Daniel's been thinking about action minutes ever since we did that episode on treating a casual friend like a diplomat. He wants to unpack what they actually are, why they make speakers self-edit and whether that's a feature or a bug, the specific skills required to produce them, and why this practice has near-universal application across the business world. So let's start with the obvious question — what exactly is an action minute, and how is it not just a fancy name for what we all pretend to do with a notepad?
An action minute is a record that captures only what was agreed and who is responsible for executing it. Not the discussion, not the debate, not who said what. The diplomatic origin is exactly what Daniel described — a junior diplomat sits in on a bilateral meeting with the explicit brief to produce this document, and they're not there to participate. They're there to observe through a very specific lens.
So the thing most of us call meeting minutes — the chronological "X raised a concern, Y responded, Z circled back" — that whole narrative layer gets thrown out entirely.
Completely. Traditional minutes are backward-looking. They're a historical record of the conversation. Action minutes are forward-looking. They don't care who argued what. They care what the room decided and who's doing it. The Lucid Meetings folks frame it as a format where each item has an owner and a deadline, but the diplomatic version is more radical. It doesn't just add action items to traditional minutes. It replaces the entire record with only the actions. The discussion is intentionally discarded.
Intentionally discarded is a strong phrase. Most people would find that alarming.
They would, and that's where it gets interesting. The diplomacy research is clear on this — diplomatic minutes are explicitly not a transcript. The minute-taker exercises what they call "considerable judgment" about what constitutes an agreement versus what constitutes discussion. That word, judgment, is doing an enormous amount of work.
So you've got someone in the room whose job is to decide what the meeting meant, and everything else — all the hemming and hawing and "we should probably look into" — just evaporates from the official record.
And that's the mechanism Daniel's asking about. The moment you introduce action minutes into a room, something strange happens to how people talk.
The self-editing.
The self-editing. Everyone in that room knows that whatever they say is going to be converted into a written record that gets circulated upward. The Management Study Guide research notes that action minutes force participants to be precise about commitments. But the deeper mechanism is anticipatory. People pre-edit their own statements because they know ambiguity will be resolved against them in writing.
Resolved against them. That's the phrase I want to sit with. Because it's not neutral editing, is it? If I'm vague in a meeting and someone else is writing the record, my vagueness doesn't survive. It gets turned into something concrete — and that something might not be what I intended.
Right. In bilateral negotiations, loose language creates deniability and future disputes. Action minutes force clarity in the moment because the record is the reality. If you say "we should probably look into the Q3 numbers," in traditional minutes that gets recorded as a discussion point. In action minutes, it doesn't appear at all unless someone explicitly agrees to look into Q3 numbers and takes ownership. The minute-taker has to judge whether "probably" constitutes a decision.
And nine times out of ten, it doesn't.
It doesn't. So if you're the person who said "we should probably look into the Q3 numbers" and you wanted that to be your contribution — your way of looking engaged without actually committing to anything — you just got erased from the meeting.
That's brutal. And I can see why Daniel flagged whether this is a drawback or a feature. Because if you're the person who got erased, it feels like a drawback.
The diplomatic view is that it's absolutely a feature. The self-editing is the point. Once people internalize that only commitments get recorded, they stop floating vague suggestions and start making actual proposals with clear ownership. The meeting changes.
But there's a tension here. Not every meeting is a bilateral negotiation where the goal is to produce a communiqué. Some meetings need to surface disagreement, explore options, let people think out loud. If you bring action minutes into that kind of meeting, don't you just kill the exploration?
Yes. That's the failure mode. Action minutes work poorly when there's genuine disagreement that needs to be surfaced. If the minute-taker only records agreements, they can paper over conflict. The diplomatic safeguard is that parties keep their own minutes and compare them — the agreed record is a negotiated document. In business, a single set of action minutes can create false consensus.
So the skill isn't just knowing what to include. It's knowing when to record a disagreement as an unresolved item rather than forcing it into an agreement.
