Why industrial AI needs its own org chart
IT is built for support. Operations is built for survival. Neither is structured to make AI work in the field — which is why industrial AI needs its own accountable operating model.
These are the companion notes. The full essay lives on The Turing Pilgrim, my Substack publication.
Why this matters
The essay opens in a West Texas dispatch meeting, where a mathematically sound crew-routing model died in one sentence from a field supervisor: his people should own their day, not be micromanaged by an algorithm. The math was not wrong. The model stripped agency from the people doing the physical work, and the people deploying it were structurally insulated from that reality.
That is the pattern. Enterprise IT is built for scale, compliance, and support — measured on uptime and tickets, not field outcomes. Operations is built to protect the current way of working, because in physical systems caution is survival. Hand AI to either and you get a standard software rollout or decision-support theater. Everyone agrees the pilot was promising. Nothing changes.
The argument: carve out a separate, field-embedded AI unit — pushed by a ruthless executive mandate, pulled by giving A-players a high-accountability sandbox — that treats the model as a living system and feeds every override and ignored recommendation back into the product.
Key takeaways
- Adoption dies on agency, not accuracy. The pilot failed because it micromanaged the crew, not because the routing math was wrong.
- IT is the wrong owner. A team measured on uptime and ticket resolution cannot be accountable for field outcomes.
- Operations is the wrong owner too. The reliability instinct that keeps people alive will quietly limit AI to slicker reporting.
- The push and the pull. An executive mandate protects the unit from legacy metrics; the sandbox attracts the best builders.
- Field-embedded by design. The biggest risk of a separate unit is becoming another silo — smart people in a Houston office building for people they never meet.
Who should read it
Executives and product leaders at legacy industrial companies deciding where AI should live on the org chart — especially anyone about to hand it to IT or fold it into operations.
Related field notes
When AI Learns Physics…
Most industrial AI can notice patterns but cannot explain the physical story underneath them. Physics-informed neural networks change what the field can ask of a model.
June 11, 2026The Small Business AI Opportunity Is Hiding in Plain Sight
Small businesses are not short on ideas. They are short on capacity. AI works when it acts as a support layer around the owner — not as a replacement for their judgment.
May 14, 2026The Next AI Shift Is Delegation
The hard question is not what AI can do. It’s what we should let it do, and how.
Working through this decision?
A short note on the AI bet you are weighing is enough to start. I usually reply within one business day.