EXPOSING FAKE GURUS IN 30 SECONDS FLAT

Who Owns the Failure

The guru trade has been running the same con for 2,400 years. Here is the test that catches it.

In 1908, by his own account, a twenty-five-year-old magazine writer named Napoleon Hill was shown into Andrew Carnegie's mansion and handed the commission that would shape the next hundred years of self-improvement. Carnegie, the story goes, gave him twenty years and no money to interview the great industrialists of the age and distill the laws of success. Hill accepted on the spot. The result, eventually, was Think and Grow Rich, a book that has sold in numbers its publishers put in the tens of millions and that sits, directly or by descent, underneath almost every course, mastermind and morning-routine reel you have ever scrolled past.

Carnegie's biographer went looking for that afternoon. David Nasaw, who spent years inside the man's meticulously kept papers, put it about as plainly as a historian will: he found no evidence of any sort that Carnegie and Hill ever met.

Hill had an explanation ready for the missing evidence, and it is the detail I cannot stop turning over. His correspondence with the famous — letters from Woodrow Wilson, photographs inscribed by Edison and Bell — had been destroyed, he said, when a storage building burned in the mid-1920s.

The warehouse burned. Of course it did.

The founding document is a receipt for nothing

What Hill did leave behind, in newspaper morgues and court records, is a different kind of archive.

There was the lumber company in Alabama in 1908, which bought heavily on credit and sold cheap for cash until the creditors worked it out and Hill left the state. There was the automobile college in Washington that charged tuition while using student labour to build cars. There was the George Washington Institute in Chicago, an unaccredited school teaching the principles of success, capitalised on paper at $100,000 against actual assets valued at something like $1,200, which drew a warrant in June 1918 for violating Illinois' Blue Sky Law. There was the Federal Trade Commission action in October 1919 over false advertising in his own magazine. There was a scheme in 1922 that collected money from Ohio churches to educate prisoners, where the prison's own warden confirmed that none of it arrived.

I am not interested in Hill's private life, which was ugly, or in the pleasure of kicking a dead fraudster. I am interested in the shape of what he built, because the shape survived to this day. The founding text of the modern success industry rests on an unverifiable credential, defended by evidence that burned, written by a man whose documented commercial record is a sequence of insolvent enterprises.

It worked then , and it works now. It works so well that the structure has been reproduced, without anyone needing to be taught it, by several hundred thousand people who have never heard of Hill and would be offended by the comparison.

Not that a liar lied — liars are dime a dozen and always have been. The question is why this particular architecture is so stable that it regenerates spontaneously, in every medium, in every decade, and why the people it hurts most defend it hardest.

First, the objection, because it is a good one

Teaching for money is not the crime. Surgeons teach surgery. Trade schools charge fees and produce welders who can lay a butt joint that passes X-ray. Apprenticeships have been the transmission mechanism for competence for a thousand years and every one of them involves money changing hands. A retired operator who advises is not a parasite; she is doing what the word "profession" was invented to describe. Some skills genuinely propagate better through teaching than through doing, and a person who discovers that about themselves has learned something true rather than committed a fraud.

I sell ‘advice’. I have a small number of coaching clients I have carried for years. I write things like this one, and I would be delighted if you found them useful enough to pay for something somewhere down the track. If the test is "does money change hands for know-how," I fail it, and so does every accountant, tutor, physiotherapist and driving instructor in the country.

The ancients had this argument first and got further with it than most of the modern version manages. Protagoras, we are told by Diogenes Laërtius, charged a hundred minai — an amount so large that the figure is probably part of the legend rather than a price list. Gorgias the same. Plato sneered at them for it. Xenophon reached for language about selling yourself that I will not reproduce here.

But Plato had inherited wealth and enslaved labour, and did not need a fee. That is not a small point and modern scholarship has stopped pretending it is. The complaint about payment was partly the complaint of a man who never had to send an invoice.

What Plato was actually prosecuting was the sale of persuasion detached from truth — a technique for winning that worked identically whether or not you were right. The Sophists' offer was that you would leave more convincing than you arrived. Notice what that promise does. It cannot fail. If you win, the training worked. If you lose, you did not apply it hard enough.

Twenty-four centuries later, that is still the trick. Not the fee. The structure of the promise.

The mechanism: promises engineered so they cannot lose

The distinction that does the trick, every time:

A real transfer of skill makes a falsifiable promise. It names, implicitly or explicitly, a condition under which it has failed. Your weld passes inspection or it does not. The books balance or they do not. The patient's airway is secured or it is not. There is a moment at which reality gets a vote, and reality is not a paying customer, so it votes honestly.

