The smartest person in the room, and the hardest to follow
There is a particular kind of silence that falls over a meeting when the expert starts talking. Not the respectful kind. The other kind — the one where six people are nodding at a slide none of them understands, waiting for a gap to pretend they got it.
The engineer who built the system, the analyst who lives in the data, the founder who has run the numbers a thousand times: they open their mouths, and something goes wrong between their head and everyone else's. They skip three steps. They use a word as if it were obvious. They answer a question no one asked. And afterward, privately, they wonder why the team is so slow.
The team isn't slow. The expert is cursed.
What the curse of knowledge actually is
In 1989, three economists — Colin Camerer, George Loewenstein, and Martin Weber — gave a name to a problem they kept finding in negotiation experiments. When people knew a piece of information, they were unable to set it aside and reason as though they didn't. The knowledge contaminated their estimate of what an uninformed person would think. They called it the curse of knowledge: once you know something, you lose the ability to imagine not knowing it.
This is not arrogance, and it is not a failure of empathy in the ordinary sense. It is a limit built into how the mind models other minds. You cannot fully un-see what you have seen. When you try to imagine a colleague's confusion, you do it from inside your own understanding, and your own understanding keeps leaking in. You correct for it a little. You never correct enough.
The most famous demonstration is almost comically simple. In 1990, a Stanford graduate student named Elizabeth Newton ran a study for her dissertation. She split people into tappers and listeners. Each tapper picked a well-known song — "Happy Birthday," a national anthem — and tapped its rhythm on a table. Then she asked the tappers to predict: what fraction of listeners will name the tune?
Tappers guessed around 50 percent. The real number was two and a half. Three songs identified out of a hundred and twenty.
Why the tappers were so wrong
Here is the part that matters for your Tuesday standup. While the tappers tapped, they heard the full song in their heads — the melody, the words, the swell of it. They could not tap without hearing it. So the bare knocks on the table felt obviously like the song, because for them the knocks were wrapped in music.
The listeners heard knocks. Morse code from a stranger.
Every expert in every meeting is a tapper. Inside their head, the idea plays in full: the context, the failed alternatives, the reason step four follows step three. What comes out of their mouth is the tapping — a compressed, de-boned version that made perfect sense to the person who could already hear the tune. They mistake their private richness for public clarity, and they systematically overestimate how much got across. Not by a little. By a factor of twenty, if the tappers are any guide.
The tells, once you know to look
The curse has a signature. You can catch it in yourself and in others once you know the shape.
There is the buried assumption — a term of art dropped in without translation, because the speaker has forgotten it was ever unfamiliar. There is the compressed leap, where two thoughts that are adjacent in the expert's mind arrive with the connective tissue removed. There is the answer to the wrong question, when someone asks for the what and receives the how, because from inside the knowledge the distinction has dissolved.
And there is the quiet one, the most expensive: phantom agreement. Nobody objects, because objecting means admitting you're lost, and everyone assumes they're the only one who is. The expert reads the silence as consensus. The meeting ends. Six people walk out with six different guesses about what was just decided, and the decision itself was never really understood by anyone but the person who proposed it.
Why seniority makes it worse, not better
You might expect experience to cure the curse. It deepens it. The more expert you are, the more your knowledge has compiled itself into intuition — fast, wordless, unavailable for inspection. A beginner can explain a task step by step because they are still performing it consciously. A master has folded the steps into a single fluent motion and can no longer find the seams.
Researchers who study teaching call the related problem expert blind spot: skilled people underestimate how hard their domain is for novices precisely because their own competence has become invisible to them. This is why the person who knows the most is often the worst person to explain it — and why the second-best engineer, the one still close to their own learning, sometimes onboards new hires better than the genius who wrote the code.
Seniority also silences the correction. Nobody wants to tell the principal engineer or the VP that they lost the room. So the feedback loop that might break the curse — wait, can you back up? — gets suppressed by the very hierarchy that most needs it.
What actually helps
You cannot un-know things on command, so willpower is useless here. What works is building external checks that don't depend on your broken intuition about what's obvious.
Externalize before you explain. Write the idea down first. The curse operates hardest in live speech, where you're improvising from the melody in your head. On the page, the missing steps become visible — you can see the leap because you have to physically write the ground on either side of it.
Ask for the readback, not the nod. "Does that make sense?" invites a reflexive yes; it measures politeness, not comprehension. "Can you tell me how you'd explain this to someone on your team?" measures the real thing. If they can't reconstruct it, it didn't land — and now you know exactly where.
Name your assumptions out loud, on purpose. Before the meeting, list the three things you'll be tempted to treat as obvious. Say them anyway. The cost of over-explaining to the one person who already knew is trivial; the cost of losing the five who didn't is the whole meeting.
Recruit the newest person as your instrument. The most recently confused colleague is your most accurate gauge of clarity, because their tune isn't playing yet. Ask them first. Watch their face, not the senior people's.
The point isn't dumbing down
None of this is about simplifying until the idea is gone. It's about closing the gap between the version in your head and the version that reaches everyone else — the gap the curse tells you doesn't exist. Clarity is not the enemy of depth. It's the delivery system for it. The best explainers aren't the ones who know less; they're the ones who have learned to distrust their own sense of what's obvious.
And that distrust is hard to maintain alone, in the moment, mid-sentence, with the melody playing. It usually takes something outside your own head to show you where the tapping stopped making sense.
Where this leaves your meetings
Most teams never learn where their explanations break, because the breakdown is invisible from the inside — the expert felt clear, the room stayed quiet, and the misunderstanding only surfaces weeks later as work built on a decision nobody actually understood. MeetingMortem exists to make that gap visible: it looks at how your meetings actually run and surfaces the patterns — the phantom agreement, the buried assumptions, the moments a room went silent instead of clear — so the curse stops costing you quietly. If you've ever left a meeting sure it went well and been proven wrong, it might be worth seeing what your meetings look like from the listener's side.