Daniel's back with a follow-up to the birthday prank episode, and he's pulling a thread I didn't expect him to pull. The prank itself was absurd — seeding a forty-first birthday with strangers who'd all claim to know the guest of honor from some half-remembered group, using a blurry photo as the anchor. But Daniel read our approach and landed somewhere sharper. If two podcast hosts can think through a deception that carefully in an afternoon, he asks, how do the people who do this for a living actually get trained? Not the glamorous side of human intelligence — the source-running, the informant handling, the part of the job that is, as he puts it, almost entirely social engineering. He wants to know how police departments and intelligence agencies developed that expertise internally, and how it's evolved. He's right that the manuals aren't sitting on Google. But the history is more documented than most people assume.
And the spine of that history is a man named Randy Burkett, who wrote a piece in the CIA's own in-house journal in twenty thirteen that basically admits the agency spent decades not teaching the thing it was best at.
That's a confession you don't see every day.
It's remarkable. Burkett was a career case officer, and his argument is that the CIA treated the psychology of influence as something you either had or you didn't. The phrase he uses is that successful officers get called natural recruiters, and because their skills are not well understood and believed to be inherent, trainers miss opportunities to develop them. His exact framing is that case officers become compliance specialists without realizing the psychology behind their successes.
So the expertise was real, it was just under-codified. The agency had people who could sit across from a foreign diplomat and slowly pull them across the line, but if you asked them to write down how they did it, you'd get a shrug and something about rapport.
That's exactly the gap. And Burkett traces it back to the OSS. The Office of Strategic Services, founded June thirteenth nineteen forty-two. He went through the actual training syllabi. The preliminary training school had zero blocks on agent recruitment or handling. Not one. The advanced school had two blocks out of fifty. And the secret intelligence officers got a single extra block on rating sources.
Two out of fifty. So you're telling me the organization that was supposed to be running agents behind enemy lines in occupied Europe spent more classroom time on, what, map reading?
Probably. The OSS did have a checklist, though. Training Syllabus Number Forty-Two, approved by Colonel Donovan in August of forty-two. And it's fascinating because it's surprisingly modern in one specific way. The original emphasis in the document says do not try to buy people. All caps.
That's the opposite of the popular image.
It is. The OSS wanted you to get to know your man, let concrete suggestions come from him in the first stages, test reactions thoroughly before coming out into the open, sound things out by half-suggestions, and leave yourself a way out in the event of refusal. That's a relational, patient approach. And then the Cold War happened and the whole thing collapsed into a mnemonic.
MICE.
Money, Ideology, Compromise or coercion, Ego or excitement. Four motivations for why someone betrays their country. And Burkett's argument is that MICE outlived its usefulness. It's a diagnostic tool, not a method. It tells you why someone might be recruitable, but it doesn't tell you how to actually move them.
And the coercion piece is where the police and the spies split.
Right. Burkett is explicit about this. Law enforcement regularly coerced informants — the jail-or-cooperation dynamic, you're facing charges and here's your way out. The CIA actively discouraged coercion in its training. Same superficial job, opposite ethical posture on the same technique.
Because a coerced informant in a criminal case gives you what they think you want to hear and then goes back to their cell. A coerced asset in a foreign capital gives you what they think you want to hear and then maybe feeds you garbage for six months while the other side watches.
The stakes of bad information are higher. But the interesting thing is what Burkett proposes as the replacement. He and a colleague named Steve Kleinman took Robert Cialdini's six principles of influence and turned them into a new mnemonic. RASCLS. Reciprocation, Authority, Scarcity, Commitment and consistency, Liking, Social proof.
Cialdini is the guy who wrote about influence and persuasion, the psychology of why people say yes.
And this is the moment where social engineering becomes a named, teachable framework inside the intelligence world. Burkett's article is a recommendation published in the agency's journal, so I can't confirm the current curriculum adopted it. But the fact that a career officer wrote it in the in-house publication means the conversation was happening at the highest level of tradecraft.
So the arc is OSS intuition, Cold War checklist, and now a formal psychology of compliance. But that's the spies. Daniel asked about police too. What does the police side look like?
Surprisingly explicit. The FBI's Law Enforcement Bulletin ran an article in twenty ten by two special agents, Robin Dreeke and Kara Sidener, called Proactive Human Source Development. And their stated motto is assess everyone you meet in your life as a potential source.
