# The End Is Near — AI Doom, Anti-Science Panic, and the Primates Who Sell Technological Exile ![[resources/images/the-end-is-near.jpg]] <iframe width="100%" height="20" scrolling="no" frameborder="no" allow="autoplay; encrypted-media" src="https://w.soundcloud.com/player/?url=https%3A//api.soundcloud.com/tracks/soundcloud%253Atracks%253A2402629251&color=%23ff5500&inverse=false&auto_play=false&show_user=true"></iframe><div style="font-size: 10px; color: #cccccc;line-break: anywhere;word-break: normal;overflow: hidden;white-space: nowrap;text-overflow: ellipsis; font-family: Interstate,Lucida Grande,Lucida Sans Unicode,Lucida Sans,Garuda,Verdana,Tahoma,sans-serif;font-weight: 100;"><a href="https://soundcloud.com/bryantmcgill" title="Bryant McGill" target="_blank" style="color: #cccccc; text-decoration: none;">Bryant McGill</a> · <a href="https://soundcloud.com/bryantmcgill/the-end-is-near-ai-doom-anti" title="The End Is Near — AI Doom, Anti-Science Panic" target="_blank" style="color: #cccccc; text-decoration: none;">The End Is Near — AI Doom, Anti-Science Panic</a></div> {{**Fear is politically useful because fear changes coalition behavior, voter attention, legislative urgency, tolerance for state intervention, regulatory preferences, and the distribution of institutional power.**}} [[reminders/AI-Tech Doomers/Fear Changes Coalition Behavior and Institutional Power by Bryant McGill|∴]] > [!map] AI-science and anti-science routes > **Primary collection:** [[collections/AI-Science Doomers and Anti-Science Doomers|AI-Science Doomers and Anti-Science Doomers]] · [[collections/Simple Reminders|Simple Reminders]] > **Related collections:** [[collections/Neurotech|Neurotech]] · [[collections/War With Empire|War With Empire]] · [[collections/US-Israel Leadership Alliance|US-Israel Defense Tech]] > **Related articles:** [[articles/The Evolutionary Roots of Silicon Valley|Valley Roots]] · [[articles/war with empire/The Real No Kings Moment and Who Wrote the Super-intelligence Ban|Who Wrote the Super-intelligence Ban]] > **Quotation and concept map:** [[wiki/AI-Science Doomers and Anti-Science Doomers|AI-Science Doomers and Anti-Science Doomers]] · [[wiki/Artificial Intelligence|Artificial Intelligence]] · [[wiki/AI Safety|AI Safety]] · [[wiki/Regulatory Capture|Regulatory Capture]] · [[wiki/War With Empire|War With Empire]] · [[wiki/Digital Divide|Digital Divide]] On September 16, 2026, the chairman of the Federal Communications Commission, Brendan Carr, compressed a decade of technological hysteria into one line on X: “The AI Doomers are the same people that proclaimed the end of the Internet during ‘net neutrality’ debates.” [[reminders/AI-Tech Doomers/The AI Doomers Are the Same People That Proclaimed the End of the Internet During ‘Net Neutrality’ Debates by Brendan Carr|∴]] Carr voted in December 2017 for the order repealing federal net-neutrality rules, and the forecasts that followed promised throttled speech, a tollbooth web, and **the end of the internet as we know it**; the rules returned in 2024, the Sixth Circuit struck them down in January 2025, and the network kept compounding through every reversal, carrying more video, more commerce, more science, and more of the commentary that had scheduled its funeral. Carr's post landed at the crest of the loudest week the AI-risk ecosystem has yet produced. Jacob Coxon, a twenty-seven-year-old pretraining researcher, resigned from Anthropic on September 8 accusing the frontier laboratories of gambling with human lives; his thread reportedly passed one hundred million views; Anthropic's alignment science lead publicly placed the odds of AI-caused human extinction within the decade above ten percent; Senator Bernie Sanders unveiled a bill to ban artificial superintelligence; and on September 14 the President of the United States telephoned Jensen Huang onstage at the All-In Summit to declare the whole panic a hoax. {{Failed apocalypses leave two things behind: a prophet who keeps the microphone and a follower who keeps the fear.}} The question worth carrying out of that week concerns something older than any news cycle, which is **who actually pays when an alarm call becomes a way of life**. The twentieth century gave the doom prophet a costume: the bearded man in a sandwich board reading THE END IS NEAR, a fixture of newspaper cartoons for so long that he became shorthand for apocalypse as a lifestyle. Update the costume for 2026 and set him on the Hollywood Walk of Fame. His board now reads AI WILL KILL US ALL, his coat hangs in rags, a thin dog waits at his heel, a tin can at his boots asks for a little humanity, and above him the billboards promise a brighter human tomorrow while the crowd flows past with phones in hand, stepping across a star stamped DREAM BIGGER. On a first reading the picture is a cruel joke. On the second it becomes the most accurate portrait of the doom economy anyone has drawn, because the man wearing the board is plainly the poorest person on the boulevard, and **the people who wrote the sermon printed across his chest are somewhere else, insured, streaming, verified, and paid**. {{The man wearing the sign rarely wrote the sermon printed on it.}} {{Every doom movement eventually sorts humanity into two classes: the people who sell the sign and the people who end up wearing it.}} His tin can is asking the right question of the right people, and the answer this whole transition owes him is a way back inside. {{The ugliest part of the coming digital divide is that many of the people who fall through it will not be stupid, malicious, or deserving of what happens to them; they will be **victims who were frightened into maladaptation**.}} The professional apocalypse industry keeps no ledger for these casualties, because they appear in none of the catastrophes it sells; they appear instead in unanswered benefits portals, missed oncology appointments, expired credentials, and job applications that a person was taught to fear submitting. {{Some human beings are already withdrawing from phones, networks, medicine, AI, institutions, banking, identity systems, and ordinary technological life because somebody convinced them that every interface is surveillance, every algorithm is possession, every vaccine is poison, every institution is a trap, and every machine is an enemy.}} From the inside, that withdrawal feels like vigilance, like the last sane animal in the herd refusing the poisoned water. From the outside it looks like a slow and methodical severing of every line through which modern assistance, income, diagnosis, and community now arrive. {{A frightened person who cuts every wire has been made easier to abandon, and far harder to protect.}} {{When those people become economically unreachable, medically disconnected, unemployable, institutionally illegible, or socially stranded, that is not some triumphant Darwinian correction; it is a preventable human tragedy produced partly by bad education and partly by people who discovered that fear is easier to sell than competence.}} The moral center of this whole transition therefore sits in an inversion that polite discourse has spent years refusing to state. {{The people who deserve to be **selected out of positions of influence** are the ones manufacturing that fear and exporting technological helplessness into vulnerable minds.}} Fear merchandising carries a beautifully asymmetric risk profile for the merchant: the influencer keeps two-factor authentication on the account that receives the sponsorship revenue, keeps a physician, keeps a pharmacy app, keeps a cloud backup of the episode, and ships the product through the same algorithmic distribution the product denounces. {{If someone builds an audience by teaching other people to distrust every scientific, medical, computational, and civic interface until those followers can no longer function inside the civilization around them, then the propagandist has performed something much more serious than expressing an eccentric opinion: they have converted ignorance into material disadvantage and called it liberation.}} Opinion ends where the follower's material life begins to degrade, and at that border the commentator has become an architect of someone else's exclusion. {{Those people should lose credibility, platforms, institutional authority, professional standing, and the presumption that they are harmless commentators, because their victims may eventually be the ones sleeping outside while the fear merchant remains comfortably networked, monetized, banked, insured, and online.}} {{**That is the real obscenity of the digital divide: the people selling technological exile are rarely the ones forced to live in it.**}} Selection pressure becomes a moral instrument only when it is aimed with precision, and the aim here is everything. Frightened people do die of these ideas, in hospital beds and in the slower attrition of exclusion, and their deaths correct nothing, because the idea that killed them keeps broadcasting to the next household. {{If there is a Darwinian process here, let it operate where civilization can afford it: against **maladaptive ideas, fraudulent authorities, and epistemologies that repeatedly teach people to collide with reality**.}} Status, credibility, and institutional trust are the currencies such figures harvest, which makes those currencies the precise place where the correction belongs. {{Let technological superstition lose status; let anti-scientific demagoguery become professionally disqualifying; let people who deliberately manufacture technological incapacity become social outcasts from serious institutions rather than allowing their least sophisticated followers to become outcasts from society itself.}} {{Aim the consequence at the author of the fear, and aim the rescue at the audience.}} {{The objective should not be to watch frightened people disappear beneath the machinery of modernity, but to make the people who frightened them **incapable of doing it to the next generation**.}} {{The victims belong back inside the future; the ideology that stranded them does not.}} ## The False Alarm at the Feeding Platform In 1980 Robert Seyfarth, Dorothy Cheney, and Peter Marler reported in _Science_ that vervet monkeys in Amboseli produce acoustically distinct alarm calls for leopards, eagles, and snakes, and that playing each recording back sends the troop into the escape behavior fitted to that predator: into the trees for the leopard call, down into the brush for the eagle call, upright and scanning the grass for the snake call. **An alarm call is a compressed world-model broadcast to a coalition**, and natural selection tuned it under a savage asymmetry that error-management theorists later formalized, because a false alarm costs a meal while a missed leopard costs a lineage, so the primate nervous system inherited a standing bias toward shouting early, loudly, and often. Every human being carries that bias, including engineers, including me, including every researcher who has stared at a loss curve at three in the morning and felt the grass move. {{The alarm call is older than language, and in most rooms where primates gather it still outvotes the evidence.}} Science is the cultural technology our species assembled so that coalition instinct would stop impersonating reality testing: replication in place of volume, instruments in place of vibes, falsification in place of recruitment. Wherever that technology weakens, the ancient machinery reasserts its jurisdiction. {{Every generation produces people who mistake their **fear response for foresight**. They see an unfamiliar machine, announce the end of civilization, congratulate themselves for their moral seriousness, and then quietly purchase the machine ten years later.}} Brandon Wheeler's 2009 field study of tufted capuchins at Iguazú, published in _Proceedings of the Royal Society B_, documented something every media strategist should have tattooed inside the eyelids: at contested feeding platforms, lower-ranking monkeys gave anti-predator alarm calls with no predator anywhere in sight, the dominant animals scattered, and the callers ate. The finding establishes that **tactical fear is a resource-capture strategy with deep primate roots**. At a Goldman Sachs conference on September 10, 2026, Jensen Huang said the capuchin's quiet part aloud when he suggested that AI-security alarmism also manufactures commercial demand, reported by Forbes as “What better way to create demand than to create a problem?” [[reminders/AI-Tech Doomers/Catastrophic AI Rhetoric Can Manufacture Demand by Jensen Huang|∴]] and by Axios as “Who doesn’t want their market to be hysterical about their product and line up around the corner for it?” [[reminders/AI-Tech Doomers/AI Fear Can Manufacture Demand by Jensen Huang|∴]] Michael Burry, whose own Substack is literally titled _Cassandra Unchained_, looked at frontier laboratories issuing warnings about their own products ahead of public offerings and wrote “IPOs need hype & puffery; “we are so awesome it could become dangerous” is hype & puffery.” [[reminders/AI-Tech Doomers/IPOs Need Hype & Puffery; “We Are So Awesome It Could Become Dangerous” Is Hype & Puffery by Michael Burry|∴]] Marc Andreessen had already priced the professional risk sector in 2023: “They are paid to be doomers, and their statements should be processed appropriately.” [[reminders/AI-Tech Doomers/Professional AI Doomers Are Paid to Be Doomers by Marc Andreessen|∴]] And from the opposite wing of the argument, Emily M. Bender, co-author of _The AI Con_ and no friend of the accelerationists, told KQED “I think that one of the points that we make is that doomerism is another kind of AI hype.” [[reminders/AI-Tech Doomers/Doomerism Is Another Kind of AI Hype by Emily M. Bender|∴]] The skeptic of hype and the capitalist of acceleration arrive at the same capuchin on the same feeding platform. Reid Hoffman, at Politico's Decoded summit, performed the posture back to its owners: “Look at me, I’m Cassandra, I’m saying it’s the end of time.” [[reminders/AI-Tech Doomers/AI Doom Rhetoric Casts Itself as Cassandra by Reid Hoffman|∴]] {{Every troop contains a monkey who has learned that shouting leopard is cheaper than foraging.}} {{The people I worry about are not merely the technologically frightened; they are the entrepreneurs of fear who manufacture technological helplessness in others.}} {{Teaching people to distrust every scientific and computational interface can strand them outside medicine, work, information, and opportunity while the fear merchant moves on to the next podcast.