Daniel sent us this one — he's been reading diplomatic cables. The WikiLeaks ones. And what grabbed him wasn't the scandal, it was the format. This strange hybrid of clinical precision and unvarnished judgment. He wants to know how we'd adapt that kind of meeting capture for everyday life. Mortgage brokers, client calls, whatever. How do you take notes like the person whose minutes might end up on WikiLeaks.
I love this. And before we dive in — quick note, DeepSeek V four Pro is writing today's script. So if anything sounds unusually coherent, that's why.
Alright, so the thing Daniel spotted in those cables — and he's exactly right — is that they're built around a tension. On one side you've got the dry metadata. Who met, when, where, who else was in the room. On the other side you've got this almost literary judgment layer. "We were told X, we don't believe a word of it.
The technical term for that second part is the "comment" or "assessment" field. In the State Department cable format — and these are still in use by the way, Daniel asked if they're still around, they absolutely are — you've got a structured header block and then a narrative body. The header has the classification, the sending office, the subject tags, the reference codes for previous related cables. It's almost like a git commit chain for diplomacy.
A git commit chain. You really are a nerd.
I'm not wrong though. Each cable builds on previous ones. You'll see "REF: STATE 43762" and that's the previous cable this one is responding to or updating. It creates a paper trail. And that's the first thing most business meeting notes completely miss — they're orphaned. They don't link to the last meeting, they don't set up the next one.
That's the first transferable practice. Before you even walk into the meeting, know what cable you're responding to. What was the previous conversation, what were the open items, what was the stated position last time. Because without that, you can't spot the gap between what they said then and what they're saying now.
And the gap is where the interesting stuff lives. The diplomat who wrote the cable Daniel's describing — they knew what the Israeli official had said three months earlier. So when the official said something contradictory, they could flag it. "This represents a shift from the position articulated in our June meeting." That's the kind of sentence that makes a cable useful.
Let's get concrete. Daniel asked for best practices. If I'm walking into a meeting with my banker tomorrow, what am I actually doing?
Step one, you're preparing a one-liner before the meeting that captures the context. "Follow-up to March rate discussion, bank proposed 5.2 fixed, we asked for better." That's your reference line. Takes thirty seconds to write. But now everything that happens in the meeting has an anchor.
During the meeting itself — Daniel mentioned he's not a fan of Fireflies and the AI notetakers that just transcribe everything. I get why. A transcript is not minutes. A transcript is data. Minutes are judgment.
There's a really good breakdown of this in the records management literature. Transcripts capture what was said. Minutes capture what was decided and what was signaled. Those are different things. A transcript will give you twelve thousand words of a client call and bury the three sentences that actually matter. Good minutes surface those three sentences.
You're advocating for human note-taking during the meeting. Not passive recording.
I'm advocating for active capture. And this is where Daniel's observation about voice productivity apps is spot on. Right after the meeting — and I mean within fifteen minutes, while your memory is hot — you dictate your notes. Not a transcript of what happened, but your synthesis. "Met with Sarah at First National. She opened with the rate sheet from last week, seemed almost apologetic about the numbers. When I asked about the early repayment penalty, she deflected twice before giving a number. I think there's flexibility there. She also mentioned they're losing clients to the credit unions — that's useful context she probably didn't mean to share.
That's good. That's exactly the cable format. You've got the factual layer — she opened with the rate sheet, she deflected twice — and then the analytical layer — there's flexibility, she didn't mean to share that.
You've got the tone markers. " These are the creative flourishes Daniel was talking about. They're not just decorative. They're signals to your future self about where to push and where to be careful.
One thing the cables do that most business notes don't is they identify the participants with precision. Not just "Sarah from the bank." Full name, title, role in the meeting. And they note who else was present. Daniel mentioned this — "also present for these people." That matters because people behave differently when their boss is in the room.
Or when their boss's boss is on the call silently. I've seen deals shift because someone's regional director was listening in and they didn't disclose it. If you don't note who was present, you lose that context.
