the wrong direction.

More force is not a strategy.

June describes the leadership playbook most of us were trained on as more force, more resources, more control. Her verdict is that it works until it doesn't. In chaos, rigidity breaks.

That is the most accurate description of enterprise CX in 2026 I have heard from anyone outside our industry.

Adding AI to a process you have never questioned is more force. It is faster and cheaper and pointed in exactly the same direction as before. Your vendor will hand you a deflection number that proves it worked. That number will not tell you whether the process should have existed.

What breaking the mirrors looks like in a contact center.

I worked with a team redesigning a field services journey. Sixteen steps, start to finish. AI was designed into it from day one rather than added afterward.

Partway through, three of those sixteen steps turned out to exist only because, years earlier, a human being could not track enough information at once to skip them. They were workarounds for a limit that no longer applied. We removed them. Cycle time dropped twenty-three percent.

Nobody could have written that business case in advance. And nobody would have found it by asking how AI could make the existing sixteen steps run faster.

That is the difference between adding AI and charting a new course.

You cannot do it alone, and this is where most programs die.

June's second move at IG was bringing people with her. Long-tenured employees, first job for many of them, deeply invested in the product she was walking away from. So she pulled a hundred of her top people into a room and had them build the tactics for the new direction themselves.

Her description of what changed: they went from bystanders to advocates.

We are running our own version of this at Now Analytics right now.

We are building AI voice agents for personal injury intake. The people designing them are not only engineers. They are our own agents, the ones who have spent years taking those calls. They know which question makes a caller hang up. They know what someone sounds like right before they decide not to sign. They know the difference between a case worth pursuing and a caller who needs to be pointed somewhere else.

None of that is written down anywhere. It lives in the people doing the work.

So we put them in the design room. They are shaping how the voice agent handles hesitation, where it hands off to a human, and what it should never attempt on its own. They are picking up prompt design and conversation flow while they do it, which makes them more valuable whatever comes next.

Two things happen when you work this way. The product gets better, because the knowledge going into it is real. And the team stops seeing AI as the thing that ends their job, because they are the ones building it.

Now think about how your AI rollout reached the floor. An email. A vendor demo. A training deck. And a room full of agents quietly working out whether this is the thing that replaces them.

You will not get adoption from people who believe they are being replaced. You will get quiet resistance, and it shows up in your metrics as a tool with a low acceptance rate and a business case that never lands.

AI has won the IQ fight. The EQ fight is still open.

The line I keep coming back to is June's point about what AI cannot do. All the rigorous technical work she trained for, AI can now handle. What it cannot do is connect with people, communicate clearly, and reach them.

For a CX leader that is not a philosophical observation. It is a staffing model.

If AI takes your routine volume, everything left for your humans is harder. Higher emotional stakes. More ambiguity. Angrier customers who already tried the bot and failed. Your handle time goes up. Your hiring profile changes. Your training changes. Your pay bands change.

Almost nobody budgets for that, and then they act surprised when CSAT slides twelve months in.

The question.

Lee did not win that fight by getting stronger. He won it by taking away the thing that made his opponent dangerous.

Your competitors are not beating you on your process. They are beating you inside it. And they will keep beating you inside it, because they have been running the same maze as long as you have.

So here is the question worth putting in front of your leadership team this quarter.

If someone launched a competitor to you tomorrow with no contact center, no legacy platform, and no thirty-year-old process to protect, what would they build, and how long would it take them to start taking your customers?

Tell me on LinkedIn. I read every comment.