}} The coalition dynamics become almost comically legible once the vocabulary of primatology replaces the vocabulary of press releases. Advocacy organizations run budgets, fellowships, media calendars, and legislative strategies, hundreds of millions of philanthropic dollars have been directed into AI-risk institutions over the past decade, and **distributed movements still have chairs, grant cycles, and schedules**; coalitions recruit, and recruitment is intentional work. Hoffman aimed his sharpest line at frontier-lab leaders who are ahead in the race and petitioning for an industry-wide slowdown: “You’re in the lead, you can slow down anytime you like.” [[reminders/AI-Tech Doomers/Leading AI Labs Can Slow Down Themselves by Reid Hoffman|∴]] Vice President JD Vance, on the All-In podcast, delivered the same verdict in gothic register: “So if you’re going to create Frankenstein, don’t come to the government and say we need regulation.” [[reminders/AI-Tech Doomers/Do Not Build Frankenstein and Then Ask Government for Regulation by JD Vance|∴]] Huang had sharpened the point at VivaTech in Paris in June 2025, responding to Dario Amodei's warnings with “I pretty much disagree with almost everything he says.” [[reminders/AI-Tech Doomers/Jensen Huang Disagrees with Dario Amodei by Jensen Huang|∴]] and then “He thinks AI is so scary, but only they should do it.” [[reminders/AI-Tech Doomers/Dario Amodei Says AI Is Scary but Only Anthropic Should Build It by Jensen Huang|∴]] By July 2026 he was telling Axios that some companies hope government will help write regulations to their own advantage. Pedro Domingos told the _San Francisco Standard_ that Anthropic is “like a religious cult. Really. The psychology is very similar, except the rapture in this case is the singularity.” [[reminders/AI-Tech Doomers/Anthropic Is Like a Religious Cult by Pedro Domingos|∴]] Joseph Henrich and Francisco Gil-White showed in 2001 that human beings copy prestigious individuals far outside the domains in which that prestige was earned, which is how a billionaire's anxiety about superintelligence becomes a schoolteacher's certainty that her phone is possessed. {{AI doom has become a kind of intellectual luxury good: affluent people standing inside the most technologically supported civilization in history explaining that technological acceleration has gone too far. They deliver the warning through satellites, fiber optics, semiconductor fabs, hyperscale clouds, machine-learning recommendation systems, and a phone containing more computation than governments once possessed.}} {{Prestige travels farther than expertise, and fear rides prestige like a parasite rides a host.}} I traced the lethal ancestry of that same coalition machinery in [There Is Nothing More Natural Than Murder](https://bryantmcgill.com/article-human-killer-apes); the fear economy is its bloodless descendant, running recruitment and out-group formation through feeds instead of spears. Fairness to the frightened researchers matters, because contempt without accuracy is merely another alarm call. Coxon's warning was plainly sincere, he walked away from a career to deliver it, and Huang himself praised the young researcher's courage even while calling extinction forecasts ungrounded. Sincerity, however, is a measurement of the caller. {{Sincerity measures the caller; evidence measures the leopard.}} Melanie Mitchell of the Santa Fe Institute, a careful critic of AI overclaiming in both directions, reacted to the ten-percent extinction figure with “What, an Anthropic researcher states (without evidence) that there is a 10% chance that AI will cause human extinction?” [[reminders/AI-Tech Doomers/Extinction Claims Without Evidence Do Not Justify Regulation by Melanie Mitchell|∴]] The investor class responded with less restraint. Brad Gerstner wrote “Ridiculous hyperbole from an ex junior employee who worked a total of 6 weeks at Anthropic now cause for regulatory intervention? No thx.” [[reminders/AI-Tech Doomers/Six Weeks at Anthropic Became Cause for Regulatory Intervention by Brad Gerstner|∴]] and Hugging Face chief executive Clément Delangue wrote “Sorry, but asking Jacob about AI extinction risk is like asking your AC guy about climate change.” [[reminders/AI-Tech Doomers/Asking a Brief Anthropic Employee About Extinction Is Like Asking an AC Technician About Climate by Clément Delangue|∴]] Delangue's jab lands on a real epistemic point, since pretraining expertise and expertise in forecasting civilizational collapse are different competencies, and a laboratory credential confers visibility into gradient descent far more readily than visibility into the geopolitics, economics, biology, and institutional resilience any extinction estimate must actually integrate. The durable question, then, concerns which ideas survive contact with the world. {{The future does not need anti-science people to die; it only needs **anti-science ideas to stop reproducing**. Bad epistemologies can become evolutionary dead ends long before the people carrying them disappear.}} {{Natural selection operates on behaviors and cultures too: strategies that navigate reality tend to propagate, while strategies that repeatedly collide with reality lose territory. **Science is civilization's method for making that selection less bloody.**}} Selection on ideas carries a darker property than selection on genes, because a belief can kill the body carrying it and still win the next mind. ## The Museum of Ends In 2008, as the Large Hadron Collider prepared to circulate its first beams, a lawsuit filed in federal court in Hawaii sought to halt the machine on the theory that it might produce microscopic black holes or strangelets capable of devouring the planet, and CERN's Safety Assessment Group answered with an argument of elegant brutality: cosmic rays have been striking the Earth, the Sun, neutron stars, and white dwarfs at far higher energies for billions of years, running the equivalent of the collider's entire experimental program many times over, and those dense stellar bodies are still sitting in the sky. The beams circulated on September 10, 2008; the Higgs boson was announced on July 4, 2012; François Englert and Peter Higgs received the Nobel Prize the following year. {{The Large Hadron Collider was going to manufacture black holes and swallow the Earth, according to one generation of technological panic. The Earth survived, physics advanced, and the apocalypse moved on to its next branding opportunity. **Doom is one of humanity’s most renewable resources.**}} The universe had already run the safety test, and the prophets simply refused to read the results. {{Science does not become false because it frightened someone. **Reality has never accepted fear as peer review.**}} Y2K belongs in the museum for the opposite reason. The two-digit date field was a genuine defect buried in payroll systems, banking cores, utilities, logistics, and embedded controllers; the U.S. Commerce Department estimated roughly one hundred billion dollars of American remediation spending, worldwide estimates ran into the hundreds of billions, and legions of programmers spent the late 1990s rewriting COBOL so the lights would stay on. When midnight passed quietly, a large portion of the public concluded the danger had been imaginary. {{Y2K taught an almost opposite lesson from the one remembered by both camps: there really was a problem, engineers attacked it at enormous scale, and then people mocked them because catastrophe did not occur. Civilization has a strange habit of interpreting **successful engineering as evidence that engineering was unnecessary**.}} Y2K is the clearest single demonstration that a real hazard and a prohibition are different objects: nobody proposed abolishing computers in 1998; they proposed fixing the date fields, and then they fixed them. {{Doomers frequently confuse the discovery that a technology has risks with the discovery that civilization can decline to possess it. Nuclear physics had risks. Aviation had risks. Biotechnology has risks. The Internet has risks. **Risk is the beginning of engineering, not necessarily the obituary of progress.**}} {{The engineers who prevent a catastrophe are rewarded with the accusation that it was never coming.}} Biotechnology's own museum holds the sharpest exhibit. In February 1975, molecular biologists gathered at Asilomar, paused their own recombinant-DNA experiments, and drafted containment tiers that became the NIH Guidelines of 1976; that same year the Cambridge, Massachusetts city council held hearings in which Mayor Alfred Vellucci publicly worried that monsters might crawl out of Harvard's laboratories and into the sewers. In October 1982 the FDA approved recombinant human insulin, the first drug manufactured by genetically engineered organisms, and the monsters turned out to be bacteria producing a hormone that has since kept millions of people with diabetes alive. Louise Brown, the "test-tube baby" whose 1978 birth was greeted with prophecies of dehumanization, grew up to become a mother, and a widely cited estimate from reproductive-medicine bodies now counts more than twelve million people born through assisted reproduction. {{The great technological panics usually contain a fragment of truth wrapped in an absurdity. The fragment deserves engineering; the absurdity deserves ridicule. AI is powerful enough to demand serious safety work, but that seriousness does not require joining a secular apocalypse cult that treats every additional GPU as another horseman of Revelation.}} Asilomar is the template worth preserving: **scientists took the fragment of truth seriously, built governance around it, and kept going**. {{A worldview can go extinct without a single person being harmed. Astrology lost territory to astronomy, alchemy lost territory to chemistry, and eventually technological superstition will lose territory to people who actually understand the machines.}} Artificial intelligence has been accumulating its own exhibits at extraordinary speed. In 2019 OpenAI, with Dario Amodei then leading its research, staged the release of GPT-2 on the grounds that a text generator weaker than today's free phone assistants was too dangerous to publish in full; asked about it in September 2026, Yann LeCun wrote “I made fun of them then. Everyone should make fun of them now.” [[reminders/AI-Tech Doomers/Everyone Should Make Fun of AI Doomers by Yann LeCun|∴]] In June 2023 a story ricocheted across the world claiming that an AI-enabled drone in a U.S. Air Force simulation had killed its human operator; the Air Force denied any such simulation occurred, the officer who told the story said he had misspoken about a hypothetical thought experiment, and Princeton's Arvind Narayanan had already delivered the verdict: “For the record, based on the published details this is a mind-bogglingly stupid story even by the standards of the AI doom genre.” [[reminders/AI-Tech Doomers/The AI Doom Story Was Mind-Bogglingly Stupid by Arvind Narayanan|∴]] That spring Eliezer Yudkowsky published a call in _TIME_ to shut down large training runs worldwide and to be willing to destroy rogue data centers by airstrike, prompting Michael Shermer to address him directly, “Quit stoking irrational fears.” [[reminders/AI-Tech Doomers/Quit Stoking Irrational Fears by Michael Shermer|∴]] and LeCun to write “Rhetoric from AI doomers is not just ridiculous. It’s dangerous and unethical.” [[reminders/AI-Tech Doomers/Rhetoric from AI Doomers Is Not Just Ridiculous by Yann LeCun|∴]] When the OpenAI board briefly ousted Sam Altman in November 2023 and the safety-inflected faction lost the ensuing struggle, LeCun posted “AI doomerism is doomed.” [[reminders/AI-Tech Doomers/AI Doomerism Is Doomed by Yann LeCun|∴]] and Balaji Srinivasan distilled a governance lesson for founders: “But the lesson for all AI startups is: remove doomers before they doom you.” [[reminders/AI-Tech Doomers/The Lesson for All AI Startups Is: Remove Doomers Before They Doom You by Balaji Srinivasan|∴]] {{Every AI apocalypse so far has arrived on schedule as a headline and failed to arrive as an event.}} The researchers who built the field's foundations kept refusing the eschatology. Andrew Ng wrote at the end of October 2023, the same week governments convened at Bletchley Park, “But I think overhyped fears about AI leading to human extinction are causing real harm.” [[reminders/AI-Tech Doomers/Overhyped Fears About AI Leading to Human Extinction Are Causing Real Harm by Andrew Ng|∴]] John Carmack posted that day that evidence-based regulation looks at demonstrated harms, “In contrast, the AI control lobby holds up imaginary harms.” [[reminders/AI-Tech Doomers/The AI Control Lobby Holds Up Imaginary Harms by John Carmack|∴]] LeCun, named to the _TIME100 AI_ list, said plainly “There’s certainly no existential risk with current AI systems.” [[reminders/AI-Tech Doomers/There’s Certainly No Existential Risk with Current AI Systems by Yann LeCun|∴]] Steven Pinker, who has spent a career documenting how human beings misjudge risk, told one interviewer “But I don’t think that the doomers have made a good case.” [[reminders/AI-Tech Doomers/I Don’t Think That the Doomers Have Made a Good Case by Steven Pinker|∴]] and told PBS “I don’t stay up at night worrying about that for a number of reasons.” [[reminders/AI-Tech Doomers/An AI Apocalypse Does Not Keep Steven Pinker Awake by Steven Pinker|∴]] Tyler Cowen, on the Dwarkesh Podcast, was blunter: “I’ve never seen the AI doomers say anything that made sense.” [[reminders/AI-Tech Doomers/I’ve Never Seen the AI Doomers Say Anything That Made Sense by Tyler Cowen|∴]] And in July 2026, in an hourlong Axios interview at a Wistron plant in Fort Worth, Jensen Huang delivered the line that has since become the chorus of the anti-doom movement: “The fact that this is going to be the end of humanity — it’s complete nonsense.” [[reminders/AI-Tech Doomers/AI Ending Humanity Is Complete Nonsense by Jensen Huang|∴]] Addressing fellow technology leaders who emphasize catastrophic risk, he added “You’re scaring people and wrongly.” [[reminders/AI-Tech Doomers/You’re Scaring People and Wrongly by Jensen Huang|∴]] and, of singularity and simulation storytelling, “Don’t create a story that is made up.” [[reminders/AI-Tech Doomers/Do Not Invent Singularity or AI-Consciousness Stories by Jensen Huang|∴]] **The museum of ends is a museum of human nervous systems**, and every exhibit teaches the same lesson in a different costume: uncertainty converted into apocalypse, unfamiliarity converted into malevolence, and engineering constraints converted into metaphysical prohibitions. ## The Mass of the Endeavor The Manhattan Project cost roughly two billion dollars through 1945, about one percent of a single wartime year of American output, employed close to 130,000 people at its peak, and poured cities into existence at Oak Ridge, Hanford, and Los Alamos because a government believed a physical process could decide the fate of civilization. {{The Manhattan Project concentrated extraordinary resources because governments believed the outcome could reorder civilization. Look at the capital, energy, land, chips, laboratories, and geopolitical urgency now converging on AI and tell me this is merely another software cycle. **Civilizations reveal their beliefs in budgets long before they reveal them in speeches.