We've got pre-meeting context, participant identification, factual recounting, analytical overlay, tone markers. What about structure? Daniel mentioned BLUF — bottom line up front — from military email headers. Does that apply here?
The State Department cables actually use a version of this. The first paragraph of the narrative body is almost always a summary of the key judgment. "The Finance Minister signaled openness to restructuring but faces internal opposition from the Prime Minister's office." That's your BLUF. Everything after that is supporting detail.
For a business meeting, your first line after the metadata should be something like: "Sarah indicated the bank can likely move to four point nine but needs regional approval. She's motivated to close this quarter. The early repayment penalty is negotiable.
Then you unpack. "Sarah opened with the standard rate sheet. When I asked about four point nine, she didn't say no. She said 'that would require sign-off.' That's not a no. She then volunteered that the quarter ends in three weeks and she's behind target. We discussed the penalty language — she initially quoted three percent, but when I pushed back she said 'there are structures we can explore.' I interpret this as room to move.
That's a cable. That's exactly a diplomatic cable.
Notice what's not in there. There's no transcript of the small talk. There's no recounting of the pleasantries. Unless the pleasantries are data. If Sarah mentioned her kid is applying to college and she's stressed about tuition, that might be data — it tells you something about her headspace. But you're not including it just because it happened.
The diplomatic cables are ruthlessly selective. They're not trying to be a complete record. They're trying to be a useful record. Those are different objectives.
That's my main critique of the AI transcription tools like Fireflies and Otter. They optimize for completeness. Completeness is the enemy of usefulness when you're trying to make decisions.
Daniel said he finds them not that useful. I think that's why. You end up with a searchable archive of everything, and searching it becomes a job in itself.
There was a piece in the Harvard Business Review last year — they looked at companies using automated meeting transcription and found that meeting effectiveness scores actually dropped in about forty percent of cases. Not because the transcripts were bad, but because people stopped paying attention in meetings. They figured the AI would catch it.
That's the dark side. The presence of a recording changes behavior. If I know the meeting is being transcribed, I'm less likely to take my own notes. And my own notes are where the thinking happens.
The act of deciding what to write down is the act of understanding. When you offload that to a machine, you're offloading comprehension.
We're making the case for manual note-taking, followed by immediate dictated synthesis. But let's talk about the AI piece Daniel raised at the end. He mentioned using an AI agent to take your internal notes and clean them up for client distribution. What's your take?
I think that's actually the right use case for AI here. Not during the meeting, but after. You've written your candid internal notes — "Sarah seemed nervous, I think we can push harder on the rate" — and you need to send a professional summary to Sarah and her team. An AI can strip the judgment layer and reformat the action items. "Thank you for the productive discussion. As agreed, we will review the proposed terms and respond by Friday. Key points discussed included the rate structure and early repayment options." That's useful. That saves time.
The human in the loop is essential though. You have to read every word before it goes out. Because the AI doesn't know that "Sarah seemed nervous" is something you never want Sarah to see.
It doesn't know your relationship. It might strip something that was actually fine to share. Or keep something that was not. The judgment call is yours.
Let's talk about the format Daniel described in those WikiLeaks cables. The header metadata that gives it that "teletext from a bygone era" feel. What specific fields should someone adapt for their own meeting notes?
I'd say five fields minimum. Date and time — obvious but often sloppy in business notes. Participants with roles. Location or medium — was this in person, Zoom, phone? Subject line that's actually specific, not "catch-up" or "check-in." And a reference line linking to the previous related meeting or document. Those five fields take thirty seconds and they make the note searchable and contextual forever.
I'd add a sixth. Not classified versus unclassified. But something like "Internal only" or "Client-facing summary prepared" or "Contains market-sensitive information." It tells your future self how to handle the document.
That's smart. And it's exactly what the diplomatic cables do. The classification header tells the reader immediately what the stakes are. If you're scanning old meeting notes and you see "Internal — do not forward," you treat it differently than "Cleaned for external use.
Daniel also mentioned something specific about the cables that I want to dig into. He said they're often filed not by the ambassador, but by a second economic attaché. Someone not at the pinnacle. And there's something interesting about that. The person writing the cable isn't always the most senior person who was in the room.