That's the hardest judgment call. And this is where the cognitive load of the minute-taker becomes the real story. The skill of selective omission — deciding what to leave out — is harder than capturing what was said. It's real-time triage of verbal content into "decision-relevant" and "noise." And the person doing it is typically the most junior person in the room.
That's the part that's always struck me as slightly insane. You put the least experienced person in the room, tell them not to participate, and then their document becomes the official record of what happened. The power dynamic there is... something.
It's enormous and almost never acknowledged. The junior diplomat is producing the document that will define what the meeting meant to everyone above them. They're not participating in the discussion, but they're making a hundred micro-decisions about what counts as an agreement, what's a commitment versus a pleasantry, whose vague statement gets sharpened into an action item and whose gets dropped.
I'm trying to imagine being that person. You're twenty-four, it's your first posting, and the ambassador just said something ambiguous about troop movements. You have to decide whether that goes in the record.
And the ambassador knows you're making that decision. Which loops back to the self-editing. The ambassador isn't just self-editing for the counterparty across the table. They're self-editing for their own junior minute-taker, because they know the cable goes to the foreign ministry tonight and their ambiguity will either be resolved into something concrete or vanish entirely.
So the minute-taker is almost a silent participant. They're not speaking, but their presence shapes the conversation.
They're the audience. And in diplomacy, the audience is always the real participant. The person across the table matters less than the capital reading the cable tomorrow morning.
Which brings us to Daniel's question about universal application. Because most business meetings don't have a foreign ministry waiting for the cable. But they do have a CEO, a board, a skip-level manager who wasn't in the room.
And that's where this gets practical. Every organization has the same problem — meetings that produce discussion but not decisions. Action minutes are a forcing function for decision-making. The Lucid Meetings research frames this as accountability infrastructure. The document itself creates the expectation that commitments will be tracked.
But if you try to bring this into a regular business meeting, you're going to hit resistance immediately. People treat minutes as a CYA record. They want their objection recorded, their concern noted, their alternative proposal memorialized. Action minutes take that away.
They do. And the diplomatic insight here is counterintuitive. In bilateral meetings, the CYA function is served by the fact that the action minute is the only record. There's no competing transcript to argue against. If you want your objection recorded, you have to state it as a formal reservation, not just a comment. The format forces you to be explicit about what you're disagreeing with and why.
So in a business context, if someone says "I'm not sure about the timeline" and the minute-taker doesn't record it, that person learns very quickly that they need to say "I object to the timeline and want my objection noted" or it didn't happen.
And that changes the meeting culture. The first month, nothing gets captured because the notetaker doesn't know what counts as an agreement and nobody's speaking in commitments. The second month, the team starts pre-stating decisions because they know only commitments get recorded. By the third month, the meeting itself changes — less discussion, more decision-making.
I want to push on that timeline. Three months feels optimistic.
It probably is. But the pattern holds. I've seen teams try this and the pain point is always the same — the notetaker has to make judgment calls in real time, and they're going to get some of them wrong. Someone's going to say "I never agreed to that" and the notetaker's going to say "you said we should move forward with the vendor" and the person's going to say "I said we should explore moving forward, that's different."
And now you've got a dispute about what was said, but there's no transcript to check against because you deliberately didn't create one.
That's the tradeoff. The diplomatic solution to this is fascinating — in many bilateral meetings, there's a pre-circulated draft communiqué. The meeting isn't about generating text from scratch. It's about agreeing the text that's already been drafted. The discussion is structured around specific language, not open-ended exploration.
So the business parallel would be pre-circulating draft action items before the meeting and using the meeting to confirm them.
Yes. And that flips the entire meeting design. Agendas become lists of decisions to be made, not topics to be discussed. You're not meeting to talk about Q3 numbers. You're meeting to decide whether to investigate the Q3 variance, who's doing it, and by when. The discussion serves the decisions, not the other way around.