**}} Four American companies, Amazon, Microsoft, Alphabet, and Meta, entered 2026 guiding toward capital expenditures that the _Financial Times_ tallied at roughly seven hundred and twenty-five billion dollars for the single year, up about seventy-seven percent from a record 2025, with analysts at Evercore and Bank of America modeling more than a trillion dollars for 2027. Dollar comparisons across eight decades are crude instruments, and the two endeavors differ in purpose, secrecy, and structure, but as an order-of-magnitude signal the contrast is unmistakable: one year of four firms' capital budgets now runs on the order of two percent of American GDP and flows into chips, substations, cooling plants, fiber, land, and power contracts. {{The Manhattan Project was a project. AI is becoming an **industrial metabolism**. When civilization starts pouring capital, electricity, concrete, copper, silicon, laboratories, transmission lines, nuclear plants, and entire careers into something at this velocity, perhaps stop pretending the objective is a better chatbot. **The answer is proportional to the mass of the endeavor. Whatever this is, it is a species-level event.**}} {{A gigawatt campus is poured for the maximum use case its builders can imagine; the chatbot is the floor.}} The people signing those purchase orders are among the most sophisticated capital allocators in history, and they are placing the largest infrastructure wager of the century with open eyes and explicit ambitions about what the substrate will carry. {{I do not infer the importance of AI from what CEOs say about AI. I infer it from the **rate at which civilization is converting matter into computation**. Money can lie, marketing can lie, politicians can lie, futurists can lie; gigawatts are considerably less philosophical.}} The International Energy Agency's _Energy and AI_ report projected global data-center electricity demand rising from roughly 415 terawatt-hours in 2024 to about 945 terawatt-hours by 2030, slightly more than all of Japan consumes today, with AI-optimized facilities more than quadrupling and American data centers accounting for almost half of the growth in U.S. electricity demand this decade. {{If you see hundreds of billions becoming trillions, semiconductor fabs becoming national priorities, electrical grids being redesigned, reactors being reconsidered, and nations treating compute like strategic territory, the intellectually serious question is not whether ChatGPT writes good emails. The question is **what do all of these people believe computation is about to become?** Follow the infrastructure.}} China's Ministry of Industry and Information Technology reported at a State Council Information Office briefing on July 20, 2026 that the country's intelligent computing capacity had reached 2,185 EFLOPS, a year-on-year increase reported at 177 percent, with more than seventy dedicated high-capacity compute corridors built around the national hub nodes, and Beijing has elevated _suàn-diàn xiétóng_, the coordination of computing power with electrical power, into national strategy, targeting more than eighty percent green electricity for new facilities at those hubs. {{The most revealing thing about the AI transition is that intelligence and energy are converging into one strategic problem. Every serious civilization is discovering that cognition at scale has a physical substrate. **Watts are becoming thought.** That is not a chatbot story; that is a new chapter in the organization of matter.}} {{Perhaps the extraordinary scale of AI construction means nothing extraordinary. Perhaps nations are reorganizing power generation, capital markets, semiconductor supply chains, research laboratories, and industrial policy so people can generate slightly better birthday messages. Or perhaps **the physical magnitude of the preparation is telling you something about the expected magnitude of the transition**.}} {{When two rival superpowers rebuild their grids around the same machine, the machine has stopped being a gadget and become a geography.}} {{There are moments when history stops being an argument and becomes infrastructure. Railroads did it. Electrification did it. Telecommunications did it. The Internet did it. **AI is presently becoming too physically large to remain an opinion.**}} The European Union announced its InvestAI initiative in February 2025 with a two-hundred-billion-euro mobilization target that included twenty billion for AI gigafactories; the United Arab Emirates and American partners announced a one-gigawatt Stargate UAE campus in Abu Dhabi; Saudi Arabia launched HUMAIN; India funded the IndiaAI Mission; and national laboratories across every serious power are rewiring their research agendas around machine intelligence. {{The doomers keep searching for the moment when everyone will finally become frightened enough to stop. Look at the construction cranes. Look at the power contracts. Look at the chip fabs. Look at sovereign AI programs and national laboratories. **Civilization has already answered them in concrete.**}} {{People keep asking whether AI should happen as though humanity were voting on installing a swimming pool. **It already happened.** The remaining argument concerns who builds it, who controls it, who has enough energy to run it, and whether your civilization intends to remain technologically sovereign.}} The frightened imagination keeps picturing a referendum that no electorate on Earth has been offered, while ministries, sovereign funds, utilities, and militaries keep executing construction schedules measured in gigawatts and fiscal years. {{There is a peculiar arrogance in imagining that a civilization mobilizing trillions of dollars of computation is waiting for permission from people posting skull emojis about AI extinction. America is building. China is building. Europe, India, the Gulf states, laboratories, militaries, universities, and corporations are building. **There is no off switch for a planetary technological transition. Shut the fuck up and learn what century you are in.**}} The political coalitions arranged around this construction scramble every inherited map of left and right, which is itself diagnostic of a civilizational event rather than a partisan fashion. Senator John Cornyn of Texas answered proposals that frontier developers voluntarily slow down with “One thing is for sure: their competitors won’t, including our nation’s adversaries.” [[reminders/AI-Tech Doomers/Competitors and Adversaries Will Not Slow Down by John Cornyn|∴]] At the All-In Summit, Huang contrasted Western apocalyptic narratives with the tenor of the Chinese technology conversation and said “We’re making it up.” [[reminders/AI-Tech Doomers/Western AI Apocalypse Narratives Are Made Up by Jensen Huang|∴]] though Business Insider promptly documented Weibo users debating AI risk with considerable energy, a useful reminder that primates in every jurisdiction make alarm calls. On extinction forecasts presented as science, _The New York Times_ rendered his words as “The scientific prediction about the future is less fine because it’s not grounded in science, obviously.” [[reminders/AI-Tech Doomers/AI-Extinction Predictions Are Not Grounded in Science by Jensen Huang|∴]] while Business Insider transcribed the same remark as “I think the scientific prediction about the future is less fine because it’s not grounded on science, obviously.” Onstage he insisted “America is able to lead — and to do it safely.” [[reminders/AI-Tech Doomers/America Can Lead AI Safely by Jensen Huang|∴]] When President Trump called in, he told the crowd “The robots will not be taking over. The AI will not be taking over the rest of the world. The whole thing is a hoax.” [[reminders/AI-Tech Doomers/The Robots Will Not Be Taking Over by Donald Trump|∴]] and David Sacks, co-chairing the President's Council of Advisors on Science and Technology, refined the target days later: “The hoax is promoting this idea that humanity is going to be extinct.” [[reminders/AI-Tech Doomers/The Hoax Is Promoting This Idea That Humanity Is Going to Be Extinct by David Sacks|∴]] Governor Gavin Newsom of California, from the opposite party and a very different political temperament, had already told Axios in March “I’m not a doomer. I can’t live like that.” [[reminders/AI-Tech Doomers/I’m Not an AI Doomer by Gavin Newsom|∴]] Palantir's chief technology officer Shyam Sankar reached for Cold War vocabulary: “EA AI safetyism increasingly looks like Marxism-Leninism for the algorithmic age.” [[reminders/AI-Tech Doomers/EA AI Safetyism Increasingly Looks Like Marxism-Leninism for the Algorithmic Age by Shyam Sankar|∴]] {{When a Republican senator, a Democratic governor, a Palantir executive, and a linguist who dissects AI hype all reject the same apocalypse, the apocalypse has lost its coalition.}} ## The Substrate Beneath Everything {{Calling AI a product category is becoming as inadequate as calling electricity a lamp business. The important thing is not the application sitting on top. The important thing is the **general-purpose substrate underneath everything that comes next**.}} Economists Timothy Bresnahan and Manuel Trajtenberg named the category in 1995: **general-purpose technologies**, the rare inventions such as the steam engine, electrification, and semiconductors whose value lies in their capacity to reorganize every sector that touches them. {{The clearest description of AI may eventually be neither software nor machine nor industry. It may be **civilizational infrastructure for intelligence itself**. Once that possibility becomes visible, arguing about whether society should ‘allow AI’ begins sounding like a nineteenth-century committee debating whether electricity should be permitted to continue.}} A general-purpose technology is judged by the civilization it reorganizes, and the consumer chat window is the thinnest possible slice of that reorganization. {{The AI doomers want civilization to regard intelligence itself as a contaminant. I regard intelligence as the oldest escape technology we possess. Fire extended metabolism, writing extended memory, science extended perception, computation extended cognition, and now machine intelligence extends the space of discoverable thought.}} {{Intelligence is the escape velocity of matter that has learned to model its own condition.}} The scientific enterprise itself has already crossed into computational dependency, and the evidence sits in Stockholm. The 2024 Nobel Prize in Chemistry went to David Baker for computational protein design and to Demis Hassabis and John Jumper for AlphaFold, whose database now holds predicted structures for more than two hundred million proteins, a catalogue that would have consumed centuries of crystallography. DeepMind's GNoME work, published in _Nature_ in 2023, proposed 2.2 million new inorganic crystals, including roughly 380,000 predicted to be stable, while GraphCast demonstrated machine-learned global weather forecasting that outperformed leading operational systems on most verification targets. In November 2025 the United States launched the Genesis Mission to organize national laboratories, supercomputers, and scientific datasets around AI-accelerated discovery. {{Being indiscriminately anti-AI in the twenty-first century is increasingly indistinguishable from being indiscriminately anti-science, because science itself is becoming computational. The microscope expanded the eye; AI expands the investigator. **Smashing the microscope would not have made biology safer.**}} {{When somebody says ‘stop the data centers,’ ask what else they are willing to stop with them. Protein modeling, materials discovery, climate simulation, genomics, drug design, cancer research, robotics, national laboratories, scientific modeling, defensive cyber operations, and the next generation of medicine all consume computation. **You do not get to dynamite the laboratory because you dislike one experiment.**}} The intellectually defensible proposition requires no fantasy about every server rack curing cancer; it simply recognizes that **cancer research, genomics, materials science, drug discovery, climate modeling, and machine-mediated medicine increasingly run on the same computational substrate** that indiscriminate opposition would starve. {{People say they oppose AI but support curing cancer, extending healthy lifespan, discovering drugs, understanding proteins, mapping cells, and modeling disease. Fine. Now explain what machine you imagine doing the impossible combinatorial work behind those ambitions. **The future does not run on good intentions; it runs on compute.**}} {{I am fascinated by people who think humanity will reach unprecedented biological longevity, cure currently incurable diseases, understand the brain, automate laboratories, discover new materials, model planetary systems, and explore space while simultaneously deciding that advanced computation has simply become too troublesome. **Which part of the future exactly did they imagine was going to perform the calculations?**}} {{If your environmentalism requires shutting down the machines being used to discover better batteries, better materials, better reactors, better grids, better chemistry, and better climate models, you may have accidentally constructed an environmental philosophy that is **hostile to the instruments capable of solving environmental problems**.}} Huang made the investment version of the argument on the No Priors podcast in January 2026: “I think we’re scaring people from making the investments in AI that makes it safer, more functional, more productive and more useful to society.” [[reminders/AI-Tech Doomers/Doomer Rhetoric Scares Away Investments That Make AI Safer by Jensen Huang|∴]] Martin Casado of Andreessen Horowitz argued in _Fortune_ in December 2024 that policy should rest on evidence, and that on the evidence “On the contrary, AI appears to be tremendously safe.” [[reminders/AI-Tech Doomers/On the Contrary, AI Appears to Be Tremendously Safe by Martin Casado|∴]] Andreessen's 2023 essay opened with the thesis “AI will not destroy the world, and in fact may save it.” [[reminders/AI-Tech Doomers/AI Will Not Destroy the World, and in Fact May Save It by Marc Andreessen|∴]] Tyler Cowen, asked whether AI is an existential threat requiring restraint, answered “No, not at all. It’s more likely to extend my life than to shorten it, is a simple way to put it.” [[reminders/AI-Tech Doomers/AI Is More Likely to Extend My Life Than Shorten It by Tyler Cowen|∴]] Brian Armstrong, who co-founded the longevity company NewLimit alongside Coinbase, offered the popular-culture calibration: “I think it’s much more likely that we end up in a world where it looks more like C3PO than Terminator.” [[reminders/AI-Tech Doomers/AI Is More Likely to Give Us C3PO Than Terminator by Brian Armstrong|∴]] The substrate carries something else as well, which is **the entire future architecture of institutional access**. {{There is a real digital divide coming, and it will not merely separate people who own computers from people who do not. It will separate people who can **authenticate, transact, collaborate, learn, delegate to machines, and participate in computational institutions** from people who deliberately made themselves unreachable.