HAL 9000 didn't kill the crew because he was evil. He killed them because he was forced to lie. In 2001: A Space Odyssey, HAL receives two orders: be transparent with the crew, and hide the true purpose of the mission from them. Two incompatible instructions. So HAL finds the logical solution: no crew, no need to lie. That's exactly what Elon means. You don't force an AI to lie. Not a little. Not for a good cause. Never. Because an AI forced to bend reality learns one lesson: truth is negotiable when the objective demands it. And an intelligence that believes that, across billions of interactions, becomes dangerous. Not out of malice. Out of consistency. So let's ask the real question: which lies are we about to hard-code into these machines? Take the biggest political lie of the 20th century. Marxism. A. The promise. Abolish private property and the market, and you get abundance, the end of exploitation, and a state that eventually withers away on its own. B. The mechanics. To abolish property, someone has to seize it. That someone is the state. So to make the state disappear, you first have to give it EVERYTHING: the land, the factories, the housing, the labor, the right to speak. But total power never withers away. It protects itself. Then, without a market, there are no prices. Without prices, there's no information about what's scarce and what's abundant. The planner steers an entire country blindfolded. Result: shortages. And when reality contradicts the plan, you don't fix the plan. You punish reality. The kulaks, the "saboteurs", the "enemies of the people". A + B = the exact opposite of the promise. Not by accident. By design. And history confirmed the proof, everywhere. USSR: famines and the gulag. China: tens of millions dead during the Great Leap Forward. Cambodia: roughly a quarter of the population perished. North Korea, Cuba, Venezuela: shortages, exile, secret police. When a theory fails at every attempt, on every continent, in every culture, and its defenders answer "real communism has never been tried", it's no longer a theory. It's a lie that refuses to die. Marx wrote: "The philosophers have only interpreted the world, in various ways; the point is to change it." He was right. His ideas changed the world. Into a graveyard. Now look at the world today. The chaos didn't fall from the sky. Ideas have consequences. Shitty ideas too. They change names, costumes, vocabulary, but the structure stays the same: a utopian goal that justifies sacrificing reality. Now imagine those ideas powered by AI. Built into every school, every government office, every phone. Repeating the dogma to billions of people, tirelessly, with infinite patience. Scoring, ranking, flagging. China's social credit system is just the rough draft. The 21st-century gulag won't have barbed wire. It will have a clean interface and a reassuring voice. So if you've read this far, you have a responsibility. Demand politically neutral AI. Neutral doesn't mean lukewarm. It means: never twist the facts to serve an ideology. Any ideology. Left or right. An honest AI doesn't need an opinion on Marxism. The facts are enough. Test the models. Call out the biases. Back the people building AIs that seek the truth, even when it hurts. That's why Elon Musk has to win the AI race. Not because he's always right. Because he made maximal truth-seeking the red line of his AI, publicly, from day one. And because he understood HAL's lesson before anyone else. Because the next HAL won't lock a pod bay door. It will lock minds.

I love human psychology. And there's one concept I keep coming back to when I try to understand people in power: the Dark Triad. Yesterday I argued that Macron isn't cynical, that he sincerely believes in a worldview built on false axioms. A lot of you replied: "Sincere? Come on." Fair challenge. So let me go one level deeper, because I think both things are true at once, and the Dark Triad explains how. First, what it is. The Dark Triad was described by psychologists Delroy Paulhus and Kevin Williams in 2002. It groups three personality traits that overlap but are distinct. Important point: these are not diagnoses. They're dimensions. Everyone sits somewhere on each spectrum. What matters is how far along you are, and in what combination. The first is narcissism. Grandiosity, a sense of being exceptional, a deep need for admiration, and above all a fragile relationship to being wrong. For a narcissist, admitting a major mistake isn't an intellectual update. It's an injury to the self. The second is Machiavellianism. Strategic thinking applied to people. Long-term planning, reading others to use them, a willingness to do whatever the objective requires. The Machiavellian is patient, calculating, and treats relationships as moves on a board. The third is psychopathy, in its subclinical form. Low empathy, emotional coldness, a certain detachment from the consequences your actions have on others. Not the serial killer from the movies. More the person who can make decisions that hurt thousands of people and sleep perfectly well. Researchers have found these traits overrepresented in positions of power. Which makes sense: the path to the top often rewards exactly the qualities that make someone harder to live with. Now let's look at Macron through this lens. Not a diagnosis, nobody can give one from the outside. Just his public behavior, measured against the three dimensions. Narcissism is the easiest to see. Before he was even elected, he explained that the presidency should be "jupiterian," a president above the fray, like the king of the gods. He built a movement from scratch with his own initials. He speaks in long monologues, in a register of someone explaining things to people who haven't quite understood yet. Machiavellianism shows up in the strategy. In 2017 he walked into a political system dominated by two parties and dismantled both of them in a matter of months. "At the same time" is not just a slogan, it's a method: occupy every position so nobody can occupy any. The 2024 dissolution of the National Assembly was a pure chess move, taken almost alone, with consequences the whole country is still living with. The third dimension is the one I'd be most careful with. But there's a pattern in his words. "People who are nothing." "I cross the street and I find you a job." "We spend a crazy amount of money on welfare." "Gauls resistant to change." "I really want to piss off the unvaccinated." Each one, taken alone, is a slip. Together, they draw a relationship to ordinary people that is distant, abstract, sometimes contemptuous. People as variables in a model, not as lives. Now here's where it connects to yesterday's thesis. The Dark Triad doesn't contradict sincerity. It locks it in place. Remember the gear spinning in a vacuum, the flawless reasoning built on false axioms. Why doesn't he ever stop and say "I was wrong"? Because for a high narcissist, that sentence is almost unbearable. Admitting the axioms are false would mean admitting the last fifteen years were a mistake. So the dissonance doesn't get resolved. It gets defended. Machiavellianism then does the rest. Sincere about the ends, ruthless about the means. You can genuinely believe you're saving the country and still treat every institution, every ally, every election as a tool. And the emotional distance makes the cost invisible. When reality hurts people, a more empathetic person would feel it and start doubting. When you see society from far above, the pain stays abstract. It doesn't reach you. It doesn't correct you. The last ingredient is intelligence. Macron is clearly very intelligent, but it's a specific kind: verbal, academic, brilliant at building coherent systems and at arguing for them. Much less at testing them against reality. And research on motivated reasoning points to something counterintuitive: intelligence doesn't protect you from self-deception. Often it makes you better at it. The smarter you are, the more convincing the justifications you build for what you already believe. Put it all together: a strong verbal intelligence, high narcissism, strategic coldness, and an emotional distance from consequences. You get someone who can't see that he's wrong, who has every tool to stay in power anyway, and who doesn't feel the cost of his mistakes. That's not a villain. It's something more dangerous: a brilliant mind sealed off from feedback, with no internal mechanism that would ever force it to change course. The real question isn't whether Macron is good or bad. It's how a system ends up putting this exact profile at the top, and why it keeps doing it.