Most companies I talk to want AI to solve a complexity problem.
AI does not reduce complexity. It reflects it. Every broken process you feed into an AI system comes out the other side faster, cheaper, and more broken. The mess does not disappear, it amplifies.
I call this organizational gravity. The idea is simple. AI accelerates the direction an organization is already moving. If you are moving toward simplicity, AI compounds that. If you are moving toward complexity, AI compounds that too.
The direction was already set before the technology arrived.
The room where it happens
Last year I was at an executive offsite, setting priorities for 2026. Small room. Twenty leaders, and someone asked the group a question: how do we reduce cost using AI?
The answers came fast. Automate sales reporting. Deflect customer service calls. Cut content production time.
I listened for a while. Then I asked a different question.
What does the customer actually need from these processes? Not what we do in them. What they need out of them.
The room went quiet for a moment.
That question is where transformation starts. Not with the technology. Not with the process map. With an honest answer about what work is worth doing at all.
The form nobody questioned
Here is a concrete example. One organization I worked with had a standard order intake process. A form arrived. Nothing moved forward until every field was complete. The team had built an automated reminder workflow to chase the missing fields.
They wanted to make the chasing faster.
What we did instead was ask why each field existed. Who used it. What decision it informed. When we finished that exercise, more than half the required fields turned out to be unnecessary. They had been copied from an older version of the process and nobody had questioned them in years.
The 30 percent reduction in customer wait time did not come from automation. It came from elimination.
That is the sequence. Eliminate first. Then simplify what remains. Then accelerate. Not the other way around.
Why AI makes this more urgent
For years, complexity was expensive but manageable. Organizations added layers slowly, the friction was annoying but not fatal. We could hide it and afford the OpEx to add labor for the manual processes.
AI removes that friction. It makes the broken process run at speed. It removes the natural drag that used to slow things down enough for humans to notice something was wrong.
That is the risk most executives are not talking about. Not that AI will replace people. That AI will entrench the organizational decisions that should have been revisited long ago.
The organizations winning with AI right now are not the ones with the most sophisticated models. They are the ones that did the hard, unglamorous work of simplification first. They eliminated the processes nobody should have been doing. They tightened the ones that remained. Then they put AI on top of something worth accelerating.
What this means in practice
Before any significant AI investment, the right question is not: what can we automate?
It is: what are we trying to achieve? What are our customers willing to pay for?
That question is harder. It requires the executive team to agree on priorities, which means agreeing on what does not make the list. It requires killing projects that are already in flight. It requires someone in the room willing to ask how a process benefits a customer before agreeing to make it faster.
That is not a technology decision. It is a leadership decision.
AI will reflect the organization you have built. If you want different results, build a different organization first.
The technology will take care of the rest.