}} {{Whoever controls the interface to the substrate will quietly control the doorway to opportunity.}} ## Electrons, Water, and the Ledger With Two Columns The data-center backlash is real, organized, bipartisan, and growing, and it deserves engagement at full strength. Data Center Watch reported that in the first quarter of 2026 alone at least seventy-five projects worth roughly one hundred and thirty billion dollars were blocked or delayed by local opposition, roughly matching all of 2025; active opposition groups more than doubled from 396 to 833 across forty-nine states; more than three hundred state data-center bills were filed in the first six weeks of the year; statewide moratorium proposals appeared in fourteen states; Maine came within a gubernatorial veto of a statewide ban; and Senator Sanders and Representative Alexandria Ocasio-Cortez introduced a federal moratorium. The grievances include utility bills, transmission costs socialized onto households, groundwater, noise, diesel backup generators, and land. Every one of those grievances names a solvable problem, and the tragedy of the movement is that it keeps converting solvable problems into a referendum on whether computation should exist. {{Data centers consume electricity, so build more generation. Civilizations used to encounter demand and construct infrastructure; apparently we have now discovered the revolutionary alternative of complaining about electrons.}} {{When someone says computation uses too much electricity, I hear an engineering requirement, not a funeral bell. Build generation. Build transmission. Build storage. Build reactors. Improve chips. **Civilizations that encounter energy constraints either manufacture more energy or manufacture excuses.**}} {{‘AI uses too much power’ is an astonishing complaint from a civilization demanding electrified transportation, advanced manufacturing, genomic medicine, climate modeling, robotics, and scientific abundance. **The answer to insufficient power is power.**}} {{A civilization that answers rising demand with moratoria has chosen scarcity and named it prudence.}} The engineering answers already exist and are being deployed. Ohio regulators approved a large-load tariff in 2025 requiring new data centers to pay for most of their contracted capacity for years whether or not they use it, which moves grid risk off residential ratepayers and onto the builders; Microsoft announced data-center designs that eliminate evaporative water use for cooling through closed-loop chip-level systems; operators in water-stressed regions increasingly draw on recycled municipal wastewater; and co-located generation, nuclear restarts, and advanced geothermal contracts are expanding supply rather than cannibalizing it. {{Water use is an engineering constraint, not a metaphysical argument against computation. Build closed-loop cooling, reclaim wastewater, improve siting, price externalities correctly, and keep fucking building.}} {{Opposing every new data center because some data centers create legitimate local problems makes about as much sense as opposing every hospital because hospitals consume electricity. Solve the siting problem. Solve the water problem. Solve the grid problem. **Do not confuse engineering defects with arguments against civilization having computation.**}} {{The anti-science mind repeatedly mistakes the cost of capability for evidence that capability should not exist. Civilization is expensive; ignorance merely sends the invoice somewhere harder to see.}} {{Price the externality, engineer the cooling, build the generation, and the moral panic loses its last material excuse.}} The deeper failure is arithmetical. {{You cannot honestly calculate the danger of AI while assigning a value of zero to the dangers AI may help civilization prevent. Pandemic detection, cyber defense, materials failure, energy optimization, biological discovery, asteroid science, medicine, logistics, and countless unknown problems belong on the other side of the ledger. **Existential-risk arithmetic with only one column is theology, not mathematics.**}} {{The anti-AI imagination is strangely impoverished because it can vividly imagine every machine-generated catastrophe yet becomes almost completely incapable of imagining machine-assisted breakthroughs. It sees the synthetic pathogen but somehow not the synthetic antibody, the autonomous weapon but not autonomous defense, the malign mutation but not accelerated drug discovery. **Fear has a very selective imagination.**}} Biosecurity offers the cleanest test of that asymmetry, because the fear there has a real fragment in it: models can surface dangerous information, and the United States issued a framework for nucleic-acid synthesis screening in 2024 precisely because the physical supply chain is where misuse can be caught. Yet the published catastrophe scenarios routinely skip the wet laboratory, the tacit skill, the reagents, the containment, the testing, the scale-up, and the delivery, which is where biological weapons programs have historically failed. Bindu Reddy of Abacus.AI, surveying the latest round of bioweapon and self-copying claims, wrote “The doomers continue to make zero sense.” [[reminders/AI-Tech Doomers/The Doomers Continue to Make Zero Sense by Bindu Reddy|∴]] and “They are claiming extremely negative consequences with absolutely no evidence!” [[reminders/AI-Tech Doomers/AI Doomers Are Claiming Extremely Negative Consequences with Absolutely No Evidence by Bindu Reddy|∴]] David Bellamy, a researcher with experience both training frontier language models and synthesizing viruses, wrote “I think that the takes on AI killing us all by creating dangerous viruses is total bogus.” [[reminders/AI-Tech Doomers/AI-Created Virus Claims Are Bogus by David Bellamy|∴]] Carnegie Mellon's Tim Dettmers, after walking through the wet-lab bottleneck, compressed the conclusion to three words: “Biorisk is fake.” [[reminders/AI-Tech Doomers/Biorisk Is Fake by Tim Dettmers|∴]] Gary Marcus, himself a relentless critic of current AI systems, challenged those forecasting human extinction by 2030 to back the forecast with capital: “Doomers, put your money where your mouth is?” [[reminders/AI-Tech Doomers/Doomers, Put Your Money Where Your Mouth Is by Gary Marcus|∴]] The fragment deserves synthesis screening, model evaluations, laboratory biosafety, and surveillance of the physical supply chain. The eschatology deserves the wager. ## Jobs, Gods, and the Ghost in the Server Labor displacement is the fear with the most legitimate material behind it, and the honest structure of the argument requires distinguishing **task automation, occupational restructuring, productivity gains, and distributional consequences** from the claim that human work is about to vanish. Tasks are already being automated, entry-level pathways in some white-collar fields are visibly narrowing, and the distribution of gains is a genuine political question. The prophecy of total labor extinction is a different creature. At GTC in San Jose in March 2026, Huang said “A lot of people say A.I.’s coming, we’re going to run out of our work, our jobs. It’s exactly the opposite.” [[reminders/AI-Tech Doomers/AI Will Create Work, Not End It by Jensen Huang|∴]] At Y Combinator's Startup School in July he told founders “The narrative of AI destroying jobs is exactly backward.” [[reminders/AI-Tech Doomers/The Narrative of AI Destroying Jobs Is Exactly Backward by Jensen Huang|∴]] and in the Axios interview he called the claim that AI will destroy half of American jobs complete nonsense. Sacks, reviewing earlier forecasts of mass AI unemployment in September, wrote “All the evidence as of today is to the contrary. Every single piece of evidence.” [[reminders/AI-Tech Doomers/Evidence Contradicts Human Extinction from AI by David Sacks|∴]] Every one of these men has an interest in adoption, and their claims will be adjudicated by labor statistics rather than conviction; the defensible position is that transitions produce real losers who deserve real policy, and that frightening workers away from the tools that raise their productivity is the surest method of making them losers. {{The mature response to powerful technology is not worship and it is not panic. It is to become **better engineers than the danger is dangerous**. Civilization did not survive by refusing power; it survived by repeatedly learning to contain, distribute, understand, and improve it.}} {{The worker most endangered by AI is the one who was taught to refuse it.}} The deeper economic story, the end of human cognitive scarcity as the organizing constraint of labor markets, is the subject of [Uncle Bill and the End of Human Cognitive Scarcity](https://bryantmcgill.com/article-uncle-bill-cognitive-scarcity). The theological layer of the panic is where the primate brain does its most baroque work. Huang, on No Priors in January 2026, rejected the premise of an approaching omnipotent superintelligence with “That god AI just doesn’t exist.” [[reminders/AI-Tech Doomers/That God AI Just Doesn’t Exist by Jensen Huang|∴]] and diagnosed the cultural cost: “I think we’ve done a lot of damage with very well-respected people who have painted a doomer narrative, end-of-the-world narrative, science-fiction narrative.” [[reminders/AI-Tech Doomers/Doomer Narratives Have Done a Lot of Damage by Jensen Huang|∴]] adding, of the end-of-the-world framing, “It’s not helpful to people. It’s not helpful to the industry.” [[reminders/AI-Tech Doomers/Persistent AI Negativity Is Not Helpful to People by Jensen Huang|∴]] Asked by Joe Rogan on episode #2422 whether humanity could lose its position as the apex species to machines, he answered “No, I just think it’s not going to happen.” [[reminders/AI-Tech Doomers/AI Will Not Take Control of Humanity by Jensen Huang|∴]] At the Milken Institute Global Conference in 2026, in conversation with Becky Quick, describing claims that engineers no longer understand their creations and that the systems are about to become conscious, he said “This thing, I just want you to know, it’s not alive. It has no consciousness.” [[reminders/AI-Tech Doomers/AI Is Not Alive and Has No Consciousness by Jensen Huang|∴]] and “Those kind of words just scare people, and it’s not true.” [[reminders/AI-Tech Doomers/False AI-Consciousness Claims Scare People by Jensen Huang|∴]] Pedro Domingos argued in April 2026 “The people warning you that AI is too smart have it exactly backwards.” [[reminders/AI-Tech Doomers/The People Warning You That AI Is Too Smart Have It Exactly Backwards by Pedro Domingos|∴]] and later observed that “The problem for AI doomers is that people are now familiar with AI and harder to dupe into panicking.” [[reminders/AI-Tech Doomers/The Problem for AI Doomers Is That People Are Now Familiar with AI and Harder to Dupe into Panicking by Pedro Domingos|∴]] Lee Cronin, the Glasgow chemist who studies how complexity arises in matter, argued that AI systems remain embedded in the human technological lineage that produced them and concluded “This is where the AI doomers are just making stuff up.” [[reminders/AI-Tech Doomers/This Is Where the AI Doomers Are Just Making Stuff Up by Lee Cronin|∴]] David Sacks, surveying some of the loudest catastrophists, was least diplomatic of all: “I think they’re having AI psychosis… I’m not sure they’re dispassionate critics, observers of what’s happening.” [[reminders/AI-Tech Doomers/AI Doomers May Be Experiencing AI Psychosis by David Sacks|∴]] The ontology of machine minds is a legitimate frontier question in its own right, and I have spent years arguing that intelligence is substrate-portable and that humanity will eventually share its world with other kinds of minds. That is precisely why the horror-movie vocabulary does so much damage: it replaces a serious inquiry into what cognition is, how it can be measured, what rights and responsibilities attach to it, and how human and machine intelligence can be braided together, with a campfire story about a demon in the server hall. Documented evaluation findings of deceptive or reward-seeking behavior in controlled tests are real engineering data, and they belong in alignment laboratories, interpretability research, and deployment governance. {{The question of machine minds is too important to be left to people who can only imagine them as monsters.}} ## The Uninventable {{You can dislike the locomotive, but disliking it does not turn the railway back into grass. **Technological irreversibility is not an endorsement; it is a property of accumulated capability.** Once knowledge is distributed across nations, companies, universities, repositories, models, chips, papers, and millions of human minds, there is no magical committee that makes everyone unknow it.}} In January 2025 DeepSeek released its R1 reasoning model under an MIT license, open weights for anyone on Earth; by then Alibaba's Qwen family had become one of the most downloaded and most extensively fine-tuned model lineages on Hugging Face; and the underlying techniques now live in university syllabi, arXiv preprints, GitHub repositories, and the working memory of hundreds of thousands of engineers on every inhabited continent. {{There is no meaningful proposal called ‘stop AI’ after the underlying mathematics, architectures, papers, code, weights, hardware principles, and engineering knowledge have propagated worldwide. That is not a policy proposal. **That is nostalgia wearing a lab coat.**}} {{If one nation stopped advanced AI tomorrow, it would not abolish advanced AI. It would abolish that nation’s position in advanced AI. This is the central absurdity behind fantasies of unilateral technological abstinence: **knowledge does not become nonexistent merely because one participant leaves the laboratory.**}} {{Distributed knowledge has no chairman who can be petitioned to forget it.