Reading the Room (Virtually)
The skill we didn’t realize was so physical
In a real conference room, you’re never just exchanging information. You’re constantly taking in tiny signals—someone’s shoulders tense, a glance drops, a smile flickers and disappears. You feel when a joke lands weird. You sense when a point inspires energy or quietly shuts people down.
Those micro-signals act like a built-in feedback loop. They help you adjust in real time: soften your tone, clarify a point, pause and ask a question, invite someone in.
On video calls, a lot of that vanishes.
Even with cameras on, we’re working with tiny squares, compressed audio, delayed reactions, and people multitasking just out of frame. We’re often talking into a kind of polite silence, trying to interpret a few nods and the occasional “mm-hmm.” It can feel like presenting into a void—unsure whether you’re connecting or just broadcasting.
Maya’s problem: the instincts go dull
Maya leads a distributed team across three time zones. She’s experienced, empathetic, and in-person she’s the kind of manager who can sense when someone is struggling before they say a word.
But remote work changed the texture of that. Her instincts feel duller now. Calls feel flatter. She notices something is “off” with one team member, but she can’t pinpoint what. No hallway chat. No small moments before or after meetings. No subtle cues.
So she does what many of us do in this new environment: she guesses. And guessing is exhausting.
Remote work isn’t just a logistical shift—it’s a relational one. When the feedback loop disappears, we can default to assumptions: They’re disengaged. They’re mad. They don’t care. Or the opposite: Everything’s fine, I’m overthinking. Either way, we’re making decisions with less data than we’re used to.
A new tool for an old question: “How will this land?”
Maya starts using AI in a surprisingly grounded way—not to “read” her team, but to help her anticipate how her communication might be received.
Before sending an important email about a major organizational change, she asks the AI something like:
- “How might this come across to someone feeling uncertain?”
- “What emotional tone does this strike?”
- “What might feel missing to a reader who’s anxious about change?”
The AI points out something simple but powerful: her email leads with the changes before acknowledging the feelings those changes might trigger.
That doesn’t make her email “wrong.” But it does shape how it lands.
So she rewrites. She opens by naming uncertainty. She signals care and context. Then she explains what’s changing and why.
Same facts. Different emotional architecture.
And often, that’s the difference between “This feels scary” and “I don’t love this, but I understand it.”
The remote communication trap: we write in isolation
Written and asynchronous communication has a hidden problem: you don’t get immediate feedback.
In person, if your message is confusing, you see confusion. If it hurts, you see withdrawal. You can repair in the moment.
But with email, Slack, and async updates, you compose alone and send into the unknown. Then you wait—sometimes for hours—for any sign of how it landed. If it lands badly, you only find out after it’s already created tension, anxiety, or a flurry of private side conversations.
That’s where AI can help—not as a mind reader, but as a proxy audience. A second set of eyes that forces you to ask, before you hit send: What are the likely interpretations of this?
Better one-on-ones: changing the opening changes the outcome
Maya begins using the same approach for conversations, too—especially the delicate ones.
She’s preparing for a one-on-one and wants to address a pattern: the team member seems less engaged lately. In her head, the honest opener is:
“I’ve noticed you seem less engaged.”
The AI flags what that line can imply: judgment, accusation, a conclusion already reached. Even if Maya doesn’t mean it that way, the wording can put someone into defense mode.
A different opener creates a different emotional doorway:
“I wanted to check in—how are things going?”
It doesn’t erase accountability. It doesn’t avoid the issue. It simply opens space first, which often makes the truth easier to say.
When Maya uses that approach, her team member actually opens up—sharing challenges Maya didn’t know about. Now the conversation can move from vague concern to real support and concrete adjustments.
The honest limitation: AI can’t read people
AI isn’t perceiving emotions. It’s pattern-matching, making educated guesses based on language and common human reactions.
So the value isn’t in the AI being “right.”
The value is in what it prompts you to consider: alternate perspectives, hidden assumptions, unintended tones, missing acknowledgments. It’s like a mirror held up to your message—one that asks, “Are you sure this says what you think it says?”
Used well, it doesn’t replace empathy. It strengthens it.
Adaptation, not a crutch
Maya notices the void starts to feel a little more like conversation. Not because AI made her communication perfect, but because it helped her restore something remote work took away: a feedback loop.
The human elements still matter most—genuine care, real curiosity, actual relationships. AI can’t do those for you. But it can help you communicate with more awareness when the room is pixelated, distributed, and asynchronous.
Reading the room was always a skill. The room has changed. Learning new ways to sense how your words might land isn’t weakness—it’s adaptation.
And in a world where so much is said without seeing the impact, a little extra reflection before sending can be an act of care.
