Every week, a leader from an architecture, engineering, or construction firm asks me some version of the same question: with everything happening in AI, where do we even start?
I understand the anxiety behind that question. The noise around AI has never been louder. Every software vendor now sells an AI feature. Every conference has an AI track. Every board meeting has someone asking what the firm is doing about it. When everything is urgent, nothing is clear.
After two decades working at the intersection of AI and the built world, first as a researcher, then leading AI R&D at Autodesk, and now advising AEC firms through YegaTech, I have learned that clarity rarely comes from knowing more about the technology. It comes from asking better questions about your business. Here are the three I ask every executive team I work with.
1. Where does your firm actually lose time and margin?
Not where AI is impressive. Where you bleed. Proposal development that consumes senior staff for weeks. Quality reviews that catch problems after they become expensive. Knowledge that walks out the door every time someone retires. Project data scattered across systems that never talk to each other.
The best AI strategy starts with an honest inventory of friction, not a tour of tools. When you map your losses first, you evaluate every AI opportunity against a number, not a feeling. Some of the highest-return AI applications I have seen in this industry are unglamorous. Nobody keynotes about automating specification reviews. But the firms doing it quietly gained margin their competitors are still looking for.
2. What decisions would you make differently with better information?
AI is not primarily an automation technology. It is a decision technology. The firms that win with it are not the ones that automate the most tasks. They are the ones that make better calls: which projects to pursue, how to price risk, where to assign their best people, when a schedule is quietly slipping.
Ask your leadership team to name five decisions they make on judgment alone that they wish they could make on evidence. That list is an AI roadmap. It is usually more valuable than anything a vendor will show you.
3. Who in your organization is already experimenting?
In almost every firm I visit, someone is already using AI, usually quietly, sometimes against policy. A project engineer summarizing RFIs. A marketer drafting proposals. A designer testing generative tools on their own time.
These people are not a compliance problem. They are your early warning system and your first hires for any AI initiative. Find them, learn what they have discovered, and build guardrails that let them experiment safely instead of secretly. Culture, not technology, is what separates firms that transform from firms that talk about transforming. I wrote Disrupt It because I watched too many capable organizations lose not to better technology but to better cultures.
The filter
When the next AI headline lands in your inbox, run it through the three questions. Does it address friction we actually have? Does it improve a decision we actually make? Do we have people ready to test it? If the answer is no three times, close the tab and get back to work. If it is yes even once, you may be looking at something worth a pilot.
Hype is loud. Strategy is quiet. The leaders who thrive in the age of AI will be the ones who can tell the difference.