Which sounds efficient in theory. In practice, I suspect most organizations would find this deeply uncomfortable. People like the ambiguity. Ambiguity is deniability.
And that's why the self-editing question Daniel raised is the central tension. Is it a bug that people self-edit when they know everything's being recorded as commitments? Or is it a feature that forces clarity and eliminates the kind of vague language that lets everyone leave the room thinking they agreed to different things?
I think it depends on what the meeting is for. If the meeting is exploratory — a brainstorming session, a strategy offsite — action minutes are the wrong tool. You want the mess. You want the half-formed ideas. But if the meeting is supposed to produce decisions and it consistently doesn't, action minutes are a surgical intervention.
And most business meetings fall into that second category whether people admit it or not. The calendar invite says "discuss Q3 pipeline" but everyone knows the real question is whether we're funding the expansion or not. The discussion is a ritual. The decision is the point.
So the minute-taker's job is to cut through the ritual and extract the decision. Which brings us back to the skill question. Daniel asked what's required to craft these minutes well. And I think we've established it's not a clerical task.
It's not. The core skill is real-time triage — distinguishing between statements that constitute agreements and statements that constitute discussion. That requires understanding the subject matter well enough to know what a commitment sounds like in context. It requires enough social awareness to read the room and know when a nod is a yes versus when it's a polite way of not saying no. And it requires the confidence to leave things out.
That last one is harder than it sounds. The natural instinct when you're junior is to include everything. If you include too much, you're thorough. If you leave something out, you might be wrong.
And the diplomatic training deliberately breaks that instinct. The diplomacy.edu research emphasizes that the minute-taker is not producing a transcript. The word "considerable judgment" appears repeatedly. They're being trained to exercise authority they don't formally have.
Which is terrifying. You're twenty-four, you've been in the foreign service for eighteen months, and you're supposed to exercise considerable judgment about what the ambassador and the foreign minister actually agreed to. If you get it wrong, the consequences aren't theoretical.
No, they're geopolitical. A miscorded commitment can become a diplomatic incident. The research mentions cases where the agreed record becomes the basis for treaty language. The minute-taker's judgment call ends up in international law.
That's a lot of pressure for someone who's probably still figuring out where the cafeteria is.
It is. And yet the system works often enough that it's been the standard practice for centuries. The junior diplomat burden is a feature, not an accident. It forces the senior people to be clear because they know the person recording them might not have the context to interpret ambiguity correctly.
Wait, say that again. The fact that the minute-taker is junior is what forces clarity?
Yes. If the minute-taker were a peer who'd been in the portfolio for ten years, they could fill in the gaps. They'd know that when the ambassador says "we'll consider your proposal" in that particular tone, it means no. But the junior diplomat doesn't have that context. They record what was said, or — in the case of action minutes — what was agreed. So the ambassador has to be explicit. They can't rely on shared understanding because the person writing the record doesn't share it.
That's almost elegant. The inexperience is the mechanism.
It's counterintuitive but it makes sense once you see it. The junior minute-taker is a forcing function for precision. Everyone in the room has to speak in a way that someone with limited context can accurately capture. No shorthand, no winks, no "as we discussed previously."
So if you're bringing this into a business context, you might actually want someone relatively junior taking the minutes. Not despite their juniority, but because of it.
And it flips the usual assumption that the most senior person should control the record. The senior person has too much context. They'll hear what they expected to hear, not what was actually said.
There's a whole organizational design principle buried in here. The person who writes the record should be the person with the least prior context about the topic.
Or at least, the person whose lack of context is a known constraint that forces everyone else to be explicit. The minute-taker's blankness becomes a tool.
I'm now thinking about all the meetings I've been in where the most senior person took their own notes and then sent around a summary that bore almost no resemblance to what anyone else heard. The gap between what was said and what was recorded wasn't a bug in their process. It was the entire point.
