Three months in. Three things I have tried, been surprised by, and am still figuring out about building an AI-augmented leadership practice.
I started Q1 thinking the breakthrough would come from finding the right AI tool. The perfect prompt engineering. The ideal workflow automation. Three months later, the breakthrough was none of those things.
The breakthrough was changing how I think about delegation. Not to people, but to systems. The mental model shift from "AI as a search engine" to "AI as a thinking partner" changed everything about how I prepare for keynotes, how I synthesize research for clients, and how I draft proposals.
The tool mattered less than the mental model. If you are waiting for the "right" AI tool before you start, you are optimizing the wrong variable.
The fear in every AEC conference hallway: "Will AI replace what I know?" After three months of deep daily use, I can say with confidence that the answer is the opposite of what most people expect.
AI is spectacularly good at generating plausible-sounding content. It is spectacularly bad at knowing when that content is wrong in ways that matter. The more domain expertise you have, the better you are at catching the errors that would cost a client real money or real safety risk.
This is the Material Intelligence argument in practice. Twenty years of looking at concrete failures means I can spot when an AI-generated assessment sounds right but misses the mechanism. That judgment is not replaceable. It is, in fact, the thing that makes AI useful instead of dangerous.
(AI + MI) × HI™ is an original framework developed by Jigar B. Desai.
I speak to firm leaders every week. The gap I see is not technological. It is conversational. Most leadership teams have not had an honest discussion about what AI means for their business model, their talent strategy, or their client relationships.
They have had the "should we buy Copilot" conversation. They have not had the "what does this mean for how we price our expertise" conversation. The first conversation costs $30 per seat per month. The second conversation determines whether the firm is relevant in five years.
If you lead an AEC firm and you have not had the second conversation, that is where the urgency is. Not in the tool. In the strategy.
I use AI to draft, synthesize, and iterate. But there is a line somewhere between using AI as leverage and using it as a crutch. I have not found it yet. Some weeks I catch myself accepting an AI draft that is 85% right and calling it done, when the last 15% is where my actual point of view lives.
The ½ this quarter: I am still calibrating how much I should let AI do the work versus how much the work is what makes the thinking sharp. The answer is probably different for every task. I am watching for the pattern.
Get the next edition delivered. No spam.