People sometimes introduce me on stage as the holder of eight U.S. patents in AI, and I can see the audience picture something cinematic: a flash of insight, a sprint to the patent office, applause. The truth is less glamorous and far more useful.
Every one of those patents, spanning AI, manufacturing, and building design, came out of the same unremarkable process: a real problem that annoyed someone enough to demand a better answer, a long series of ideas that did not work, and a team patient enough to keep asking why. If you lead innovation in architecture, engineering, construction, or manufacturing, the lessons from that process matter more than the certificates.
Constraints are the raw material
None of my patents began with the question, what could AI do here? They began with constraints that would not move: a design that had to meet code, a structure that had to stand, a process that could not add a single hour to a schedule already promised to a client. The built world is a world of constraints, and that turns out to be an advantage. Constraints tell you exactly where the valuable problems are. When people complain that a limitation makes innovation impossible, I have learned to hear it as a signal that we are close to something worth patenting.
The invention is rarely the hard part
Here is what surprised me most across those eight patents: the technical insight was usually the fastest step. The slow steps were everything around it. Convincing people the problem was worth solving. Getting access to the data that proved it. Translating between the language of engineers and the language of executives. Innovation in established industries is mostly a communication and trust problem wearing a technology costume. This is why I tell leadership teams that hiring brilliant technologists is necessary but nowhere near sufficient. If your organization cannot carry an idea from the person who has it to the person who can fund it without losing its meaning, no amount of talent will save you.
Better questions beat better answers
One of the ideas I am proudest of was later used in the design of the world's first AI-designed car, a story Fast Company covered. What made it work was not a clever algorithm. It was reframing the question, from how do we automate what designers do, to how do we let designers explore options no human would have time to consider. The first question replaces people and produces resistance. The second one augments people and produces champions. That single reframing, from replace to augment, changed my career, and it eventually became the title of my first book. Organizations that ask how AI can replace their people get sabotage. Organizations that ask how AI can amplify their people get transformation.
Innovation is a habit, not an event
Eight patents sounds like eight events. It was actually one habit sustained for years: noticing friction, asking why three more times than is polite, and writing down ideas before judging them. Firms are no different. The ones that innovate consistently are not the ones with innovation departments. They are the ones where questioning the default is a daily, unremarkable act, safe for a first-year engineer and expected of a senior partner.
You do not need a patent portfolio to compete in the age of AI. You need the habits that produce one. Start with the friction in front of you, protect the people curious enough to poke at it, and keep asking better questions. The certificates are optional. The habit is not.