Sunk Cost and Motivated Reasoning: Why IQ Doesn't Protect Against Bad Decisions — Intelligence-Rationality Correlation Is Only 0.2–0.35 Author: Дністер Published: 2026-09-17T03:01:34.000Z Language: en URL: https://neurodrift.org/en/blog/chomu-rozumni-lyudy-pryjmayut-durni-rishennya/ Original (Ukrainian): https://neurodrift.org/blog/chomu-rozumni-lyudy-pryjmayut-durni-rishennya/ Tags: rationality, cognitive-bias, decision-making, motivated-reasoning, self-deception Original source: https://neurodrift.org/blog/chomu-rozumni-lyudy-pryjmayut-durni-rishennya/ The correlation between IQ and rationality is only 0.20–0.35. fMRI shows that when a smart person clinches a convenient conclusion, the brain's logic region goes dark while the reward region lights up 'like a hit for an addict'. A longread about the inner lawyer — the smartest enemy of your thinking — and how the architecture of desire grabs the wheel before intelligence can sit down behind it. ----- Up front: a smart person, certain she "just looked at the facts objectively". Through the glass you can see the truth — her intelligence is a lawyer foaming at the mouth to persuade an empty jury box, while the judge's chair gathers dust. In the aquarium an axolotl smiles calmly — a creature that never grows up. In 2006 Drew Westen slid thirty people into the tube of an MRI scanner and showed them words. Not images of horror — just quotes. Fifteen die-hard Democrats, fifteen die-hard Republicans, on the eve of the 2004 election. Eighteen sets of statements flashed on the screen, in which each person's favorite candidate contradicted himself: first promising one thing, then doing the opposite. The task — rate how hypocritical it was. Thirty adults wheeled into the most expensive instrument of the age to catch the cheapest operation of the psyche. And here is what the scan showed. When a partisan's brain digested the contradiction of the other side's candidate, cold logic was at work. But when his own man lied on the screen, the dorsolateral prefrontal cortex — the zone of conscious reasoning — didn't switch on at all. Dark. Instead, the emotional circuits lit up. And when the person finally clinched a convenient explanation ("well, actually he meant something different") — the reward circuits glowed. The same ones that give an addict a hit. Westen put it coldly: the partisan "spins the cognitive kaleidoscope until he gets the conclusion he wants". Reread that slowly. The reasoning zone — zero activation. The high zone — a flash. A "smart decision" here was not a computation but a painkiller: the brain wasn't counting — it was numbing the ache of an inconvenient truth and rewarding itself for the escape with a hit, like a dealer who is his own client. The architecture of desire grabbed the wheel before intelligence could sit down behind it. A sapper on a minefield never optimizes for speed. He moves slowly, checks every centimeter before the step, because he knows one thing the smart person at the negotiating table does not: the price of a single careless move is not an error but the end. So in the highest-stakes decisions, your brain behaves like the exact opposite of a sapper. It throws out an answer instantly, with the air of confidence, and steps forward without even glancing under its foot. And where the sapper runs his probe along the ground, your intelligence is already sprinting toward the tripwire — convinced that speed itself is competence. And here is why I'm writing this at all. For years I traded on the belief that intelligence would save me: that being the smartest person in the room was enough to keep bad decisions away. I built businesses, entered investments, exited them, and I made every fatal decision precisely when I felt smartest. Not dumbest. Smartest. And only years later did I understand that this is not a coincidence but a mechanism. This text is about two characters in your head, and you've mistakenly put the wrong one in charge. And about how one of them quietly robs you exactly when you're certain you're thinking most clearly. This mechanism has always existed. But 2025–2026 made it more destructive than any era before. Not because people got dumber — but because the tools for self-deception got smarter. ChatGPT, Gemini, Perplexity give confident, well-formatted answers to any question in seconds — and a smart person, instead of verification, gets ready-made ammunition for his lawyer. The information glut of the 2020s did not improve decisions: it handed more rounds to the inner defender. Where a biased person once had to dig up convenient data herself, now an AI assistant finds it on request and serves it with the air of an authoritative source. This is not the peak of access to knowledge — it's the peak of the confident wrong conclusion. And it is exactly now — not in the 1970s, not in the 2010s — that we need to talk about the inner lawyer seriously. The antagonist named "the inner lawyer" This story has a specific villain, and he has a face. It isn't stupidity. Not a low IQ. Not fatigue. It's a well-paid, eloquent, impeccably dressed defense attorney who lives in your head. Let's call him the Inner Lawyer. His job is not to find the truth. His job is to win the case for a verdict already passed. Desire has already decided everything: buy this company, don't exit this deal, don't admit this mistake, vote for your own side. The lawyer enters the case after the verdict — and builds flawless proof beneath it. He doesn't argue with your desire. He serves it. And, unlike an ordinary lawyer, this one issues no invoice — because you pay in years of your life, and the receipts never arrive. And here is the worst joke of the whole text, the one meant to make you squirm: the higher your intelligence, the more expensive and more convincing this lawyer. A dim defender would lose the case — he'd babble nonsense, you'd spot the holes in the logic yourself and change your mind. A brilliant defender wins any case, even a losing one: he'll find the data, build the argument, anticipate objections and demolish them. A smart person is not a person with a better judge in their head. They're a person with a more expensive lawyer. And they walk out of the meeting certain of their own objectivity precisely because they've just deceived themselves more talentedly than anyone. Memorize this phrase, because it will come back: he doesn't argue with your desire — he builds proof for it. The more talented the lawyer, the more convincingly he wins a losing case. Up front you're "just explaining the facts" — behind the glass your intelligence triumphs before a wall of convenient evidence, while the stenographer erased the record long ago. The axolotl, meanwhile, regrows another amputated limb — and another excuse. A smart person is not a person with a better judge in their head. They're a person with a more expensive lawyer. And they mistake winning in court for the truth. First piece of evidence: IQ and rationality are different axes Let's start with the biggest mistake we all unconsciously make: that smart = rational. That it's one scale. The numbers say otherwise, and they're unpleasant in exact proportion to how pleasant we thought we were. Keith Stanovich, a psychologist who has studied thinking for thirty years, measured the link between IQ and rationality proper — the ability not to be a "cognitive miser", to avoid distortions, to have the right "mindware". One clarification, without which the number lies: 0.20–0.35 is not a single correlation but a range across different tasks and subscales. Avoiding cognitive laziness gives r≈0.20–0.30; sufficient mindware r≈0.25–0.35; probabilistic and scientific thinking r≈0.20–0.35. The range describes the overall spread, not a single figure. For the non-statisticians, here's the bottom line: these are nearly independent things. To know a person's IQ and try to predict whether they think rationally is roughly like knowing a basketball player's height and trying to predict whether they sing well. There's some hint of a link. But betting money on it is suicide disguised as statistics. Stanovich gave the phenomenon a name: dysrationalia — the inability to think and act rationally despite perfectly sufficient intelligence. An IQ test, he says, measures the processor's power. But it says nothing about which task you aim that processor at. You can own the fastest processor in the world and spend all day flawlessly computing how to justify your own stupidity. That's the architecture of desire: the question isn't "how powerful is your mind" but "whom does it serve". And by default, it serves the lawyer. Now the sharpest part. You might think: but at least smart people see their own biases better? No. West, Meserve and Stanovich (2012) measured the "bias blind spot" — a person's confidence that they're less biased than others — and got a result exactly opposite to intuition: all markers of cognitive sophistication correlated positively with the size of the blind spot. They didn't shrink it — they grew it. Higher ability = a bigger blind spot, not a smaller one. Smarter people see their own distortions worse. Intelligence here is not glasses but a rearview mirror, where objects conveniently appear smaller than they are — and the more expensive the optics, the more convincingly the reflection lies. So the lawyer doesn't just win the case. He also convinces you there was no trial at all — that you simply "looked at the facts objectively". He doesn't argue with your desire — he builds proof for it, and then erases the transcript of the hearing. A fair objection: maybe IQ does save you after all? Before going further — an honest stop. Not everyone who defends a link between intelligence and good decisions is wrong. Research really does show: higher IQ correlates with longer health, higher income, fewer serious accidents, lower risk of chronic illness. Stanovich himself never claimed IQ is useless — he said IQ and rationality are not the same thing. A smarter person does, on average, make more careful everyday decisions: won't forget to check meds for interactions, will better assess risk statistics, will avoid an obvious subscription scam. But here's what that doesn't explain: why exactly in the highest-stakes decisions — where identity, status or big money is on the line — the statistical protection of IQ vanishes. Zagorsky showed it with bankruptcy. Tetlock — with expert forecasts. Kahan — with polarization among the most numerate. In other words, IQ is body armor that holds flawlessly at the range and disappears exactly in the patch