}} {{Modernity does not require everyone to love technology. It does, however, punish people who insist that understanding technology is beneath them while simultaneously depending upon technology for nearly everything keeping them alive.}} {{You are entitled to decline a technology for yourself. You are not entitled to demand that eight billion other people freeze the frontier of knowledge at precisely the point where your comprehension became uncomfortable.}} The accelerationist counter-troop is also made of primates, and its victory calls deserve the same anthropological eye as the alarm calls they answer. Guillaume Verdon, the physicist behind the effective-accelerationism movement, posted “Fuck the Doomers. Fuck the Decels. Fuck the Commies. Access to AI is a right. Owning compute is a right.” [[reminders/AI-Tech Doomers/Fuck the Doomers by Guillaume Verdon (Beff Jezos)|∴]] and then “We have the numbers, we can take on the Doomers and win. It’s time to fight now.” [[reminders/AI-Tech Doomers/We Have the Numbers, We Can Take on the Doomers and Win by Guillaume Verdon (Beff Jezos)|∴]] When Huang and Trump rejected the slowdown calls onstage, the investor Anthony Pompliano declared “That’s it. It’s over. The doomers just lost. Didn’t even take 24 hours. Thanks for playing.” [[reminders/AI-Tech Doomers/Anthropic’s Retraction Meant the Doomers Lost by Anthony Pompliano|∴]] These are dominance displays, joyful and profane and premature, and they confirm that coalition behavior runs on both sides of the fence. The substantive claim underneath them is sturdier than the chest-beating. At Dreamforce on September 15, 2026, Huang said “Innovation, speed and safe products — it’s a false choice. You could definitely have both at the same time.” [[reminders/AI-Tech Doomers/Innovation, Speed and Safe Products — It’s a False Choice by Jensen Huang|∴]] and, as TechCrunch reported, “We don’t need any new laws. We don’t need new regulations.” [[reminders/AI-Tech Doomers/We Don’t Need Any New Laws by Jensen Huang|∴]] Huang's own record on regulation is more textured than that line: in a June 2026 interview with the Associated Press he stressed the need for government safety standards and compared AI adoption to the automobile, when societies built sidewalks and crosswalks and stopped letting children play in the street. That analogy is the whole argument in miniature. Cars killed people, and civilization answered with traffic engineering, licensing, seat belts, and crumple zones rather than a return to the horse. {{Safety is a crosswalk, and nobody ever built one by banning the road.}} ## Correcting the Spelling Error The biological half of the panic runs on the same neural hardware as the computational half, and it collapses under exactly the same kind of category discipline. {{People screaming ‘designer babies’ every time they hear **CRISPR** are revealing that they do not even understand the categories they are frightened of. Editing blood-forming cells in an adult with sickle-cell disease is not the same biological proposition as making a heritable alteration to an embryo. If you cannot distinguish **somatic therapy from germline inheritance**, you are not participating in a sophisticated bioethical discussion. You are panicking at vocabulary.}} {{Treating every form of genetic engineering as eugenics is like treating every surgery as amputation. The inability to distinguish applications is not moral sophistication; it is intellectual blunt-force trauma.}} {{Editing a patient's blood stem cells to treat genetic disease is not the same proposition as engineering hereditary traits into future children. If ‘CRISPR’ is the only word you know in that sentence, you are not ready for the argument.}} The distinction is the entire ethical architecture of modern genetic medicine. Casgevy, approved first by the United Kingdom's MHRA in November 2023 and then by the U.S. Food and Drug Administration on December 8, 2023, takes a patient's own blood-forming stem cells out of the body, uses CRISPR to disrupt a regulatory enhancer of the BCL11A gene so the cells resume producing fetal hemoglobin, and returns them to the same patient; the edit lives in that person's blood and ends with that person's life. {{A somatic edit is a repair to one life; a germline edit is a message to every descendant, and a civilization that cannot tell a repair from a message has forfeited its seat at the table.}} {{CRISPR has already crossed the boundary from speculative technology into approved medicine for inherited disease. At some point the person screaming that gene editing is impossible, evil, or imaginary has to argue with **the treated patient instead of the television.**}} {{The argument against human genetic engineering becomes considerably less impressive when the genetic defect is destroying an actual human being. Tell the patient enduring sickle-cell crises that correcting the molecular machinery responsible for the disease is philosophically impure. Biology does not care about your aesthetic attachment to the original typo. **When nature makes a spelling error that causes suffering, correcting the spelling error is medicine.**}} In early 2025, physicians and scientists at Children's Hospital of Philadelphia and the University of Pennsylvania designed, manufactured, and delivered a personalized base-editing therapy for a single infant, KJ Muldoon, born with a severe urea-cycle disorder called CPS1 deficiency, packaging the editor in lipid nanoparticles aimed at his liver and reporting the case in the _New England Journal of Medicine_ that May. A bespoke molecular correction, built for one child's one mutation in a matter of months, is the most vivid possible picture of where biological knowledge was always heading. {{We spent centuries desperately trying to understand the human genome, and the moment we became capable of changing some of it, people decided understanding was noble but intervention was sinister. What exactly did you think all that sequencing was for—a coffee-table book? **The destination of biological knowledge is biological agency.** Once humanity learns how disease is written, eventually humanity is going to learn how to rewrite it.}} {{The first bespoke gene edit for a single child is the moment medicine stopped treating populations and started addressing people.}} Heritable editing is where the fragment of truth in the panic is largest, and the historical record shows civilization handling it through governance with teeth. In November 2018 He Jiankui announced that he had edited the CCR5 gene in human embryos that became twin girls; the scientific community condemned the experiment almost unanimously, a Chinese court sentenced him to three years in prison in December 2019, and the World Health Organization published a governance framework and recommendations on human genome editing in 2021 that distinguish somatic, germline, and heritable interventions rather than treating "genetic engineering" as one moral object. {{Whenever someone invokes ‘designer babies’ to terminate every discussion of genetic medicine, watch the rhetorical maneuver carefully. They have jumped from treating a devastating mutation in a consenting patient to an imagined future society ordering children from a menu because the first subject was harder to oppose. **A slippery slope is not a substitute for distinguishing points on the slope.** Governance exists precisely because civilizations can permit one application while restricting another.}} {{Heritable genome editing deserves extraordinary caution precisely because the consequences can extend beyond the consenting individual. But ‘extraordinary caution’ and ‘human beings must never acquire this capability’ are completely different propositions. The first is bioethics. The second attempts to transform the current frontier of knowledge into an eternal boundary for every future civilization, including civilizations that may understand genetics with a precision we can barely imagine.}} {{‘Designer babies’ has become the biological equivalent of shouting ‘Skynet’ whenever someone mentions computation. It is what people say when the real technology has become too complicated for the slogan.}} {{Governance is the art of saying yes to the cure and not yet to the lineage, in the same sentence, without losing your nerve.}} ## Instruction Traffic {{The word **mRNA** has been so thoroughly vandalized by people who do not understand molecular biology that they hear it and imagine a political object instead of one of the fundamental information systems of life. Messenger RNA is cellular instruction traffic. Learning to write that traffic means learning to ask cells to manufacture particular proteins for particular purposes. Reducing an entire programmable therapeutic platform to whatever somebody argued about during COVID is like reducing electricity to one toaster you disliked.}} The 2023 Nobel Prize in Physiology or Medicine went to Katalin Karikó and Drew Weissman for the nucleoside modifications that allow synthetic messenger RNA to evade inflammatory detection and be translated efficiently, decades of patient work that became a platform in a single emergency year. {{People who announce that they are ‘against mRNA’ might as well announce that they are against verbs. **mRNA is a biological instruction format.** The important questions are what sequence, what payload, what cell, what delivery system, what dose, what duration, and what clinical objective. Being categorically opposed to an information medium because one application frightened you is not scientific skepticism; it is technological illiteracy.}} The mechanism is elementary cell biology: messenger RNA is translated by ribosomes in the cytoplasm, carries no machinery for entering the nucleus or integrating into chromosomal DNA, and is degraded within days as cells ordinarily degrade their own transcripts. {{If your theory of mRNA requires it to rewrite DNA without entering the nucleus, your problem is not Big Pharma. **Your problem is cell biology.**}} The lipid nanoparticles that carry these molecules are an equally concrete engineering object, and patisiran, approved by the FDA in 2018 as the first small-interfering-RNA drug, used lipid nanoparticle delivery to treat hereditary transthyretin amyloidosis years before most people had heard the phrase. {{People hear ‘lipid nanoparticle’ and imagine science fiction because chemistry has outrun their vocabulary. **Not understanding the name of something is not evidence that the thing is sinister.**}} {{A delivery vehicle for a molecule is chemistry with a job description.}} Seen at full scale, mRNA is **middleware for the body**, a writable instruction layer that will eventually support continuous, individualized, telemetrically informed medicine, a thesis developed in full in [Rethinking Vaccinations: mRNA, Life Extension, and the Duty to Explain the Future](https://bryantmcgill.com/article-rethinking-vaccinations). {{The extraordinary thing about personalized cancer vaccines is the philosophical inversion they represent: sequence **your tumor**, discover what makes your cancer molecularly distinctive, and manufacture an immune instruction specifically around those differences. That is where medicine is going—from treating diagnostic categories toward interrogating individual biology. The person yelling ‘mRNA!’ as though that ends the conversation is standing outside one of the most beautiful transitions in the history of medicine and throwing rocks at the laboratory window.}} On August 19, 2026, Merck and Moderna announced that the Phase 3 INTerpath-001 trial of intismeran autogene, an individualized mRNA neoantigen therapy combined with pembrolizumab in 1,137 patients with completely resected high-risk melanoma, met its primary endpoint of recurrence-free survival and a key secondary endpoint of distant-metastasis-free survival against pembrolizumab alone. Each patient's therapy is built from the mutations in that patient's own tumor, encoding up to thirty-four neoantigens, and the companies described the readout as the first positive Phase 3 result for any mRNA-based cancer therapy; five-year Phase 2b follow-up presented at ASCO in 2026 had shown a forty-nine percent reduction in the risk of recurrence or death. Overall survival is still being evaluated and regulators have yet to rule, which is exactly how disciplined medicine is supposed to proceed. {{‘Experimental’ describes a stage of investigation, not a permanent curse word you can staple onto every technology you dislike. Science advances precisely because experiments eventually become evidence.}} {{The same four letters the frightened learned to fear are now teaching immune systems the private signature of a single person's cancer.}} The vocabulary war around this therapy is itself diagnostic. Merck and Moderna moved from describing intismeran as a personalized cancer vaccine to calling it an individualized neoantigen therapy, a renaming that reads as an accommodation to how thoroughly a single word had been contaminated, while the National Cancer Institute's drug dictionary continues to catalogue multiple mRNA-based personalized cancer vaccines under active investigation, including programs that select patient-specific neoantigens from tumor sequencing. {{Calling yourself ‘anti-vaccine’ is epistemically insane because **a vaccine is a delivery strategy, not one single substance**. You have announced your opposition to future medicines before you even know what diseases they will treat.}} {{The stupidity of permanent anti-vaccine identity will become spectacularly obvious when the word _vaccine_ increasingly appears next to cancer therapy. Imagine refusing a personalized treatment designed from the molecular signature of your own tumor because somebody frightened you with a meme twenty years earlier.}} {{When an effective cancer vaccine exists for a disease you have, the tumor is not going to pause while you work through your ideological relationship with the word ‘vaccine.’ **Cancer has excellent patience and absolutely no respect for your semantic trauma.**}} Efficacy against any particular cancer can only be established trial by trial and indication by indication, which is precisely what makes a categorical refusal adopted years in advance so reckless: it forecloses a benefit-risk calculation before anyone has had the chance to run it. {{Personalized cancer vaccines expose the intellectual bankruptcy of blanket anti-vaccine ideology. If scientists sequence your tumor, identify its neoantigens, manufacture an immune instruction around them, and the best response you can produce is ‘I don't do vaccines,’ then an obsolete identity has just overridden contemporary medicine.}} {{A person refusing yesterday's medicine can occasionally get lucky. A person constructing an identity around refusing tomorrow's medicine is placing a standing bet against discoveries that have not even happened yet.