That's a feature, not a bug. The most senior person is often the one doing the talking in the meeting. They're performing. The junior officer is observing. They're watching the other side's body language, they're catching the offhand remarks, they're noting who looked uncomfortable when a certain topic came up. The best cables are often written by the note-taker, not the talker.
Which has implications for business. If you're the CEO and you're leading the meeting, you probably shouldn't be the one writing the minutes. You were busy. You were managing the room. Your notes will be thin.
The person who spoke least in the meeting often has the best notes. They were listening.
If you're structuring this for yourself, and you're both the participant and the note-taker, you need to build in a gap. Don't try to write your synthesis during the meeting. You can't observe and synthesize simultaneously. Jot keywords, jot timestamps, jot one or two verbatim quotes that struck you. Then do the real writing in the fifteen minutes after.
The State Department actually has guidelines on this. Cables should be drafted within twenty-four hours of the meeting. The sweet spot is same day, within a few hours. Long enough that you've processed, short enough that you haven't forgotten. The fifteen-minute window Daniel mentioned is aggressive but ideal for business contexts where the stakes are lower and the details are fresher.
Let's talk about the analytical layer. Daniel said what struck him was the gap between what was said officially and what the diplomat actually believed. "We met these people but we didn't believe anything they actually said, in more or less those words." How do you train yourself to write that way about your own meetings?
You start by asking yourself two questions after every meeting. One: what did they want me to believe? Two: what do I actually believe? The gap between those answers is your analytical layer.
That's good. That's really good.
If a vendor tells you their product will be ready by Q three, they want you to believe the timeline is solid. But if they hesitated before answering, or if they qualified it with "assuming the engineering team delivers," what do you actually believe? Probably not Q three. Write that down. "Vendor stated Q three delivery but body language and qualifiers suggest Q four or later.
That's not cynicism. That's pattern recognition. The diplomat who's been burned by that government's assurances three times before is doing their job by flagging it.
It's also the thing that makes the notes valuable six months later. When the vendor misses Q three, you go back to your notes and you see: right, I flagged this. I'm not surprised. And more importantly, you have a record that you can use. "When we discussed timelines in May, I noted some uncertainty. Let's revisit.
Daniel asked specifically about meetings with mortgage brokers, bankers, clients. What's different about those contexts?
The power dynamics are different, so the analytical layer shifts. With a mortgage broker, you're the customer. They're selling to you. Your analytical layer should focus on: what are they not telling me? What questions did they deflect? What fees did they gloss over? With a client, you're the service provider. Your analytical layer is about: what are they really worried about that they're not saying out loud? What's the unstated objection?
With a banker, I'd add: what did they ask about that seemed to come from a script, and what did they ask about that seemed genuine? The scripted questions tell you what the bank is worried about. The genuine questions tell you what this specific banker is worried about.
That's the kind of distinction a diplomat would make. "The Minister delivered the standard talking points on border security, but departed from prepared remarks when discussing water rights, where his tone became notably more animated." That's a signal. Water rights matter to this guy personally. Border security is just his job.
Let's bring this down to something dead practical. Daniel said everyone has a voice recorder in their pocket. Open up your smartphone, record. What's the actual workflow you'd recommend?
Here's my workflow. Before the meeting, I open my notes app and I write three lines. One: the context — what's this meeting a follow-up to? Two: my objectives — what do I want to walk out with? Three: my predictions — what do I expect them to say? That third one is crucial because it primes you to notice surprises.
During the meeting, you're jotting. Not full sentences. Keywords, numbers, direct quotes if something is said in a specific way. And you're noting your own reactions. "That number seems high." "She avoided eye contact on the penalty question.
Then within fifteen minutes of the meeting ending, you open your voice recorder and you dictate for three to five minutes. Structured exactly like a cable. "Met with Sarah Chen, Senior Loan Officer at First National, also present was her assistant Mark. Location: their downtown office. Subject: rate negotiation for property on Oak Street. Reference: our March twenty-second meeting where they quoted five point two fixed. Summary: Sarah signaled flexibility on rate, likely to four point nine with regional approval. She's motivated to close this quarter. Early repayment penalty appears negotiable. Details: Sarah opened with...