of field where they're actually shooting: where the decision touches the ego. And it's precisely these decisions that cost a person the most. The rest is just noise on the practice range. Second piece of evidence: the better you count, the harder you lie If the previous point could still be written off as abstraction, this study hits precisely and painfully. Dan Kahan and colleagues took over a thousand people and gave them a dull "numeracy" task — the ability to read data off a table. A classic problem: does a new cream reduce a rash? The numbers are built so that a surface glance gives one answer and the correct calculation gives the opposite. The result was expected: people with high numeracy solved the task better. Logical. And now the trick. The same people were given the same numbers, with the same mathematical structure — but labeled differently. Not "rash cream", but "data on whether a ban on carrying guns lowers the crime rate". And a reverse miracle happened. Accuracy among the numerate fell. And polarization between camps didn't shrink — it grew precisely among the most numerate. That is, the best counters diverged hardest into their ideological corners. Give a person a calculator — she'll count. Tell her that the answer determines who she is — and the calculator starts spitting out exactly what she needs to hear. Think about the mechanics. The same person. The same brain. The same math. The only difference: whether the answer touched identity. When it was about cream — the skill of counting served the truth. When it was about guns — the same skill instantly switched sides and started serving the tribe. High intelligence didn't make the person a better judge. It made them a better lawyer. This is a fatal blow to our favorite illusion: "give me more data and I'll decide correctly". More data is more ammunition. And which way to shoot is decided not by the data. It's decided by the lawyer. And the smarter you are, the better he shoots — just not at the target you'd find it comfortable to think. More data is more ammunition. And which way to shoot is decided not by statistics, but by your inner lawyer. The smart one just hits more accurately. Third piece of evidence: everyone digs the pit, even the most select Maybe all this concerns some abstract "people", and the real elite — students at the planet's best universities — thinks differently? Shane Frederick tested this with a task you're about to fail yourself, or just failed. A bat and a ball together cost $1.10. The bat costs $1.00 more than the ball. How much does the ball cost? Almost every brain instantly throws out an answer: 10 cents. Fast, obvious. And wrong. If the ball costs 10¢ and the bat is a dollar more, then the bat is $1.10, together $1.20. The correct answer is 5 cents. Intuition lied, and lied confidently — without hesitation, with the air of a person who knows for certain. Frederick gave this test of three such tasks to students. Look at the result and feel the chill: Cognitive Reflection Test (CRT): average score out of 3 at elite universities UniversityAverage score (out of 3)What it means MIT2.18The best result — and even that isn't full marks Princeton1.63Less than two out of three Harvard1.43Less than half Bat-and-ball problem>50% wrongOver half at MIT/Princeton/Harvard gave the intuitive wrong answer on this very item Let me be honest, because the lawyer adores round numbers: "over half get it wrong" is about one item of the test, that same bat-and-ball problem, not all three. On the full test MIT averages 2.18 out of 3 — the best result, and even that isn't full marks. But the point only gets sharper for it: on the simplest braking question, half of the planet's most select students fell into the same pit as a supermarket cashier. Not because they're dumb — they obviously aren't. But because intelligence and the braking of intuition are different muscles. A smart brain generates a brilliant answer faster than anyone. But who checks whether the first one is correct? The sapper runs his probe before the step — slowly, because he knows the price. The MIT student steps onto "10 cents" without glancing under his foot, because he's used to his first thought always having been right. The speed of computation and the brake of self-checking are not one muscle. We train the first all our lives and never the second. And then we wonder why the fastest crash the loudest. Fourth piece of evidence: intelligence doesn't insure even the wallet "Fine," the skeptic will say, "logic puzzles, sure. But in real life smart people are richer, more successful, they don't go bankrupt." Jay Zagorsky took data on 7,403 Americans from a multi-year panel study and checked directly: does high IQ insure against a financial hole? The relationship turned out not to be linear but quadratic — that is, curved and treacherous. The share of those who maxed out their credit card to zero doesn't fall evenly with intelligence. It creeps from 7.7% in the lowest group (IQ ≤75) to a peak of 12.1% at IQ 90 — and yes, you read that right: the peak of the financial hole falls not on the dumb but on those slightly above average. Then it drops to 5.4% at IQ 115 — and creeps back up again (the journal Intelligence, 2007). Zagorsky's conclusion was merciless to our ego: high intelligence doesn't insure against bankruptcy, missed payments and a financial hole. Smart people go to zero too — they just assemble, beforehand, the most convincing presentation of why this time the risk was justified. The only investor who buys that pitch sits in the same chair as the presenter. And now picture the most expensive historical example of how smart people escalate a losing bet. 1962, the budget estimate for the supersonic Concorde airliner — £70 million. British and French engineers and officials — the best technical minds of two nations. By 1976, the final development cost reaches £1.5–2.1 billion. An overrun of 21–30 times. The British government privately admitted it was a "commercial disaster which should never have been started". And still poured money in. Why? Because to exit meant admitting that the previous smart people — they themselves — had been wrong. Here the lawyer plays full out: "we've already put in so much, we can't just drop it all now". "A little more" is the official currency of sunk costs: you pay in it for twenty years, and it converts into real money exactly once — at the moment the project is finally written off. This is the classic sunk-cost fallacy, and its genius is that it's dressed in prudence. It sounds like thrift and is the opposite: you burn the second billion so as not to admit the first one is gone. The smarter the person, the more elegant the argument they'll build to avoid exiting a losing game into which they've sunk their ego. Concorde is not a story about stupidity. It's a story about twenty years of brilliant rationalizations of what everyone already knew in their heart. "We've already put in so much — we can't drop it now." Up front: dignified prudence; behind the door an accountant pours the second billion into the bonfire so as not to admit the first one burned, while a clerk turns away from the grapes, muttering they're sour anyway. The axolotl has grown fat and smiles contentedly amid the wreckage — it still has never been wrong. Concorde is not a monument to stupidity. It's a monument to twenty years of brilliant rationalizations of what everyone had long known in their heart. Fifth piece of evidence: expertise gives confidence, not accuracy Philip Tetlock ran a study that should have closed down a whole profession, but didn't, because the lawyer is stronger than the data. Twenty years, 284 expert forecasters — the people paraded on television as oracles — 82,361 concrete predictions about politics and economics. The result: the average expert turned out only slightly more accurate than a "dart-throwing monkey" and often lost to simple mathematical extrapolation. But that's not the most important part. The most important part is that forecast accuracy did not rise with status, expertise or access to information. More knowledge did not mean better predictions. The monkey, at least, isn't paid a fee and seated in a studio to explain why it was right when it wasn't. Tetlock split the experts into two types. "Hedgehogs" know one big idea and see the whole world through it — they're confident, charismatic, beloved guests in the studio. "Foxes" hold many models in their heads, doubt, say "it depends". And it was the hedgehogs — the confident carriers of one big truth — who lost to the foxes systematically. More knowledge without doubt and self-correction gives not accuracy. It gives confidence. And a confident expert is a lawyer with a doctorate: he isn't wrong less often, he defends his error more beautifully and more loudly, while you nod. Stanovich separately formalized the face of our antagonist. He showed that myside bias — bias in favor of one's own side — is special among all cognitive distortions: ubiquitous, unpredictable and, crucially, uncorrelated with intelligence, executive functions or any prized traits of the mind. That is exactly why the cognitive elite sincerely consider themselves "objective and purely rational" — though they're biased exactly as much as everyone else. Their bias is just dressed in a better suit and looks down on everyone else's. Where the lawyer came from: he's not a bug, he's a feature Here an honest question arises: if self-deception is so harmful, why didn't evolution file it off? The answer is uncomfortable. Because it was never a bug. It's a feature. Nature didn't err in installing a lawyer instead of a judge: it optimized you not for truth but for survival in the pack. Look at Peter Wason's selection task, 1966. People are given four cards and a logical rule to verify. In abstract form, fewer than ~10% solve the task correctly. Why? Because the brain instinctively looks for cards that confirm the hypothesis, not ones that could refute it. This is confirmation bias, a term Wason himself introduced back in 1960. And now the same logical task, but in a real-life wrapper: you're a bouncer at a bar, the rule is "anyone under 18 doesn't drink alcohol", in front of you are cards with ages and drinks, whom do you check? Success jumps to 75–85%. The same brain, the same logic — suddenly it works. Because your inner logician evolved not for abstract truth but to catch cheaters in the tribe and win social arguments. He's not a professor but a bouncer: he comes alive only when there's a stranger to catch red-handed, and