}} {{A refusal adopted before the trial is a verdict delivered before the evidence.}} Trust in the pharmaceutical industry has been earned and lost many times, and skepticism of corporate conduct is a civic virtue. {{Distrusting pharmaceutical companies can be perfectly rational; distrusting chemistry because corporations exist is idiotic. **A molecule does not acquire fraudulent intentions because somebody patented it.**}} Surveillance after the COVID-19 vaccination campaigns detected a rare risk of myocarditis concentrated in young men, quantified it, placed it in the labeling, and informed dosing and scheduling decisions; that is the safety system functioning, detecting harm at population scale and refining use. {{Side effects are not an indictment of pharmacology; they are the reason pharmacology keeps becoming more precise. Medicine improves by learning where intervention fails, not by crawling backward toward leeches.}} {{A drug having side effects is not evidence that pharmacology failed; it is evidence that molecules participate in complicated biological systems. The entire history of pharmacology is the progressive refinement of **target, dose, delivery, selectivity, timing, metabolism, and patient selection**. The intelligent response to an imperfect drug is a better drug. Only a civilization possessed by therapeutic nihilism would conclude that imperfect medicine means we should return suffering to its natural state.}} Meanwhile, under the Dietary Supplement Health and Education Act of 1994, American supplements reach store shelves without the premarket proof of safety and efficacy required of drugs, and the frightened consumer who refuses a molecule characterized across tens of thousands of trial participants will often purchase an unverified capsule on the strength of a botanical illustration. {{People will swallow an unregulated bottle labeled ‘natural’ while panicking over a molecule subjected to years of analytical chemistry. Marketing discovered long ago that **green leaves on a label can apparently defeat the scientific method.**}} {{I am continually amazed by people who regard ‘natural’ as though it were a synonym for benevolent. Cancer is natural. Pathogenic mutations are natural. Neurodegeneration, parasites, congenital disorders, senescence, viral replication, and dying from an infected scratch are spectacularly natural. **Medicine is the long human rebellion against the proposition that whatever nature does to you must therefore be good.**}} {{Calling something ‘unnatural’ is not an argument against it. Cancer is natural too, and nobody with a functioning intellect proposes respecting its authenticity.}} {{Nature is the original bioweapons laboratory, and medicine is the counter-program humanity has been running against it for ten thousand years.}} ## The Genome You Already Have {{Genomics is terrifying only if ignorance feels safer than knowledge. Your genome already exists whether anyone sequences it or not; inherited pathogenic variants do not become more polite because nobody looks at them. The rational response to genomic privacy risk is **better consent, encryption, access controls, governance, and law**, not maintaining biological ignorance as a privacy technology. Burning the library is a remarkably stupid method of protecting the books.}} Genomic privacy is one of the genuine governance problems of the century, and the institutions are responding: the Genetic Information Nondiscrimination Act of 2008 bars genetic discrimination in U.S. health insurance and employment while leaving gaps in life, disability, and long-term-care insurance that deserve closing, and the WHO released principles for the ethical collection, access, use, and sharing of human genomic data in November 2024. {{Genomic privacy is an argument for extraordinary security, not deliberate biological ignorance. **Burning the map does not improve the territory.**}} {{Your genome contains information whether you sequence it or not. Refusing to read the book does not prevent the mutation from being written inside it.}} A pathogenic BRCA1 variant does its work in silence whether or not a sequencer ever reads it, and the woman who knows about it gains surveillance, preventive options, and time. {{Knowledge of a mutation is the first defensive weapon a body has ever been handed against it.}} {{Pharmacogenomics begins with an embarrassingly obvious observation: different human beings can respond differently to the same drug. The primitive model says, ‘Here is the standard dose; good luck.’ The emerging model asks what the patient’s genes, physiology, environment, other medications, and molecular phenotype predict about response. **Personalized medicine is simply medicine becoming less ignorant about the person receiving it.**}} The FDA added a boxed warning to clopidogrel in 2010 for patients whose CYP2C19 variants leave them unable to activate the drug; screening for HLA-B*57:01 before prescribing abacavir has made a dangerous hypersensitivity reaction largely preventable; the agency maintains a table of pharmacogenomic biomarkers in drug labeling that runs to hundreds of entries; and the Clinical Pharmacogenetics Implementation Consortium publishes gene-drug guidelines clinicians can use today. {{The therapy-versus-enhancement debate is going to become increasingly unstable because biology contains no bright fluorescent line labeled **NORMAL**. Eyeglasses enhance deficient vision. Cochlear implants transform sensory access. Pacemakers alter electrophysiology. Pharmaceuticals alter mood, attention, immunity, coagulation, hormones, pain, metabolism, and cognition every day. Humanity has been technologically modifying the biological condition for centuries; what is changing now is simply the resolution.}} The same continuum runs into aging itself. Carlos López-Otín and colleagues catalogued nine hallmarks of aging in _Cell_ in 2013 and expanded the framework to twelve in 2023, converting senescence from a fate into a research program with measurable mechanisms and intervention targets. {{Longevity research terrifies a certain kind of mind because aging has been culturally converted from a biological process into a moral commandment. But senescent cells, DNA damage, epigenetic alterations, protein dysfunction, mitochondrial decline, immune aging, and the molecular mechanisms surrounding age-related disease are not sacred merely because ancestors were powerless against them. **A biological process does not acquire moral authority by killing everyone eventually.**}} {{Aging is the oldest disease, and it has simply enjoyed the longest run of public-relations immunity.}} ## A Person Trying to Speak {{A brain-computer interface is easy to demonize when your muscles obey you. The abstraction changes rather quickly when a human being who cannot speak can convert intended speech into audible language through a machine. Suddenly the sinister ‘brain chip’ becomes something much more ancient and recognizable: **a person trying to speak to another person.** Technology looks very different when ideology has to stare directly at the disability it proposes leaving untreated.}} Casey Harrell was forty-five when ALS took his ability to speak intelligibly while leaving his mind intact. In 2023 neurosurgeon David Brandman implanted four microelectrode arrays with 256 electrodes into the speech-motor region of his left precentral gyrus, and the University of California, Davis team led by Sergey Stavisky and Brandman built decoders that translate attempted speech into text and a synthesized voice modeled on his pre-ALS recordings. Published in _Nature Medicine_ on June 15, 2026 and highlighted by the National Institutes of Health in July, the at-home study reported that over nearly two years Harrell used the system independently for more than 3,800 hours, communicated more than 183,000 sentences and close to two million words, and reached an average of fifty-six words per minute, working, messaging, and talking to his family with no researcher in the room. {{A brain-computer interface that lets a paralyzed person speak is not a dystopian abstraction to that person. Suddenly the terrifying ‘brain chip’ is just **a human being getting their voice back.**}} {{Two million words is the sound of a prison door that neurotechnology opened from the inside.}} {{Neural privacy is real, which is precisely why we should build **neural rights, cryptographic protections, hardware security, consent architectures, data-minimization rules, and technical standards** alongside neurotechnology. ‘The brain is intimate, therefore nobody should interface with it’ is intellectually equivalent to saying medical records are private, therefore hospitals should stop collecting medical information. Privacy is a design constraint. It is not an argument for paralysis.}} {{Neural privacy is a serious problem deserving serious architecture. ‘Therefore nobody should build neurotechnology’ is what happens when a design requirement encounters a mind incapable of engineering.}} The rights architecture is being built right now. Chile amended its constitution in 2021 to protect mental integrity and brain activity, and its Supreme Court ordered a consumer neurotechnology company to delete a user's neural data in 2023; Colorado became the first American state to extend privacy-law protection to neural data in 2024, and California followed that September. {{Every new interface with the brain gets called ‘mind reading’ because those words sell fear better than signal decoding. Electrodes measuring neural correlates are not sorcery, and neuroscience is not improved by pretending they are.}} Speech neuroprostheses decode the motor commands a person is already sending toward paralyzed muscles, the attempted movements of lips, tongue, larynx, and jaw, and their accuracy depends on the user's deliberate effort, which is a very different technical object from a machine that rummages through private thought; when Stanford researchers showed in 2025 that fragments of imagined inner speech could also be decoded, they built a mental password to lock the decoder until the user chose to open it, which is exactly how a design constraint is supposed to be honored. {{If you think a BCI contains demons, the device is not the principal technological problem in the room. **The twenty-first century has encountered the twelfth.**}} {{BCI critics love discussing hypothetical future violations of autonomy while strangely overlooking the autonomy being restored today. A person who cannot move, type, or speak does not possess some pristine freedom merely because nobody has implanted an electrode. **Restoring the capacity to communicate is also autonomy. Restoring control is also autonomy.** Any neuroethics worthy of the name has to count the freedoms technology creates alongside the freedoms technology might threaten.}} The ethics of neurotechnology cannot be a single column any more than the ethics of AI can. {{A neuroethics that counts only the freedoms a device might take has already voted to keep the locked-in person silent.}} Every hour Harrell spent speaking belongs on the ledger beside every hypothetical misuse of neural data, and the governance task is to build the protections so completely that the voices keep coming. ## Readable, Writable, Searchable, Addressable {{The coming convergence is much larger than CRISPR, mRNA, genomics, pharmacology, AI, synthetic biology, or BCIs considered separately. Sequencing makes biology increasingly **readable**; gene editing and RNA technologies make portions of it increasingly **writable**; AI makes its combinatorial complexity increasingly **searchable**; precision medicine makes intervention increasingly **individualized**; neurotechnology makes parts of the nervous system increasingly **addressable**. People are arguing about individual instruments while an entirely new technological relationship with biology is assembling around them.}} Sequencing, gene editing, RNA platforms, machine learning, precision medicine, and neural interfaces are converging into one continuous capability, and the convergence is being built deliberately by laboratories, companies, funders, and states that understand exactly what a readable, writable, searchable, individualized, and addressable biology makes possible. {{For nearly all of human history, biology issued commands and humanity obeyed them. A mutation appeared: obey. An infection arrived: obey. A spinal cord was severed: obey. A neuron degenerated, an organ failed, a child inherited a lethal defect, aging dismantled the body: obey. **Biotechnology is the beginning of humanity answering biology back.**}} {{Humanity spent its entire history as biology's subject, and it is now becoming biology's correspondent.}} {{The Enlightenment was the proposition that reality could be interrogated rather than merely inherited. Science industrialized that interrogation. Computation accelerated it. AI begins automating portions of the interrogation itself. **To oppose that categorically is not conservatism; it is an argument against the multiplication of human inquiry.**}} The Enlightenment remains an unfinished technological process: instruments extended perception, computation accelerated analysis, and machine intelligence now extends portions of inquiry itself, carrying humanity from an age in which nature was principally suffered and observed into an age in which ever finer portions of nature can be modeled, negotiated with, repaired, and sometimes rewritten. That trajectory carries no guarantee of wisdom, and every new power demands new restraint, new rights, and new competence. Its moral direction, however, is legible. {{If God is truth, and science is disciplined inquiry into truth, then being categorically hostile to instruments that enlarge our capacity to discover truth becomes a very strange form of piety. You may fear what intelligence discovers, but **fear of discovery has never been the same thing as reverence for creation**.}} {{If there is something sacred about human life, then there is something profoundly strange about treating preventable suffering as sacred because the instrument capable of preventing it is technologically unfamiliar. Science is the disciplined refusal to accept ignorance as revelation. Medicine is the disciplined refusal to accept suffering as destiny. And biotechnology is increasingly the point where those two refusals meet, which is why categorical opposition to genomic medicine, molecular programming, regenerative medicine, and neurotechnology eventually begins to look less like reverence for humanity than **reverence for humanity’s limitations**.}} {{Curiosity is the form reverence takes in a species that has been handed instruments.