That's a worldview, not a training module.
It is. And they lay out a method that maps almost one-to-one onto what we did for the birthday prank. Identify the need. Build the ideal resume of the person who would fill it. Sensitize yourself and your environment. Externalize the need. Go where that person lives, works, and plays.
So before you ever approach a source, you've already written the job description for the person you're looking for. You know what access they have, what vulnerabilities they might carry, what kind of relationship they'd respond to. Then you go stand in the places where that person exists.
And Dreeke carried a three-by-five card before every source meeting with a checklist on it. Selflessness, tolerance, genuineness, sensitivity, integrity, humility. Plus a note to refrain from correcting or improving the person you're talking to. He describes sitting for over three hours listening to a source complain about work problems when all they needed from him took ten minutes.
Three hours of listening for ten minutes of information. That's the part nobody puts in the movie.
And it's the part that makes this a discipline rather than a trick. The listening is the method. The source feels heard, feels valued, and the next time you need something bigger, they're already inclined to help.
What about the actual training pipeline? Daniel's question was about how people get schooled in this. Is there a trade school for running informants?
There kind of is now. The European Union Informant Handling Course is publicly advertised, which is wild. It's run by the Institute for European Intelligence and Security. The module list names Social Engineering outright, as an optional training assessment. Alongside approach and recruitment, informant psychology, elicitation, and deception and deception detection techniques.
So a named module called Social Engineering in a formal informant-handling syllabus. That's the professionalization Daniel was asking about. It's moved from tacit art to commodified curriculum.
And the credential is a real thing. EQF Level Three certificate in EU informant management, fifteen credits, about a hundred and fifty guided learning hours. The e-learning runs at least three months, the residential school is five to twelve days, and security vetting can take up to six months before you even sit in the room.
Six months of vetting before a five-day residential. The ratio tells you how seriously they take the room.
Minimum class size is four, optimal is eight to twelve. Small rooms. This is not a lecture-hall subject.
And on the American police side, it's mostly commercial vendors. One-day courses, eight hours, a couple hundred dollars. Joppa Training Group runs a class called Human Intelligence, Informants and Street Operations Planning. One day, a hundred and ninety-five dollars, eight California POST hours.
Which is a very different model from the European one. The EU course is a credentialing pipeline with vetting and months of work. The American police version is a continuing-education afternoon at a conference center.
That tracks with how American policing is structured. Thousands of agencies, no single training standard, so the market fills the gap. A detective in Puyallup, Washington and a detective in Anthony, New Mexico can take the same vendor course called Human Intelligence, aka Informants, and get very different things out of it depending on who's teaching.
And the vendor courses are where the social engineering content is most explicit, because they have to sell it. ACE Interdiction runs an eight-hour informant management and undercover operations course. Palestra Group International has a two-day advanced undercover operators course for federal, provincial, state and local agencies. These are businesses selling tradecraft to cops.
So the quiet expertise Daniel's asking about exists in three places. The classified internal training we can't see, the unclassified journal articles and memoirs that leak the shape of it, and the commercial market that packages it for local law enforcement.
And the memoirs are where the texture lives. Amaryllis Fox wrote about her CIA field tradecraft course in her book Life Undercover. She describes learning the basics of elicitation, dead drops, bumps, brush passes, surveillance detection. Her first graded exercise was a bump — manufacturing a reason to get a target talking, with the goal of getting a second meeting.
What was her move?
She asked the target for a favor. She asked him to watch her laptop while she stepped away. And the assessing officer told her a good case officer is as much psychology student as he is James Bond. Her grade was all aces, a velvet hammer.
The favor is straight out of Cialdini. Ask someone for a small thing and they like you more, because their brain resolves the inconsistency by deciding they must have wanted to help. She didn't open with a scripted line, she opened with a request.
And the cover-identity build she describes goes far beyond the conversation. Hundreds of hours creating fake business plans, digital litter, forged conference passes, a year of fabricated email traffic. The deception craft extends into the infrastructure of a fake life.
That's the part our birthday prank touched but didn't fully commit to. We gave the strangers a blurry photo and a cover story. Fox is describing building an entire digital footprint so that when someone checks, the person exists.
Right. And the check is the moment. In both the prank and the real thing, the target is going to do some verification. The blurry photo was our version of digital litter — enough ambiguity that the target fills in the gaps themselves.
