A real moment when AI synthesis needed a SME point of view

An executive emailed me a straightforward question: “How does our company rank in industry studies?”

I knew two things about this leader: he’s skeptical by default, and anything that smells selectively chosen gets discounted instantly — along with the subject matter expert who dared be so cavalier. This “simple” question was a rare opportunity for me to influence a leader at his level — and I had maybe 2 hours to respond.

How I answered would shape what he believes about our current position, how he reads every future study that crosses his desk, and where he points attention and investment.

A simple question, in other words, that wasn’t simple at all.

I started by letting AI do what it’s good at.

I gathered the well-known industry studies and uploaded them. Then I started asking — not just “How do we rank?” but:

• How does each study define performance?
• What dimensions are being measured?
• What does each say we’re strong at? Where do they say we lag?
• How do they weight perception versus operational reality?

AI was genuinely excellent here. It compressed days of reading into minutes, synthesized rankings and recurring themes, and surfaced something every analyst eventually meets: the studies didn’t agree. In one report, a capability showed up as a competitive strength. In another, the same capability was our largest gap.

And that’s where AI’s role largely ended — and mine began.

The hard part wasn’t synthesis.

I could have sent a summary. Here’s Study A, here’s Study B, here’s an average, here’s a slide with arrows. That would have been information. It would not have been helpful.

Because the executive wasn’t really asking, “What do the studies say?” He was asking, “What should I believe?”

And belief isn’t formed by synthesis.

So the next round of questions changed.

The work that followed wasn’t about editing or telling a story with data. It was about forming a point of view — and that meant questions AI could help me explore but couldn’t answer on its own.

Why do these studies disagree so sharply on the same capability? Which measures hold steady over time, and which swing with methodology? Which conclusions are safe to act on, and which would mislead if taken at face value?

Instead of trying to resolve the disagreement, I tried to understand it. The disagreement itself was the signal.

Some of the differences came down to how questions were framed, which populations were included, and whether expectations or lived experience carried more weight. Once I saw that, the contradictions stopped looking like noise and started looking like insight.

That’s not synthesis. That’s interpretation.

What I sent wasn’t an answer. It was a point of view.

A short, structured note: where the research clearly aligns, where it diverges and why, what we can say with confidence, what we should be careful not to over-interpret — and how I’d suggest reading future studies in this space.

I didn’t just answer the question he asked. I shaped how he’ll interpret similar information from now on.

What AI enabled — and what it didn’t replace.

AI made this possible in the time I had. It compressed the research, surfaced patterns and contradictions, and let me work multiple angles in parallel.

But it didn’t know which conclusions this leader would trust. How cautious or bold to be. What would influence a real decision versus create confusion. Where the real risk of misinterpretation lived. That required context, experience, and accountability — and accountability was the part that mattered to me. My name was on the reply.

Why this moment stays with me.

As executives increasingly use AI to gather and summarize information themselves, moments like this will get rarer. And more critical.

The value of subject-matter expertise isn’t disappearing — it’s shifting. From knowing more facts to helping others understand what the facts mean, deciding how much confidence to place in them, and shaping belief in ways that lead to better decisions.

AI is excellent at synthesis. But synthesis doesn’t create clarity. Clarity comes from a point of view — especially in the moments that shape how leaders think, decide, and invest.

That’s still very human work.