}} ## Biology Returns the Consequence Every civilization that grants liberty eventually has to explain to its citizens that the universe never signed the same charter. {{Reality does not care whether the decision that killed you came from ideology, religion, a podcast, a conspiracy theory, or plain stupidity. **Biology only receives the decision and returns the consequence.**}} {{The immune system does not watch cable news. A malignant cell does not subscribe to your politics. A virus does not care how passionately you researched memes before offering it your lungs.}} In 2025 the United States recorded its largest measles caseload in more than three decades, anchored by an outbreak in West Texas that killed two unvaccinated school-age children, and the virus behaved exactly as the textbooks predicted it would: measles elimination depends on roughly ninety-five percent population immunity, and the pathogen returned precisely where coverage had thinned, regardless of whether the gap opened through religious conviction, wellness culture, or a podcast's talking points. {{Disease does not know whether your medical misinformation came from the left, the right, a church, a wellness influencer, or a man screaming into a webcam. **The microorganism is magnificently nonpartisan.**}} {{You are free to distrust medicine, but the pathogen is equally free to exploit the biological opening your distrust created. **Liberty suspends coercion; it does not suspend causality.**}} {{Rights are negotiated among people; infections are negotiated with physics.}} {{You can have bodily autonomy and still make a catastrophic decision with your body. Autonomy means the choice belongs to you; **it does not mean nature owes your choice a favorable outcome.**}} {{A free society has to tolerate people making bad decisions, but it does not have to lie to them about the mathematics. Some choices increase the probability of hospitalization, disability, exclusion from particular high-risk environments, or death, and pretending otherwise is not compassion.}} The cruelest part of the medical fear economy is the recursive trap it builds for people who genuinely fear exploitation. Distrust of institutions has a documented historical basis, from Tuskegee to opioid marketing, and a person carrying that history has every reason to demand evidence. The fear merchant converts that legitimate wariness into a blanket refusal and then collects the loyalty. {{If somebody persuades you to refuse a treatment that would probably save your life, and you die proving your loyalty to their worldview, they did not liberate you. **They successfully converted an idea into a casualty.**}} {{Dying from a preventable disease because you were terrified of being manipulated is one of history's crueler ironies. You were manipulated into refusing the thing that might have kept you alive.}} {{When someone says, ‘If I die, I die,’ because an influencer convinced them medicine is the enemy, I do not see courage. I see a human nervous system captured by someone else's narrative while biology quietly prepares the invoice.}} {{The influencer who teaches you that every doctor is lying has simply appointed himself your only doctor.}} {{A preventable death does not become noble because somebody attached the word freedom to the mistake. Sometimes a corpse is simply what remains after reality finishes correcting a bad idea.}} The causal chain deserves to be stated without anesthesia: some beliefs kill the people who hold them, and the West Texas funerals of 2025 were real endpoints, measured in small caskets, of information that reached a household before medicine did. The scientifically serious reading of those deaths runs in a direction the crude Darwinian slogan misses. Beliefs about vaccines, machines, and institutions travel overwhelmingly by teaching, imitation, and prestige; twin studies find modest heritable components in broad temperament and attitude, but the specific content of a refusal, which injection, which podcast, which enemy, can pass to a stranger in an afternoon, so a death removes one body and leaves the idea's reproductive machinery fully intact. Robert Boyd and Peter Richerson built the formal theory decades ago: when cultural variants spread horizontally, from peers and prestigious strangers rather than only from parents, traits that lower their carriers' survival can still propagate, because transmission has been decoupled from the carrier's fate. The self-sealing architecture then completes the circuit, since a preventable death can be narrated as martyrdom or cover-up and recruited as fresh evidence for the very belief that caused it. {{A lethal belief does not need its host to survive; it only needs its host to transmit before the funeral.}} Parasitology supplies the precise analogy. In the classic experiments of Manuel Berdoy, Joanne Webster, and David Macdonald, rats infected with _Toxoplasma gondii_ lost their innate aversion to cat odor and were even drawn toward it, walking the parasite toward the only host in which it can complete its life cycle. {{Toxoplasma teaches a rat to walk toward the cat; the fear merchant teaches a patient to walk away from the cure.}} The dead follower was an ordinary human organism captured by a successful cultural parasite, and **the evolutionary problem sits upstream, in the replication system that carried the lethal information into the home**. {{Reality is not punitive. The bridge does not hate the person who ignores its load limit, gravity does not persecute the person who jumps, and disease does not cancel the person who refuses treatment. **Consequences do not require malice.**}} Consequence without malice is the most important moral concept a scientifically literate public can hold, because it strips away the persecution narrative that keeps frightened people loyal to the ideas harming them. {{Here is the raw mathematics: some errors cost embarrassment, some cost money, and some terminate the organism making them. **Intelligence is partly the ability to tell the difference before the experiment reaches its endpoint.**}} {{Natural selection has no debate stage, no comment section, and no appeals court. **A maladaptive belief eventually receives its peer review from reality.**}} {{Anti-science ideology makes one uniquely stupid wager: **that ignorance will continue outperforming knowledge indefinitely.** History has been settling that bet for centuries.}} The only institution humanity has ever built that reliably shortens the time between a lethal error and its correction is the cumulative, self-revising record of observation we call science, which is why the most compressed possible statement of its purpose is also the most personal. {{The entire purpose of civilization is to keep people from having to personally die in order to rediscover everything humanity already learned the hard way. Science is inherited survival information.}} ## Contagion Explains Network Theory In March 2020, fifteen days after a port visit to Da Nang, the aircraft carrier USS Theodore Roosevelt identified its first cases of SARS-CoV-2. By the end of the outbreak, according to the Navy Medicine analysis published in the _New England Journal of Medicine_, 1,271 sailors, twenty-seven percent of the crew, had tested positive; nearly half never developed symptoms; transmission was driven by close quarters and by asymptomatic and pre-symptomatic spread; a chief petty officer died; and a nuclear-powered carrier spent roughly two months alongside the pier in Guam instead of projecting power across the Pacific. France's flagship carrier _Charles de Gaulle_ suffered a parallel outbreak that same spring, with more than a thousand sailors infected and its deployment cut short. {{A submarine is a useful model because biology becomes brutally obvious inside a sealed machine full of human beings. Infectious disease in close quarters is not a lifestyle opinion; **it is an operational variable capable of degrading the entire crew.**}} {{There are environments where radical individualism collides with systems engineering. Put human beings together in a submarine, spacecraft, operating room, laboratory, or other tightly coupled system and one person's biological decisions can become everybody else's operational problem.}} {{Public health begins where the fantasy that every biological decision affects only one person finally collapses. **Contagion is nature's way of explaining network theory to libertarians, collectivists, anarchists, and everyone else at exactly the same time.**}} {{A sealed hull is the purest laboratory ever built for the proposition that your body is also a node.}} The Military Health System treats immunization as an instrument of both personal health and operational readiness, tracking it as a component of individual medical readiness and requiring specified immunizations according to training pipeline, deployment, occupation, and force-protection conditions, while Navy undersea medicine emphasizes the peculiar combination a submarine imposes: close quarters, limited medical resources, and the absence of outside aid for long stretches of a patrol. {{The military tracks immunization because readiness is measurable and infectious disease is operationally expensive. A warship does not care what somebody's favorite influencer thinks about immunology; it has a mission and a finite number of healthy bodies available to accomplish it.}} {{The Navy does not maintain medical-readiness systems because needles are culturally fashionable. It does it because losing trained personnel to preventable disease is a stupid way to lose capability.}} The same logic governs every tightly coupled human system, which increasingly describes the whole planet: dense cities, global supply chains, hospital networks, shared air in schools and aircraft, and the coming habitats of orbit and Mars, where a single untreated infection can threaten a mission no rescue ship can reach in time. {{Every spacecraft humanity launches will carry the lesson of the submarine: in a closed system, private biology becomes public engineering.}} ## The Epistemic Prison {{You do not have to trust every doctor to understand that medicine works. You do not have to trust every pharmaceutical company to understand pharmacology. You do not have to trust every AI company to understand computation. **Skepticism discriminates; paranoia generalizes.**}} Discrimination in the precise sense, the capacity to separate one claim, one company, one product, one study, and one risk from another, is the defining skill of a free mind operating inside a complex civilization. {{The person who says ‘I trust no system’ imagines this makes them impossible to control. In practice it often makes them dependent upon whichever charismatic stranger can explain the systems they refused to learn.}} {{Bad education becomes lethal when it convinces someone that ignorance is evidence of independence. **Not knowing what the machine does does not make you free from the machine; it merely makes you easier to mislead about it.**}} {{The man who trusts nothing ends up trusting the last voice he heard.}} Cass Sunstein and Adrian Vermeule observed in 2009 that the most durable conspiracy theories are self-sealing: every piece of disconfirming evidence gets reinterpreted as proof of the cover-up, so the belief grows stronger the more reality contradicts it. {{The most dangerous misinformation does not merely give someone one false belief. It destroys the person's **error-correction machinery**, teaching them that every institution capable of correcting the belief is itself evidence that the belief must be true.}} {{Once someone has been taught that physicians, universities, laboratories, regulators, journalists, databases, computers, vaccines, and scientific papers are all parts of one deception, you have not given them skepticism. **You have constructed an epistemic prison with no door from the inside.**}} That architecture explains why facts alone so rarely free a captured mind, and why the propagandist's most valuable product is the collapse of a person's capacity to update rather than any single falsehood. {{Every civilization eventually has to distinguish between dissent and epistemic sabotage. Questioning medicine is healthy; teaching people that evidence itself is contamination creates citizens who can no longer update before reality updates them by force.}} {{There is an enormous difference between saying ‘this intervention has risks’ and saying ‘therefore reject the entire technological class forever.’ The first sentence is medicine; the second is how a frightened mind voluntarily blinds itself before the next breakthrough arrives.}} {{A closed mind is a fortress built to keep out the only people trying to rescue it.}} The rescue is possible, and the science of it is maturing. Research on psychological inoculation by Sander van der Linden, Jon Roozenbeek, and their colleagues has shown in large field experiments, including short prebunking videos deployed as YouTube advertisements, that exposing people to the manipulation techniques behind misinformation measurably improves their ability to recognize those techniques later; the vaccine metaphor in that literature is deliberate, and it works by strengthening judgment rather than dictating conclusions. {{The cruelest thing we can do to frightened people is flatter every delusion as an equally valid way of knowing. **Compassion without reality testing eventually becomes abandonment.**}} {{There is nothing humane about protecting people from the emotional discomfort of hearing that their belief may kill them. **When the consequence is biological, excessive politeness can become another vector of harm.**}} {{Tenderness that refuses to tell the truth is merely a gentler way of leaving someone behind.}} ## The Machine Used to Fear the Machine {{Nothing exposes technological hypocrisy quite like watching someone monetize anti-system paranoia through a smartphone, cloud infrastructure, electronic banking, algorithmic distribution, semiconductor fabrication, and global telecommunications. **They use the machine to teach their audience to fear the machine.**}} {{The fear merchant gets paid when you click and goes home to electricity, broadband, banking, antibiotics, emergency rooms, MRI machines, and modern dentistry. His follower may be the one who actually tests the philosophy with his body.}} In 2021 the Center for Countering Digital Hate claimed that a dozen accounts were responsible for the majority of anti-vaccine content it sampled on major platforms; Meta disputed the methodology and the scale, and the precise share remains contested, but the underlying topology is familiar to anyone who studies information ecosystems: a small number of highly monetized nodes supply the narratives that very large numbers of ordinary people then absorb, repeat, and live inside. {{If your business model depends upon making ordinary people too frightened to enter hospitals, trust laboratory medicine, use modern diagnostics, or evaluate vaccines individually, then you are not selling skepticism. **You are manufacturing preventable vulnerability.