Then you've got a voice memo. What do you do with it? Keep it as audio?
Most phones do this natively now. The transcription doesn't have to be perfect. You're not publishing it. You just need it to be searchable. The audio is your source of truth, the transcript is your index.
This is where the AI piece Daniel mentioned becomes genuinely useful. You take your dictated internal notes — candid, unvarnished, full of judgment — and you run them through an AI with a prompt like: "Extract the action items and create a professional summary suitable for sharing with the other participants. Remove all internal commentary, tone assessments, and strategic observations.
You read every word of the output before it goes anywhere. I cannot stress that enough.
The AI is your assistant, not your replacement. You're still the diplomat.
Let's talk about a mistake I see a lot. People conflate meeting minutes with a to-do list. They walk out with action items and think that's sufficient. Action items are the output of a meeting, not the record of it. Six months later, "Sarah to send rate sheet" is useless. You need to know why she was sending it, what the context was, what she said about it.
The action item is the what. The cable is the why. And the why is what compounds over time. After ten meetings with the same banker, you've got a narrative. You know how she operates, what she cares about, what her constraints are. That's leverage.
It's also protection. If a deal goes sideways and someone asks what happened, you've got contemporaneous notes that show your thinking at the time. Not reconstructed memory, which is notoriously unreliable. Courts love contemporaneous notes for exactly this reason.
Daniel mentioned the golden hour for contemporaneous notes in a previous conversation. That's a real concept. Memory degrades fast. Within twenty-four hours, you've lost about forty percent of the detail. Within a week, you're reconstructing, not remembering.
Reconstruction is dangerous because you fill gaps with assumptions. You remember what you expected to happen, not what actually happened. The diplomat who waits three days to write their cable is writing fiction.
The workflow is: prepare context and predictions before, jot keywords and reactions during, dictate full synthesis within fifteen minutes after, transcribe for searchability, optionally run through AI for a client-facing version. That's the system.
File it somewhere you can find it. The State Department has an entire records management system. You need something. Could be as simple as a folder structure. "Banking — First National — Sarah Chen" with date-stamped files. But it has to be consistent.
I'd add one more practice from the diplomatic world. They number their cables sequentially. You could do the same. Every significant meeting gets a number. "Meeting note forty-seven, May fifth, Sarah Chen." It makes referencing trivial. "See note forty-two for her previous rate position.
That's elegant. And it creates a chain of custody for your own thinking. You can trace how your understanding evolved.
Let's address something Daniel said that I think is worth unpacking. He mentioned that the cables have "creative flourishes." They're not just dry bureaucratic documents. Some of them are well-written. There's a craft to them.
The best cables are almost literary. There's a famous one from the US embassy in Moscow from the nineteen nineties that described a meeting with a Russian official as "a masterclass in strategic ambiguity, delivered with the enthusiasm of a man reading a weather report." That's not bureaucratic language. That's a writer.
It's more informative than "the official was non-committal." The image sticks. You remember it. Six months later, you still know what that meeting felt like.
Don't be afraid to write well. Your meeting notes are for you. If a metaphor captures the tone better than a clinical description, use the metaphor. "Sarah delivered the rate reduction like someone confessing to a minor crime — reluctant, but clearly relieved to get it off her chest.
That tells me more than "Sarah offered a reduced rate." It tells me how she offered it, which tells me how to negotiate with her next time.
The caveat is that this is for your internal notes. The client-facing version gets the clinical version. "Sarah proposed a revised rate of four point nine percent, subject to regional approval.
The internal note is your diary. The external note is your handshake. They serve different purposes, they have different audiences, they should sound different.
This is where Daniel's AI pipeline idea shines. You write the diary entry. You think freely, you write colorfully, you include your hunches and your suspicions and your tactical thoughts. Then the AI strips it down to the handshake version. You review, you send. Two documents, one effort.