dozes off when you need to selflessly check your own favorite idea. On the savanna, the one who confidently held a position and didn't doubt out loud earned status and a seat by the fire; the one who said "I was probably wrong" lost face — and with it, dinner. So the inner lawyer is not a malfunction but an evolutionarily polished organ of survival. He brilliantly does what he arose for: holds your status, wins the argument, defends the ego from the pain of defeat. He's simply catastrophically unfit for what we demand of him today: to see reality as it is, even when it's humiliating. We've put a champion lawyer in the judge's seat — and sincerely wonder why the verdicts are always in our favor while life is somehow being lost. A systems view: a decision is the output of a circuit, not an event in the head This is where we need to step back and see the whole machine. We're used to thinking a decision is an event: a smart person sits down, weighs, chooses. In fact a decision is the output of a system. At the input — identity, fear of pain, the desire to be right, the desire not to lose status. At the output — what we proudly call "my considered decision". Between the input and the output sits the lawyer, quietly editing the spec while you're sure you're writing it yourself. And the key question is not "how smart are you" but another, scarier one: what did your system quietly set as the optimization target? If the target became "to be right" or "not to feel the pain of a mistake" — all your intelligence will be thrown at that target. Flawlessly. And the greater it is, the more flawlessly you optimize the wrong thing. Genius in the service of the wrong goal is not salvation but simply catastrophe at a higher resolution. Recall the scene from "Moneyball". A room of seasoned baseball scouts discusses players. Elderly men with decades of experience deliver verdicts "by the nose": this one hits the ball "like a real man", and that one isn't worth taking, because he has an ugly girlfriend — a sign, supposedly, of insecurity. When Billy Beane brings statistics, the head scout throws it in his face: "Baseball isn't science. If it were, anyone could do what we do. But they can't, because they don't have our intuition." There it is, the architecture of desire in pure form. Decades of "expert gut feeling" that statistics later showed to be systematically wrong. Experience here didn't reduce the error — it stamped it "tested over the years". The goal of these scouts' system had long stopped being "guess who'll play well" and quietly become "confirm that our nose is still worth something". And all their brilliant, detail-packed intelligence worked on exactly that. The lawyer defended their identity, not searched for the best player. A smart decision fails not when intelligence is lacking, but when the optimization target quietly became "to be right" rather than "to see reality" — and all the intelligence obediently serves the new goal. The question isn't how smart you are. The question is what your system quietly set as its goal — to see reality, or not to feel the pain of a mistake. Who makes a living off you getting stuck in bad decisions? Not theoretically — concretely. Consulting firms that bill for months of analysis while the project is obviously burning. Financial advisors who keep you in the wrong assets, because an exit would mean admitting their mistake along with yours. The hedgehog TV experts — the very ones Tetlock caught performing at monkey level — who are paid studio time not for accuracy but for confidence. Businesses built on sunk-cost logic: "we've already put so much into this CRM, we can't switch to another now" — and a whole ecosystem of vendors that lives off it. Dysrationalia is not just a personal tragedy. It's a market. The size of this market is proportional to the average IQ of the clients: the smarter the audience, the more refined the product for their self-deception — and the higher the price tag. The group lawyer: why a room of smart people doesn't save you, it amplifies Here the most comfortable last hope kicks in: "fine, an individual deceives himself — but a meeting of smart people will cross-check one another". No. The room doesn't switch the lawyer off. It hires one for everyone — and gives him a choir. What was a blind spot in one head becomes, in a group, a shared record that no one dares challenge any longer, because to challenge it means going up against everyone at once. Look at how a sapper works as part of a demining team. The team never relies on "we all looked and agree". On the contrary — distrust of consensus is built in: one person checks the patch, then a second independently re-walks it, as if the first never existed. Because the most expensive mine is the one "everyone has already checked". Consensus on a minefield is not proof of safety but an imitation of it: everyone stepped in the same footprint and decided the footprint was safe, because the foot ahead is still intact. A corporate meeting does exactly the opposite. The more unanimous the room, the calmer it feels at heart — though the unanimity of smart people more often means not "we're right" but "the lawyer has already convinced everyone, and the dissenter is afraid of looking stupid". The psychologist Irving