The extraction model makes an unfalsifiable promise. Mindset. Energy. Positioning. Alignment. The next level. Inner game. These are not lies, exactly, which is what makes them so durable — they gesture at things that are arguably real. But none of them names a failure condition. And a promise with no failure condition has a remarkable property: when the outcome does not arrive, the shortfall has nowhere to land except on the buyer.

That is the engine. Everything else is decoration and deceit.

Once failure belongs to the student by construction, the seller has built a machine that converts every disappointment into either a renewal or a departure that costs them nothing. The student who fails and stays buys the next tier, because the diagnosis is always insufficient application. The student who fails and leaves is self-selecting out, and takes their counterexample with them. The testimonials that remain are, by definition, the survivors.

The industry occasionally writes its own indictment.

In March 2025 the Federal Trade Commission moved on an operation called Growth Cave, which sold a "Knowledge Business Accelerator" through YouTube advertisements promising twenty to fifty thousand dollars a month in passive income. Its co-chief executive, Lucas Lee-Tyson, put the offer to camera in six words.

It is literally IMPOSSIBLE to fail.

There is the first question answered in advance, unprompted, and offered as the reason to buy. Nothing could count as this not working. Around fifty million dollars came in.

The case settled in January 2026: a judgment of $48,597,538, a permanent ban on selling business opportunities, and orders requiring Lee-Tyson to sell his house and his co-chief executive to liquidate a Rolls-Royce and a Ferrari. The props of the promise, auctioned off to repay the people who had believed it.

The same structure, different vocabulary, in the FTC's 2023 complaint against the online coaching operation Lurn: Fail 98% of the Time & Still Be Able to Make $11,453 Per Month.

That is not a slip; your failure has been pre-purchased, priced in, and stripped in advance of any power to count as evidence. Lurn's programs ran into the thousands, with coaching packages to ten thousand, and the complaint said that very few consumers, if any, actually made money. The judgment entered was $14,077,121, most of it suspended because the defendants could not pay it.

The judgment against the people teaching how to build wealth was suspended on the grounds that they did not have any.

Five questions

If "would they still be doing it" is too blunt, and "do they charge" is no test at all, then something has to take their place. What follows is a set of questions with answers you can go and check, and every one of them costs an honest merchant something to answer.

A testimonial is a sample selected on the outcome. You are being shown the winners because they won, which has a proper name — selection on the dependent variable — and a long history of being noticed. Cicero tells of Diagoras being shown the votive tablets in a temple, painted by sailors who had prayed and come home. Bacon retold it in 1620 and the answer still lands: but where are they painted that were drowned after their vows?

Note what this is not. Survivorship bias is an accident — somebody forgot to count the planes that never came back. This is not an accident. Nobody overlooked the dropouts; somebody chose, one by one, which outcomes to film. A bias is something that happens to your reasoning. This is a production line.

It is not that the testimonials are fake. Most of them are probably true. It is that a hundred true testimonials tell you nothing about your odds, in the same way that a hundred verified lottery winners tell you nothing about lotteries except that they pay out.

Ask what happened to the people who are not in the video.

What passing looks like

The wreckage is the least interesting part of this. Everybody already agrees that scammers are bad, and an essay that only prosecutes is an essay that has told you nothing you can use. The harder and more useful thing is to show what a pass looks like, because passing is neither rare nor hidden.

Aswath Damodaran teaches valuation at NYU. He also gives it away — the full semester courses, the lecture recordings, the datasets, the spreadsheets, free, from a website that appears to have been built in 2003 and optimised for nothing since.

Run him through the five.

He publishes valuations of specific companies and attaches the spreadsheet, which means you can change his assumptions and get a different number to his. That is question one, answered in the most exposed way available: here is my estimate, here is the machinery that produced it, and here is reality, coming along shortly to mark it. He then goes back and grades himself. One of his follow-up posts on Tesla is titled "An Ode to Luck: Revisiting my Tesla Valuation."

Here is a man publicly assigning a large part of his own win to luck, in a market whose entire commercial model depends on assigning your loss to your mindset.

Question two he answers by accident of format — the base rate is the archive, wrong calls left where they fell, retrievable by anyone with twenty minutes and a grudge. Question three he answers every time he is wrong in front of an audience of professionals who can tell. Question four: a course ends, and then you go and value something without him. Question five: more people who can read a discounted cash flow does not degrade the value of being able to read a discounted cash flow.

You may not agree with all of his numbers, and that is the point. There is something there to be wrong about. Try that with a mastermind.