And that's the dark side of this. The minute-taker has enormous power. They decide what the meeting meant. If you're the person writing the record and you have an agenda, the action minute format gives you more cover, not less. Because you're not distorting a transcript. You're exercising judgment about what constitutes an agreement. That's a much harder thing to challenge.
"I exercised my judgment" is a very difficult thing to argue with, especially if you're the junior person and the senior person disagrees.
It's nearly impossible. Which is why the diplomatic practice of both sides keeping their own minutes and comparing them is so important. It's a check on the minute-taker's power. In business, that almost never happens. One person writes the minutes, sends them around, and unless someone objects loudly and immediately, that's the record.
And most people don't object because they weren't paying close enough attention to remember what was actually agreed versus what was discussed. The minute-taker's version becomes the truth by default.
This is where the AI notetaking tools become interesting. Tools like Otter and Fireflies are moving toward verbatim capture. The default is record everything, transcribe everything, let the user decide later what mattered. The diplomatic insight suggests this might be exactly backwards.
Say more.
The value isn't in capturing everything. It's in deciding what matters. A verbatim transcript of a ninety-minute meeting is just a smaller haystack for the same needle. Action minutes are the needle. The diplomatic practice says the judgment has to happen in the room, in real time, by someone who understands the context enough to know what a commitment sounds like.
The next frontier for AI notetaking isn't better transcription. It's tools that produce action minutes by design, not transcripts by default.
That's a much harder problem. Transcription is a speech-to-text problem. Action minutes require understanding intent, commitment level, organizational context, the difference between "we should" and "we will," the significance of who said something versus who nodded. That's not a speech problem. That's a judgment problem.
Judgment is where the whole thing lives or dies. We keep circling back to that word.
Because it's the whole thing. The minute-taker is making dozens of judgment calls per meeting, and each one is a small act of power. Is "I'll take a look" a commitment? Is "that sounds reasonable" an agreement? Is silence consent? The format forces these questions to be answered, and the answers define what the organization does next.
I think we need to hear from someone who's actually done this job.
I was just thinking the same thing.
Hilbert: Two thousand seven. UNHCR inter-agency coordination meeting in Geneva. I was there for six weeks as a minute-taker. Only person in the room not representing a country or an agency.
What was the brief?
Hilbert: Action minutes only. Explicit instruction. No verbatim record. If it wasn't an agreement with a named owner, it didn't go in. The senior coordination officer told me on day one — "your job is to tell me what we decided, not what we talked about."
How did that feel, sitting there as the filter?
Hilbert: Terrifying for about three days. Then you realize something. You're the most powerful person in the room by accident. The senior diplomats are arguing about wording, and you're sitting there thinking — I decide what this meeting meant. I could leave out the thing the French delegate said about funding and nobody would know.
Did you ever?
Hilbert: No. But I thought about it every single day. And then I made a mistake anyway. Third week, I omitted a commitment from one of the donor countries — they'd agreed to provide supplemental funding for a camp expansion in eastern Chad. I was tired, the meeting ran long, the agreement was buried in a lot of conditional language. I didn't catch it.
What happened?
Hilbert: Nothing for three weeks. The minutes went out, nobody flagged it. Then the funding didn't arrive, the camp coordinator followed up, and the donor country said — we never made that commitment. It's not in the minutes.
Even though they had made it.
Hilbert: They had. I was there. I heard it. But it wasn't in the minutes, so it didn't happen. The action minute had literally rewritten reality. We ended up having to reconvene a side meeting to re-establish the commitment, which was humiliating for everyone. The donor country wasn't trying to welch — they believed they hadn't agreed because the record said they hadn't.
That's the thing we've been circling. The minutes don't record reality. They create it.
Hilbert: That's when I understood the responsibility. You're not a notetaker. You're an author. The document you produce isn't a reflection of the meeting. It is the meeting, as far as anyone outside the room is concerned. And if you get it wrong, the wrong thing becomes true.
Did the process change after that?