Janis called this groupthink and dissected it on disasters where a meeting of brilliant people unanimously approved an obvious stupidity: the Bay of Pigs, the Challenger. The group's shared identity became the optimization target — to preserve harmony, not to embarrass the leader — and the collective intelligence obediently lay down under that goal. The same mechanism as in one head, just multiplied by status and a quorum. A room of smart people doesn't switch the lawyer off. It hires one for everyone — and gives him a choir. The most expensive mine is the one "everyone has already checked". And here's the honest truth, without which the four questions below are naive. Willpower doesn't work here: you can't "decide to be objective", because the decision is made by that same lawyer. Debiasing researchers (Kahneman, Klein) found the only thing that gives a measurable effect — and it's not self-suggestion but a procedure that forcibly puts a second sapper on the field: Premortem (Klein): before the decision, the team imagines that a year has passed and the project was eaten by catastrophe — and each person writes why. This isn't "assess the risks" (the lawyer will numb them), but an order to enter from the side of defeat. Studies show: a premortem sharply raises the number of named risks, because it legitimizes a doubt that otherwise fears to open its mouth. Adversarial collaboration: two people with opposite conclusions search for the truth together under a shared protocol, not for victory. The lawyer doesn't switch on to refute his own thesis — so seat beside him a second one whose job is precisely to refute. Red team / a designated opponent: someone in the room is officially obligated to demolish the decision. Not "what do you think?" (they'll politely agree), but a formal role of executioner. Because a volunteer dissenter in a group goes extinct; an appointed one is protected by mandate. What all three share: they don't cultivate character. They engineer the field so that dissent is safe and verification is mandatory, not heroic. The sapper doesn't rely on the courage of the second man — he relies on a protocol that makes the second pass inevitable. Here is where the inconspicuous move hides: the honesty of a smart group is not a trait but an engineering structure. Either it was built into the process, or the room slides into the most expensive consensus precisely when it feels smartest. A tool: four questions that infuriate your lawyer A mirror without a tool is just well-phrased despair. So here's the protocol. It won't make you rational — rationality doesn't switch on at the press of a button. But it does what a sapper does before the step: forces you to check the ground with the probe before you shift your weight. Not on every step — on the ones where there might be a wire under your foot. The lawyer hates these questions, because each of them grabs the wheel from him precisely when he's already hit the gas. The logic is simple: before every decision whose cost of error is high, you consciously switch on what the brain switches off by default — the search for refutation instead of confirmation. One practical test above all: imagine that this decision is a pull request and you're the reviewer who didn't write it. Would you accept this PR from a stranger with such reasoning — or send it back for rework? If you'd send it back, but wave your own through — the signature is being put down not by a judge but by the lawyer. The lawyer-check protocol: reflex vs conscious action QuestionWhat the lawyer does by defaultWhat you do consciously The signature test "Would I make this very decision if it led to the opposite conclusion for me?" Looks for data that confirms what's already desired (as in Kahan's gun problem) Mentally switch camp/identity and see if the argument holds. If not — it's the lawyer, not analysis The refutation test "Which card would refute my hypothesis — and did I even turn it over?" Turns over only the cards that confirm (Wason: <10% look for refutation) Consciously name the evidence that should have persuaded you of the opposite, and go looking for exactly that The zero test "If I were entering this today from scratch, having invested nothing yet — would I enter?" Defends the already-invested ego and money (Concorde logic: "can't drop it now") Cut off the sunk costs. A past investment is not an argument for a future decision, just a check already cashed The hedgehog test "How confident am I — and is that confidence from accuracy or from one favorite model?" Confuses confidence with being right (Tetlock: hedgehogs are confident and lose) High confidence is an alarm signal, not a comfort. Look for a second and third model that give a different conclusion Easy to use, but uncomfortable. Once a week, take one decision that touches you emotionally — where you feel you're "obviously right" — and run it through all four questions. Not fast. Honestly. If you feel resistance inside, irritation, "what nonsense, it's all clear anyway" — congratulations, you've just heard the lawyer's voice. It's him, indignant that the judge woke up and asked to see the transcript. The stronger the resistance — the more important it is to take the questions to the end. Because the lawyer resists most fiercely exactly