Partway through a meeting, a stakeholder said, “That sounds fine.”
The words said yes. The tone said something else entirely.
I slowed down. Asked a different question. Gave him room to find what he was actually trying to say. Ten minutes later, we had a real conversation about a concern he’d been sitting on for weeks, one that would have quietly shaped every decision going forward if it hadn’t surfaced.
No AI in that loop would have caught it. Not because the tools aren’t powerful, but because of a structural limitation that doesn’t get talked about enough.
AI Works With What It’s Given
The quality of AI output depends entirely on the quality of the input. That’s not a flaw. It’s just how it works.
But it means that when the person giving the input is uncertain, vague, or holding something back, the AI works with that gap. It can’t know what wasn’t said. It can’t notice that the question someone asked isn’t actually the question they meant to ask. It processes the surface. It doesn’t reach for what’s underneath.
The human in the room does that. Not by prompting better, but by listening differently. By noticing when something feels off. By asking the question that wasn’t on the agenda but needed to be.
That’s not a workaround for AI’s limitations. It’s a distinct capability that matters more when AI is doing more of the explicit work.
What Reading the Room Actually Looks Like
There are specific moments I’m thinking of.
When someone goes quiet at exactly the wrong time, and I notice, and I don’t move past it.
When a team member says something technically accurate but with an edge that tells me there’s a conversation we haven’t had yet.
When a client starts to answer my question and then trails off, and instead of filling the silence, I let it sit, because something’s about to come out that needs space.
When I’m presenting an idea, and I can feel the room pulling slightly away before anyone says a word, and I adjust — not the slides, the approach.
What shows up in a meeting summary or a transcript is what happened in real time, with other people, and it changes what becomes possible.
Presence as a Professional Skill
My framing for AI has stayed consistent: I want it to handle what’s explicit so I can be more present for what isn’t.
The most valuable work I do happens in the space between what someone says and what they mean. Helping someone articulate a concern they hadn’t fully formed yet. Noticing when the conversation needs to slow down. Being the person in the room who makes others feel heard.
That’s not a soft skill in the dismissive sense of the phrase. It’s a competency. It takes years to build. It doesn’t transfer via documentation. As the mechanical parts of work get handled, this is what’s left. This is the work.
The people who develop this, not as a supplement to technical skill but as a skill in its own right, are going to be the ones who matter in any room they walk into. AI in the loop or not.
Leave a Reply