Hilbert: They started circulating draft action items before meetings. The meeting became about confirming the draft, not generating it from scratch. Fewer mistakes. But also less flexibility. The pre-circulated draft shapes the conversation as much as the minutes do. Maybe more.
That matches the diplomatic practice of pre-circulated communiqués. You're not discovering what you agree on. You're ratifying it.
Hilbert: That's fine if the agreement is genuine. It's dangerous if the draft is someone's wish list and the meeting is just a rubber stamp. I saw that too. The draft action items would arrive from the secretariat and the meeting was really just an hour of people nodding at things they hadn't actually agreed to.
The pre-circulation solves one problem and creates another.
Hilbert: Everything about this practice solves one problem and creates another. That's why I never recommend it unconditionally. It's a tool. Sharp on one side.
The false consensus problem. If the minute-taker only records agreements, they can make a room full of people who sort of nodded look like a room full of people who committed.
Hilbert: The nod is the most dangerous sound in any meeting. Means twelve different things depending on who's doing it. The minute-taker has to decide which meaning goes in the record.
Did you develop a heuristic for that? Nod equals yes versus nod equals I heard you?
Hilbert: I started watching hands. If someone nodded but didn't touch their pen, it wasn't a commitment. If they wrote something down, it probably was. Not foolproof. But better than guessing.
That's a remarkably physical cue for something so abstract.
Hilbert: The body knows what the mouth hasn't said yet. I learned more about negotiation watching people's hands than listening to their words. The words are the performance. The hands are the truth.
You were twenty-something, doing this in Geneva, with real consequences.
Hilbert: Twenty-four. First real assignment. The UNHCR gig was supposed to be three weeks, got extended to six. I still have the badge somewhere.
The question I keep coming back to is whether this practice scales down. Can a five-person startup use action minutes without the diplomatic infrastructure around it?
Hilbert: They can, but they need one thing the diplomats have that startups don't. A shared understanding that the minutes are the record and there's no appeal to an unwritten version of what happened. If you don't have that norm, action minutes become a weapon. Whoever writes them wins.
In a startup, the person writing them is usually the CEO.
Hilbert: Which is exactly the problem. The most powerful person in the room already controls the narrative. Give them action minutes and you've given them a tool to make their version of events the only version. That's not accountability infrastructure. That's authoritarianism with a meeting agenda.
The safeguard is either multiple minute-takers or a culture where the minutes are contestable.
Hilbert: Or a minute-taker who isn't the decision-maker. The power has to be separated. The person who decides what was agreed can't be the person whose interests are served by a particular version of the agreement. Otherwise you're just writing your own homework.
That's a structural insight. The minute-taker should report to someone who wasn't in the room.
Hilbert: That's how the UN system works. I reported to the coordination officer, not to any of the country delegates. My loyalty was to the record, not to any participant. Without that separation, the whole thing collapses into politics.
In most businesses, the person taking minutes reports to someone in the room. Usually the person running the meeting.
Hilbert: Then they're not taking minutes. They're taking dictation.
I think that's where we need to land. The practice is powerful, but the power has to be structured correctly or it becomes the opposite of what it's supposed to be.
The open question Daniel's prompt leaves us with — if action minutes create reality rather than recording it, who decides what counts as an agreement? The minute-taker has enormous power, and that power is almost never acknowledged in organizational design. We spend a lot of time thinking about who runs the meeting. Almost no time thinking about who writes the record.
As AI notetaking tools get better, the default is moving toward verbatim capture — record everything, sort it out later. The diplomatic insight suggests that's a mistake. The value isn't in capturing everything. It's in deciding what matters. The next frontier might be AI tools that produce action minutes by design, not transcripts by default. But judgment is the hard part, and judgment is where the humans are going to stay relevant for a while yet.
Thanks to Hilbert Flumingtop for producing. This has been My Weird Prompts. If you enjoyed this, leave us a review wherever you get your podcasts — it helps. We'll be back soon.