The same structure shows up somewhere much less glamorous. Watch the trades on YouTube — the machinists, the restorers, the luthiers, the people who fix electronics on camera. The part fits or it does not. The engine turns over or it sits there. Reality gets a vote on every single video, in public, and reality has never once bought a course. Those channels pass question one by the nature of the medium, which is why they can afford to be modest and generally are.

Competence, when it is real, is relaxed about being checked.

The base rates nobody puts on the sales page

They exist. Not for every operator, but for enough adjacent markets to make the shape of the distribution obvious.

In September 2024 the FTC analysed seventy multi-level marketing income disclosure statements. Even on that home ground, most participants made $1,000 or less per year. In at least seventeen of the seventy, the majority of participants earned nothing at all. The report also catalogued how the disclosures were built to mislead: leading with outliers, omitting or burying the proportion earning zero, and — this is the reliable one — ignoring participants' expenses, which routinely exceed the income.

And scamming is not reserved for wannabe gurus on YouTube - it flourishes in the hallowed halls of blue chip firms too.

Move sideways to trading education, which sells the same promise with a Bloomberg terminal in the background. ASIC reported in January 2026 that in the 2024 financial year, 68% of retail CFD investors in Australia lost money, totalling more than $458 million, of which $73 million was fees. The regulator returned nearly $40 million to more than 38,000 investors after finding that over half of the fifty-two issuers it reviewed had breached its product intervention order.

And then the coaching schemes themselves, where the numbers come from receivers and refund administrators rather than marketing departments.

The MOBE example illuminates the problem clearly. Thirty-seven thousand people, an average refund of six hundred and thirty-three dollars, against packages that ran to thousands. That is what the far end of this market looks like when a receiver finally counts it: not a jackpot with a long tail of near-misses, but a very large number of people out a few thousand dollars each, quietly, one at a time, none of them in the testimonial video.

The Australian Federal Court did not find that DG Institute's students lacked commitment. It found that specific factual representations were false — that a particular trust structure would completely protect assets from creditors when the documents supplied did not, and that the structure had been tested and upheld by the Full Court of the Federal Court in a named case, which it had not.

That is what happens when an unfalsifiable promise makes the mistake of touching something checkable. It gets checked. Both the company and its director appealed. In December 2024 the Full Court dismissed the lot.

As an aside, and backed by the full authority of one bloke's opinion and precisely no research: by my five questions, the largest guru operation in the Anglosphere has a coat of arms and a Vice-Chancellor. Universities. Ask a humanities faculty question one — what would count as this degree not working — and watch the answer dissolve into the same fog. Ask for the base rate, unfiltered by discipline and including the ones who left in second year. Ask what your unemployment costs them. Ask when it ends, now that a master's is the new bachelor's. All faculties of all universities are not a scam, but the trajectory is concerning if we apply a standard evenly rather than apply it only to people with cheap ring lights and iPhones on tripods.

What artificial intelligence actually changed

The common complaint is that AI floods the zone with slop, and that is true but it is not the interesting part.

Here is the thing to notice though: competence has always been expensive, and the signals of competence used to be expensive too — not as expensive, but expensive enough that the correlation between them held up well enough to navigate by. Writing a book clearly took years. Structuring an argument took practice. A well-made framework, a coherent forty-page document, a fluent explanation of something technical: these were never proof of expertise, but they were costly enough to be evidence of it.

The cost of the signal has now collapsed to roughly zero. The cost of the underlying competence has not moved at all.

Every heuristic you have for spotting someone who knows what they are talking about — fluency, structure, the confident handling of an objection, the apt example, the calm voice — is a signal, and signals just went on sale.

The tools aren’t bad. I use them daily and they raise the floor of what I can produce; a tool that makes a competent person more competent is not the enemy of anything. The problem is that the same tool makes an incompetent person indistinguishable from a competent one at the exact distance from which most buying decisions get made.

Which is why the standard has to get harder rather than softer, and why it has to move from style to structure. You can no longer read the prose and infer the person. Better writing is not evidence of better character and never really was; it is just that the bluff used to cost something. What you can still do is ask the five questions, because no language model can generate a published base rate that does not exist, or a refund policy the seller has not agreed to, or a defined date on which you stop needing them.

Costume is now free. Exposure is still expensive. Test for exposure.

The demand side, which is where this actually lives

There is a version of this argument that ends with unfollow them, and I have written it before and it is not sufficient. Attention is a real currency and starving the metric is a real act, but treating the guru as a supply problem misunderstands what is being transacted.