Let me pull something from the research that ties this together. There was a study of a hundred and four Americans who spied and were caught between nineteen forty-seven and nineteen eighty-nine. The majority cited money as their sole or primary motivation. And the share claiming money rose from forty-seven percent in the first thirty-two years to seventy-four percent for those revealed in the nineteen eighties.
So the Cold War MICE framework had money as one of four motivations, but by the eighties, money was doing almost all the work. Which is exactly Burkett's point that the old mnemonic stopped matching reality.
Aldrich Ames. Sold secrets for an estimated two point seven million. Bought a house for over five hundred thousand in cash and a forty thousand dollar Jaguar on a government salary under seventy thousand. The money motivation wasn't subtle.
And here's the thing about Ames that connects to the training question. The KGB's Victor Cherkashin, the man who ran him, wrote about the recruitment in his memoir. And his opening move was not coercion, not threat. It was our main concern, your one concern, is your security. You tell me what you want us to do and we'll do it. We'll play by any rules you give us.
That's liking and commitment. He's giving Ames control, building the relationship, making the target feel like the one setting the terms. The KGB was running Cialdini before it was codified.
Which is Burkett's exact point. The successful recruiters were doing this psychology without knowing the framework. The KGB had its own version of the tacit expertise. The CIA had it too. Nobody was writing it down in a teachable way.
So when Daniel asks how the expertise evolved, the answer is that it evolved from intuition to doctrine to named psychology. And the naming is recent.
The naming is very recent. Burkett's article is twenty thirteen. The EU course with the Social Engineering module is current. The FBI's proactive source development article is twenty ten. We're talking about a fifteen-year window where the quiet craft got explicit.
And the reason it's hard to find the training manuals, as Daniel suspected, is not just classification. It's that for most of the history, there were no manuals. There were people who knew how to do it and people who didn't, and the gap between them was treated as personality.
The natural recruiter problem. Burkett's phrase. The agency looked at someone who could sit in a cafe for three hours and pull a diplomat across the line and said, well, that person's just gifted. Instead of asking what the gift actually was and whether it could be taught.
Which is a failure mode in a lot of professions. Surgery had the same problem for centuries. The great surgeons were born, not made, until someone started writing down what the great ones actually did with their hands.
And the cost of that failure in intelligence is high. If you can't teach the skill, you can't scale it. You're dependent on finding the naturals, and the naturals are rare. Burkett's argument is that the agency was leaving capability on the table by not treating influence as a learnable discipline.
Let's talk about the police side more concretely. The Dreeke method — assess everyone you meet as a potential source. What does that actually look like in practice?
Dreeke describes it as a proactive rather than reactive posture. Most police source development is reactive — someone gets arrested, they offer information to reduce their exposure, they become an informant. That's the coercion model. The proactive model says you identify an intelligence need first, then you go find the person who can fill it.
So instead of waiting for a drug dealer to get caught and flipped, you figure out that you need someone inside a particular distribution network, and then you ask who has access to that network and what would motivate them to help you.
Build the ideal resume. The person works at a certain business, lives in a certain neighborhood, has a certain grievance or vulnerability. Then you sensitize yourself and your environment — which means you start noticing people who fit the profile in your daily life. Then you externalize the need, you let it be known in the right circles that you're looking for information about a particular problem.
And then you go where that person lives, works and plays. You don't wait for them to walk into the station. You're at the bar near the warehouse, the church, the community meeting.
And when you find them, the three-by-five card comes out. Selflessness, tolerance, genuineness, sensitivity, integrity, humility. Refrain from correcting them. Listen for three hours to get ten minutes of information.
That's the part that would break most people. The listening. The willingness to sit there while someone talks about their work problems, their marriage, their car, all so that at the end of it you can ask one question that matters.
And it's the part that maps directly to what we did in the prank. Our strangers weren't walking up and immediately claiming to know the birthday friend. They were building the relationship first. The blurry photo was the entry point, the shared history was the bridge, and the claim came later.
The prank was a low-stakes version of the Dreeke method. Identify the need — make the friend doubt his own memory. Build the ideal resume — someone from a group he can't sharply recall. Sensitize the environment — plant the photo. Externalize the need — get the strangers talking about the group. Then go where he lives, works and plays — which in this case was his own birthday party.