where he lied most. The finale: a court where the lawyer also sits in the judge's chair Let's return to Westen's scanner tube. A person lies in the magnet, on the screen the words of their candidate lie, and on the scan you can see what's usually hidden: the logic zone dark, the reward zone flashing the instant they clinch a convenient explanation. For the first time in history self-deception could be seen with the eyes. Not guessed — seen, as the brain rewards itself with a hit for the escape from truth. And here's the last, coldest thesis of this text. You've read it, probably, with a pleasant feeling: "yes, exactly, people are like that, my acquaintance does precisely this, our manager is a typical hedgehog". That feeling is your lawyer, who just won again. Because all the while you were reading, he quietly applied every paragraph to others. To opponents. To the dumb experts on TV. Not once — to the decision you made this week and are still proud of. He doesn't argue with your desire to be smart and right — he builds proof for it. He did it even with this text: turned the mirror into binoculars through which you survey others' flaws. This is not a reading bug. It's the antagonist's final trick, and it worked on you just now — while you were mentally assembling a list of acquaintances to forward this to. Stanovich, West, Kahan, Tetlock and Westen's tube say the same thing in different languages: you will not fire the lawyer. He's wired into the architecture deeper than any intelligence — he's older than the neocortex, he's from the savanna. IQ won't switch him off; IQ just hires him more expensively. The sapper, too, can't clear all the mines from the field — he didn't lay them there. The only thing in his power is not to step until he's checked under his foot. That's all you really control: not "becoming rational", but occasionally forcing the judge to wake up and run the probe before the step. Not always. In the highest-stakes decisions. Where the cost of an error is years of life, money, people. Because the problem was never that you're not smart enough. The problem is that your mind is the most talented accomplice in your own self-deception. And the smartest decision of your life is to admit that the smartest person in the room lies to you most convincingly of all. Especially when that person is you. And especially now, when you're already preparing the argument for why this isn't about you. Sources: Keith Stanovich — "Rational and Irrational Thought: The Thinking That IQ Tests Miss", Scientific American (IQ↔rationality correlation ~0.20–0.35; the term "dysrationalia"): scientificamerican.com — Rational and Irrational Thought West, Meserve & Stanovich — "Cognitive sophistication does not attenuate the bias blind spot", Journal of Personality and Social Psychology, 2012;103(3):506–19 (higher cognitive ability → a BIGGER bias blind spot): pubmed.ncbi.nlm.nih.gov/22663351 Kahan, Peters, Dawson & Slovic — "Motivated Numeracy and Enlightened Self-Government", Behavioural Public Policy / Yale Law (sample ~1,111; polarization rises with numeracy when identity is touched): papers.ssrn.com — abstract 2319992 Westen et al. — "Neural Bases of Motivated Reasoning", Journal of Cognitive Neuroscience, 2006; Emory University / ScienceDaily (30 partisans in fMRI; zero activation of the dorsolateral prefrontal cortex; a flash of reward circuits, "like a hit for an addict"; "cognitive kaleidoscope"): sciencedaily.com — Neural Bases of Motivated Reasoning Shane Frederick — "Cognitive Reflection and Decision Making", Journal of Economic Perspectives, 2005 (the bat-and-ball problem; MIT 2.18 / Princeton 1.63 / Harvard 1.43 out of 3; >50% wrong intuitive answer): pubs.aeaweb.org — Cognitive Reflection and Decision Making (PDF) Jay Zagorsky — "Do you have to be smart to be rich?", Intelligence, 2007; Ohio State University (7,403 people NLSY; quadratic relationship; peak 12.1% maxed-out credit cards at IQ 90 vs 5.4% at IQ 115): news.osu.edu — You don't have to be smart to be rich Concorde / sunk-cost fallacy (1962 estimate — £70m; final cost by 1976 — £1.5–2.1bn, overrun ×21–30; "commercial disaster which should never have been started"; a penalty clause for exiting the treaty): en.wikipedia.org/wiki/Concorde Peter Wason — the selection task (1966) and the "2-4-6" problem (1960, the term "confirmation bias"): <10% solve the abstract form vs 75–85% in a real-life context: en.wikipedia.org/wiki/Wasonselectiontask Philip Tetlock — "Expert Political Judgment: How Good Is It? How Can We Know?", 2005 (284 experts; 82,361 forecasts; level close to chance; "foxes" more accurate than "hedgehogs"): goodreads.com — Expert Political Judgment Keith Stanovich — "The Bias That Divides Us", MIT Press, 2021 (myside bias doesn't correlate with IQ, executive functions or prized traits): mitpress.mit.edu — The Bias That Divides Us Irving Janis — "Groupthink: Psychological Studies of Policy Decisions and Fiascoes", 1982 (unanimity of smart groups in disasters: the Bay of Pigs and others; the Challenger is cited as a textbook example): en.wikipedia.org/wiki/Groupthink "Moneyball" (2011), dir. Bennett Miller — the scouts' meeting scene and the line about intuition (a work of fiction, cited as an illustration): en.wikipedia.org/wiki/Moneyball_(film)