Lotto does not sell ‘tickets’ nor does it sell cash. It sells hope. The guru industry does not sell competence. It sells certainty, and it sells it to people who have just discovered that the world does not supply any.

That is the demand it is meeting, and it is a real demand, and it is not contemptible. Someone forty-eight years old whose industry has been hollowed out, whose savings are thin and whose confidence is thinner, is not stupid for wanting a map. They are doing the correct thing in a situation that offers no correct things. The guru's actual product is the temporary and expensive experience of not being lost. Competence, if it ever arrives, is a by-product.

Which means the honest version of the confrontation is not with the seller. It is with the buyer, and most of us have been the buyer.

Ask yourself what you were buying the last time you paid for something like this. If the answer was a specific skill you could name and later test, good — you were shopping in the real market. If the answer was the feeling of having done something about a situation you could not otherwise face, then the transaction worked exactly as designed, and it was never going to matter very much who was selling.

That is not an argument for cynicism. It is an argument for buying the thing you actually want. If what you want is competence, buy something with a failure condition attached. If what you want is not to be alone with a hard problem, buy that on purpose and from someone who will not pretend it is a business plan — a therapist, a peer group, an old colleague, a friend who will tell you the truth over a beer for the price of the beer.

The category error is what gets people robbed. Not the sellers, in the end, but the category error - and WE make that error.

Ask for the ashes

The industry's reflex is to treat the demand for evidence as a character defect in the person demanding it. Scarcity mindset. Resistance. Not ready. You will notice that this move only ever runs in one direction, and that it is deployed most aggressively at precisely the moment a checkable claim comes into view.

So make the claim checkable, and then check it. Not with hostility — with the flat, bored administrative persistence of someone processing a form. What would count as this not working? Where is your base rate? What does it cost you if I fail? When does this end? Would it still work if everyone did it?

Five questions. Thirty seconds. Anyone genuinely transferring a skill will answer four of them without blinking and give you a straight, uncomfortable answer to the fifth. Anyone selling certainty will do something else, and whatever that something else is, you will recognise it, because it will involve a warehouse that burned.

Napoleon Hill's papers were destroyed in a fire. The letters from Wilson, the photographs from Edison, the twenty years of interviews. All of it, gone, and the only surviving witness to the meeting that started everything was the man who said it happened.

A hundred years on, the fire is still burning. It burns in every "trust the process," every "I can't share client details," every set of results that cannot be produced because of confidentiality, every framework whose evidence base is the confidence of the person presenting it. The specifics change. The warehouse always burns.

Ask for the ashes anyway. Watch what they do with their hands.

Sources

• Federal Trade Commission, Analysis of Multi-Level Marketing Income Disclosure Statements, staff report, 4 September 2024.

• Federal Trade Commission, press release, FTC Acts to Stop Online Business Coaching Scheme Lurn From Deceiving Consumers About Money-Making Potential, September 2023; and refund announcement, 6 June 2024 ($2.4m to 1,922 consumers).

• Federal Trade Commission, FTC Takes Action to Stop Sprawling 'Growth Cave' Business Opportunity and Credit Repair Scam, March 2025; and FTC Secures Settlement Banning Growth Cave Defendants..., 27 January 2026 (judgment $48,597,538).

• Aswath Damodaran, Musings on Markets (aswathdamodaran.blogspot.com), including "An Ode to Luck: Revisiting my Tesla Valuation," January 2020; course materials and datasets at pages.stern.nyu.edu/~adamodar.

• Federal Trade Commission, press release, Federal Trade Commission Returns More Than $23 Million To Consumers Deceived by Online Business Coaching Scheme MOBE, April 2022.

• Australian Competition and Consumer Commission, media releases on DG Institute Pty Ltd and Dominique Grubisa: penalty judgment 9 April 2024; appeals dismissed by the Full Court, 20 December 2024.

• Australian Securities and Investments Commission, media release 26-004MR, 20 January 2026.

• International Coaching Federation, 2025 ICF Global Coaching Study (USD 5.34bn global revenue; 122,974 coach practitioners, up 15% since 2023).

• Matt Novak, "The Untold Story of Napoleon Hill, the Greatest Self-Help Scammer of All Time," Paleofuture / Gizmodo, December 2016, drawing on contemporary newspaper and court records; David Nasaw on the absence of evidence for a Carnegie–Hill meeting.

• Diogenes Laërtius on the fees of Protagoras and Gorgias; on the ancient reception of the Sophists, see modern reassessments of Plato's and Xenophon's polemic.