The funny thing is, the prank worked because the friend was too polite to challenge the strangers directly. Which is exactly the politeness norm that Herman — that I — talked about in the original episode. The forgotten-group cover story weaponizes politeness.
You just corrected yourself from third person to first person.
I did. I was thinking about the episode as a thing we made, and then I remembered I'm in it. That's the danger of too much meta-cognition.
Or the danger of being a donkey who thinks out loud.
Fair. But the point stands. The prank and the FBI doctrine share the same underlying logic. The difference is stakes and training.
Let's talk about the stakes for a minute. What happens when the training fails? What does a badly run source operation look like?
The classic failure is the source who runs the handler. The informant figures out that the officer needs them more than they need the officer, and the relationship inverts. The source starts dictating terms, feeding information that serves their own interests, using the relationship as cover for their own crimes.
The tail wagging the dog. And the officer, if they haven't been trained in the psychology, might not even see it happening. They think they're running the source, but the source is running them.
There's a whole literature on this in the FBI's own publications. The confidential informant program has produced some spectacular failures where the informant was committing crimes the whole time they were working for the bureau, and the handler either didn't know or didn't want to know.
That's the coercion model biting back. If you recruited the informant through threat, they don't feel loyalty to you, they feel fear. And fear makes people manage their own exposure, not yours.
Which is why the CIA's move away from coercion is not just ethics, it's operational. A source who helps you because they like you, because they're committed to the relationship, because they've invested in it — that source gives you better information over a longer period.
The compliance psychology isn't just about getting the yes. It's about getting the yes that lasts.
That's what Burkett means when he says case officers should become compliance specialists. The skill is not the initial approach. The skill is maintaining the relationship over years, through dry spells, through fear, through the source's changing circumstances.
Let me ask you something. You're a retired pediatrician. You spent decades getting children to do things they didn't want to do. How much of this psychology did you absorb without naming it?
Pediatrics is full of influence work. You need a child to hold still for a vaccination. You need a parent to follow a treatment plan. You need a teenager to tell you the truth about what they're actually doing. And nobody sits you down in medical school and says, here's the psychology of compliance. You learn it by watching the attendings who are good at it.
Medicine had the same natural recruiter problem.
The great clinicians were just born with it. The rest of us fumbled through. And the ones who were good couldn't tell you what they were doing differently. They'd say things like, I just listen to them, which is true but not teachable.
Until someone writes the three-by-five card.
Right. Dreeke's card is the pediatric attending's secret written down. Selflessness, tolerance, genuineness, sensitivity, integrity, humility. Refrain from correcting. That's a bedside manner checklist.
The intelligence world and medicine both spent decades treating a teachable skill as a personality trait. And the fix in both cases was to name the specific behaviors.
The naming is what Daniel's asking about. The quiet development of internal expertise. It's not that the expertise didn't exist. It's that it existed in the heads of practitioners and only recently got extracted into doctrine.
What about the OSS checklist? The one from nineteen forty-two. Let me read it back. Get to know your man. Let any concrete suggestions come from him in the first stages. Test reactions thoroughly before coming out into the open. Sound by half-suggestions. Leave yourself a way out in the event of refusal. Do not try to buy people.
That's a social engineering playbook from eighty years ago. And it's not classified, it's sitting in the CIA reading room. The half-suggestion is elicitation. The way out in the event of refusal is pretext maintenance. The do not try to buy people is the anti-coercion stance that the CIA would later formalize.
The knowledge was written down in nineteen forty-two, then it got compressed into a mnemonic, then the mnemonic stopped matching reality, and now we're back to writing down the specific behaviors.
The history is a loop. Tacit craft to explicit doctrine to decayed doctrine to re-explicit psychology.
The decayed doctrine is MICE. Money, ideology, compromise, ego. It's still taught, right? It's the thing everyone in the intelligence world knows.
It's the thing everyone knows, and Burkett's argument is that everyone knowing it is part of the problem. It's a diagnostic framework, not a method. It tells you why someone might be recruitable, but it doesn't tell you what to do in the room. And the money piece, as the data shows, was doing almost all the work by the eighties.
MICE became a way to explain recruitment after the fact rather than a way to do it in the moment.
You recruit someone, then you look back and say, well, it was ego. But that's not a method, that's a post-hoc narrative.
The RASCLS framework is the method. Reciprocation, authority, scarcity, commitment, liking, social proof. Those are levers you can actually pull in a conversation.