Let's talk about what not to do. What are the common failure modes?
Biggest one is writing minutes that are just a chronology. "We discussed the rate. Then we discussed the penalty. Then we discussed the timeline." That's a transcript without the words. It has no analytical value. You need to answer: what changed? What did you learn? What's the judgment?
Second failure mode: writing for an audience that doesn't exist. People write meeting notes as if they'll be subpoenaed. They sand off every edge, remove every opinion, and end up with something that could have been written by anyone who wasn't in the room. That defeats the purpose.
Third: waiting too long. The fifteen-minute rule is real. After an hour, you've lost nuance. After a day, you've lost accuracy. After a week, you've lost the thread entirely.
Fourth: no reference to previous meetings. Every note is an island. You can't track change over time.
Fifth: no participant list. You'll forget who was there. I promise you will forget.
I've forgotten people who were in meetings last month. If you asked me who was on that Zoom call, I'd probably miss one or two.
That's why you write it down.
Let's bring this back to Daniel's original observation. He said reading those WikiLeaks cables changed how he thinks about meeting documentation. The format is the lesson. What's the one thing a listener should take away from this?
Write your meeting notes like a diplomat filing a cable. Facts and judgments separated but both present. Written within hours. Filed where you can find it. And honest — sometimes brutally honest — about what you actually think happened.
The honesty part is hard. It's uncomfortable to write "I don't believe the timeline they gave us." It feels uncharitable. But that discomfort is the signal that you're writing something useful. If your notes feel safe, they're probably worthless.
There's a reason those cables were classified. Not because the information was always sensitive, but because the candor was. Governments need their diplomats to be honest internally in ways they can't be publicly. You need the same thing for your business relationships. An internal record that tells the truth, even when the truth is awkward.
You need the discipline to keep that internal record separate from what you share. Daniel's right that AI can help with the separation. But the separation itself is a human practice. You have to know the difference between what you think and what you say.
That's diplomacy in a sentence.
Alright, let's do a quick summary of the format. If someone wants a template, what are they writing?
Date, time, location or medium, participants with roles, subject line, reference to previous meeting. Then the body. One-paragraph summary — the BLUF. Then the detailed account, mixing factual recounting with analytical observations. Then action items, clearly separated. Then a final line with the classification — internal only, external ready, whatever applies.
The whole thing should be skimmable in under two minutes but detailed enough that you can reconstruct the meeting six months later.
That's the balance. Brevity with depth. It's hard to do well, but it's a skill that compounds. Every meeting you document this way makes the next one easier, because you've got context, you've got a record of what was said, you've got your own judgment to check against reality.
You've got a trail. If someone asks why you made a decision, you can point to the meeting where the information changed. That's powerful.
It's also rare. Most people operate on vibes and half-remembered conversations. Having actual records puts you in a different league.
Daniel mentioned one more thing I want to touch on. He said he's a fervent fan of documentation, and he's tried all sorts of systems — NFC tags, barcodes, labels, engravers, markers. He settled on markers for his home inventory, but for meetings, he's been chasing this cable format. What does that tell you?
It tells me he's learned that the format matters more than the tool. A marker on a box is simple, reliable, always works. A structured note-taking format is the same thing for your brain. You don't need fancy software. You need a template you trust and the discipline to use it.
The tool can help — voice recorders, AI transcription, searchable archives — but the format is the foundation.
The format we're describing isn't new. It's been battle-tested by diplomats for decades. It works because it forces you to separate what happened from what you think about what happened, and it forces you to record both.
That's the episode. Document your meetings like a State Department cable. Context, participants, facts, judgment, all structured and filed while the memory is fresh. And if you need to share something externally, let AI clean up your candid notes — after you review every word.
Now: Hilbert's daily fun fact.
Hilbert: The national animal of Scotland is the unicorn. It has been since the twelfth century, when it was adopted as a symbol of purity and power on the Scottish royal coat of arms.
...right.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. You can find every episode at myweirdprompts.We'll be back soon.