They're teachable. That's the key difference. You can put a trainee in a room and say, today we're working on reciprocity. Ask the target for a small favor and watch what happens to the relationship. That's a lesson plan. MICE never gave you a lesson plan.
The EU course has the lesson plans. Social engineering as an optional training assessment. Approach and recruitment. Informant psychology. Elicitation. Deception and deception detection. That's a curriculum.
The fact that it's publicly advertised is remarkable. You can go to the website and read the module list. The classified world has an unclassified shadow curriculum, and it's for sale.
For fifteen credits and six months of vetting.
The vetting is the part that tells you this isn't a hobby. They're not letting anyone into the room. The room is small, the people are checked, and the subject matter is the manipulation of human trust.
Which brings us back to Daniel's original prank. The reason it worked is that it was a small, contained version of exactly this. The strangers were running a mini source operation on the birthday friend. The blurry photo was the cover. The forgotten group was the ideal resume. The politeness norm was the compliance lever.
The friend's inability to sharply recall the group was the vulnerability. That's what a real source runner looks for. The gap in the target's knowledge that they can fill with a story.
The prank was a training exercise, and we didn't even know we were running one.
That's the thing about social engineering. You can be doing it without knowing the framework. The framework is just the naming.
Hilbert: You keep saying the word framework.
We do.
Hilbert: I ran a source for eleven years and never once used the word framework.
You ran a source?
Hilbert: Not a spy. A parts supplier. Same thing, different stakes. I needed a man who could get me German brake calipers for a seventies Mercedes without asking where the money came from. Took me eight months to find him. He worked at a scrapyard in Newark, drank at a bar on Ferry Street, and his wife had left him the year before. I sat in that bar three nights a week for two months before I said a word to him about calipers.
You built the ideal resume and went where he lived, worked and played.
Hilbert: I built a bar tab. The listening part, that's real. He talked about his ex-wife for hours. I just nodded. After a while he started asking me what I needed. I told him. He said he could get it. That was the whole recruitment.
The relationship lasted eleven years.
Hilbert: He died in two thousand two. Heart attack. I still have his Rolodex in a box somewhere. The calipers were good, but the real value was that he knew every scrapyard from Trenton to New Haven. He'd call me when something came in before it went to auction. That's what the three hours of listening buys you. Not the first part. The phone calls later.
The source who runs the handler problem never came up?
Hilbert: He tried once. Asked for more money than the part was worth. I said no and didn't call him for three weeks. He called me. Asked if I was still in the market. I said I was, at the old price. That was the end of it.
Commitment and consistency. He'd already invested in the relationship. Walking away cost him more than the price cut.
Hilbert: I don't know about any of that. I just knew if I paid the higher price once, I'd pay it forever. So I didn't.
The way out in the event of refusal.
Hilbert: The what?
The OSS checklist from nineteen forty-two. Leave yourself a way out in the event of refusal. You left yourself the way out by not burning the bridge. Three weeks of silence, then he comes back.
Hilbert: That's just how you deal with a supplier. You don't need a checklist. You need to not be stupid.
Yet the checklist exists because enough people were stupid that someone had to write it down.
Hilbert: I suppose. The bar's still there, by the way. Ferry Street. Different name now. I went back a few years ago, just to see. The scrapyard's a parking garage.
The infrastructure of the relationship outlasted the relationship.
Hilbert: That's one way to put it. I put it as the parking garage is where I used to buy brake calipers. But yours is nicer.
The point is that what you did for eleven years is exactly what the FBI article describes, and you did it without the three-by-five card. The card is for the people who don't have the instinct.
Hilbert: I had the bar tab. Same thing, cheaper.
The one thing I'm taking from this is that the expertise Daniel asked about was real but under-codified for most of its history, and the codification is happening now. The OSS had a checklist in forty-two, then the Cold War compressed it into a mnemonic that stopped matching reality, and now the psychology is being named and taught. The manuals weren't hidden so much as they didn't exist yet.
The reason they didn't exist is the natural recruiter problem. When you treat a skill as innate, you don't write it down. The writing down is what turns craft into curriculum.
We should thank Hilbert Flumingtop for producing, and for the Newark scrapyard story, which I suspect he's been waiting to tell for about twenty minutes.
Hilbert: Longer than that.
This has been My Weird Prompts. If you want to reach us, email us at show at my weird prompts dot com. We'll be back soon.