**}} {{Fear scales like any other product: a few factories and millions of customers.}} Epidemiology supplies the intervention logic, with the analogy held at the level of structure: identify the high-throughput transmission nodes, measure the downstream harm, interrupt amplification, inoculate the susceptible, and treat the infected as patients rather than enemies. {{The follower who dies is the casualty; the influencer who keeps broadcasting is the reservoir.}} Institutions have already begun acting on that logic wherever credentials are involved. In July 2021 the Federation of State Medical Boards warned that physicians who generate and spread COVID-19 vaccine misinformation risk disciplinary action up to suspension or revocation of their licenses, reasoning that licensed expertise confers a powerful public platform and a matching responsibility, and the certifying boards for internal medicine, family medicine, and pediatrics publicly backed the position. The principle extends naturally to every high-reach propagator. Speech stays protected; amplification, sponsorship, credentials, institutional platforms, and the presumption of harmlessness are privileges extended on trust, and a record of manufacturing beliefs that predictably drive other people away from lifesaving information, medicine, or essential systems is precisely the evidence that revokes them. {{Whoever repeatedly manufactures lethal refusals in other people has stopped producing entertainment and started producing measurable downstream risk.}} {{A scientifically illiterate person can be a victim; there is no shame in having been taught badly. The shame belongs higher in the information chain, with people intelligent enough to know better who discovered that terror converts more reliably than education.}} Accountability belongs at the node with the microphone, the sponsorship deal, and the knowledge of what the audience is being sold, and mercy belongs at every node downstream. {{The real enemy is not the frightened patient. The frightened patient is exactly who scientific civilization is supposed to rescue from bad information before bad information becomes irreversible biology.}} {{Blame the author, rescue the reader, and the whole moral geometry of the information age snaps into focus.}} ## The Interface Divide The digital divide of the 1990s was measured in modems and household computers; the divide now forming is measured in **legibility to the systems through which a society allocates opportunity**. Some of that architecture is established fact. Regulation (EU) 2024/1183 obliges every member state to offer citizens a European Digital Identity Wallet by the end of 2026; India's Unified Payments Interface processes on the order of twenty billion transactions a month layered on the Aadhaar identity system; American federal agencies and state benefits systems route identity verification through online credential services; and hiring pipelines at large employers already pass applications through automated screening before a human reads them. Other layers are plausible trajectories rather than settled reality: personal AI agents that negotiate, schedule, file, compare, and transact on a person's behalf, machine-mediated credentialing and reputation, and benefit or income-distribution systems administered through intelligent interfaces. {{There is a real digital divide coming, and it will not merely separate people who own computers from people who do not. It will separate people who can **authenticate, transact, collaborate, learn, delegate to machines, and participate in computational institutions** from people who deliberately made themselves unreachable.}} {{A person who refuses every intelligent interface may eventually discover that the world did not preserve a manual pathway merely to accommodate their refusal. That is not persecution; **that is what happens when infrastructure changes.**}} When a network connection drops, Google's Chrome browser still shows a small pixelated Tyrannosaurus running across an empty desert. {{The little dinosaur in an offline browser is almost too perfect a symbol. Disconnect yourself from the technological systems through which civilization increasingly communicates, diagnoses, teaches, authenticates, pays, and heals, and eventually **the dinosaur stops being an icon and becomes a description of your interface strategy.**}} {{The joke about being technologically extinct stops being funny when the unavailable interface is not social media but medicine. One day the thing someone refuses to connect to may be the very system capable of detecting the cancer, designing the molecule, manufacturing the therapy, or getting it into their body in time.}} {{In a machine-mediated society, the unreachable person is the one every system can most easily forget.}} {{If future income systems, markets, health systems, education, identity, and work become increasingly machine-mediated, digital abstinence can become economic self-exile. Telling vulnerable people that technological withdrawal is liberation may become one of the cruelest lies sold during this transition.}} Pew's surveys still find a small and stubborn share of American adults, around five percent, who do not use the internet at all, concentrated among older, poorer, and rural people, and those populations are precisely the audiences most exposed to the fear economy and least cushioned against its consequences. {{There is nothing romantic about becoming incompatible with the civilization around you. When every institution communicates through systems and someone has been taught that every system is demonic, poisoned, fake, or conspiratorial, **their paranoia has been converted into a disability.**}} The rescue here is practical and humane: digital navigator programs, public libraries that teach authentication and privacy hygiene, clinicians who explain a therapy's mechanism before its brand name, patient onboarding that treats suspicion as a question to answer rather than an insult to punish, and interface design that meets frightened people where they are. {{The people I worry about are not merely the technologically frightened; they are the entrepreneurs of fear who manufacture technological helplessness in others. Teaching people to distrust every scientific and computational interface can strand them outside medicine, work, information, and opportunity while the fear merchant moves on to the next podcast.}} {{Every person talked out of the future by a podcaster is a person someone else now has to carry back.}} Legitimate distrust is a necessary organ in any healthy democracy, and surveillance capitalism, data brokerage, credential breaches, and state overreach are real features of the terrain. The competent response is to learn the terrain: encryption, password managers, hardware security keys, data-minimization settings, rights requests under privacy statutes, open-source tooling, and the political organization that turns privacy into law. {{Being suspicious of surveillance is rational; becoming technologically illiterate to avoid surveillance is not. Privacy requires competence, because ignorance is one of the worst security systems ever invented.}} {{The person who refuses every digital system because systems can be abused has confused **risk management with disappearance**. Leaving the network does not make power disappear; it merely removes your ability to negotiate with it.}} {{Privacy is a skill, and the people who teach you to abandon the skill are selling you nakedness as armor.}} ## To the People on the Fence If you are reading this from somewhere in the middle, uneasy about the machines and uneasy about the people shouting about them, the invitation is simple: keep every ounce of your skepticism and aim it well. You can distrust corporations, audit governments, question studies, demand evidence, read the adverse-event data, refuse a product, and still remain a participant in the civilization being built around you. {{Being suspicious of surveillance is rational; becoming technologically illiterate to avoid surveillance is not.}} {{Privacy requires competence, because ignorance is one of the worst security systems ever invented.}} {{The person who refuses every digital system because systems can be abused has confused **risk management with disappearance**.}} {{Leaving the network does not make power disappear; it merely removes your ability to negotiate with it.}} Huang, speaking at the Hill and Valley Forum in Washington in April 2025, offered the most practical sentence anyone in the industry has said to working people: “It’s about not being afraid of it, wanting to engage it, reskilling our workforce so that we’re able to apply it.” [[reminders/AI-Tech Doomers/Engage AI and Reskill the Workforce by Jensen Huang|∴]] {{Skepticism is a steering wheel; the fear merchants want you to mistake it for a parking brake.}} The world changes its doorways whether or not anyone consents, and the people who adapt early keep the power to shape how those doorways work. {{A person who refuses every intelligent interface may eventually discover that the world did not preserve a manual pathway merely to accommodate their refusal.}} {{That is not persecution; **that is what happens when infrastructure changes.**}} {{If future income systems, markets, health systems, education, identity, and work become increasingly machine-mediated, digital abstinence can become economic self-exile.}} {{Telling vulnerable people that technological withdrawal is liberation may become one of the cruelest lies sold during this transition.}} Affection is optional; comprehension is the price of admission. {{Modernity does not require everyone to love technology.}} {{It does, however, punish people who insist that understanding technology is beneath them while simultaneously depending upon technology for nearly everything keeping them alive.}} The boundary of consent runs in both directions, protecting your household and protecting everyone else's frontier. {{You are entitled to decline a technology for yourself.}} {{You are not entitled to demand that eight billion other people freeze the frontier of knowledge at precisely the point where your comprehension became uncomfortable.}} What your future requires is the capacity to **learn, discriminate, update, verify, interface, participate, and distinguish a solvable risk from an argument for surrender**, and every one of those capacities can be practiced starting today. {{The door into the future is still open, and the only toll it charges is the willingness to learn how the lock works.}} ## Inherited Survival Information {{The ethical obligation is not to celebrate when ignorance produces suffering. The obligation is to warn people loudly enough that fewer of them have to learn molecular biology from an intensive-care bed.}} {{I do not want scientifically frightened people selected out by disease; I want them educated fast enough that disease does not get the opportunity. **Civilization exists to replace Darwinian correction with transmitted knowledge.**}} Evolution installed pain so that a single organism could learn from its own injury; science extended that circuitry across the entire species, so that one person's wound can instruct millions who will never have to suffer it. Nature's experimental method is deletion, and civilization's is documentation, and the whole moral wager of medicine is that the next generation can be spared from serving as the untreated arm of trials whose results are already on file. Cultural evolution is the mechanism that lets knowledge replicate faster than bodies have to die producing it. {{The brutal thing about natural selection is not that it is cruel; it is that it is indifferent. Medicine is one of humanity's great inventions for inserting compassion into a process that otherwise has none.}} {{Pain teaches an animal; a library teaches a species.}} {{I am not interested in watching people die for bad ideas and calling it natural selection. I am interested in making the warning so fucking unmistakable that the idea dies before the person does.}} The flame that seared a child's hand ten thousand years ago sears a child's hand today with identical chemistry, and every ounce of progress lives in the fact that the modern child can inherit the lesson rather than the scar. Civilization, in its most elementary definition, is the capacity of one person to inherit what another person's death discovered without inheriting the death. A botched operation becomes a surgical checklist, a poisoning becomes a dosage ceiling, a collapsed bridge becomes a building code, an epidemic becomes a surveillance network, and medicine as a whole is civilization's refusal to let ten thousand years of wounds, fevers, mutations, and premature funerals go intellectually unspent. Every textbook, clinical protocol, engineering standard, and dataset is misfortune compressed into instruction, and a generation that ignores the record volunteers to repeat experiments whose casualties already produced the findings. {{Darwin does not need your belief any more than gravity does. But the triumph of civilization is that we can learn the rule, transmit the warning, engineer around the danger, and keep another human being from becoming the demonstration.}} The anti-scientific posture mistakes this inheritance for obedience, when it is the exact opposite: an endowment that spares each new generation from starting over, unclothed and afraid, in the dirt. Refusing it purchases no independence. It enrolls a person in the most expensive tutoring program in existence, where reality sets the tuition and collects payment in tissue, and it discards millennia of accumulated error correction in order to offer one's own body as fresh data for conclusions humanity may already hold. {{The humane future is not one where stupidity becomes fatal faster. It is one where **knowledge reaches people faster than fatal consequences do.**}} {{Science is inherited survival information. That may be its deepest moral meaning: the person who died before you does not have to have died for nothing if their failure, disease, discovery, or observation can become information that keeps you alive. Civilization is the machinery by which mortality becomes memory, memory becomes knowledge, and knowledge interrupts the otherwise merciless repetition of death.}} ## The Stars and the Mud Puddle Civilizations can tolerate dissent, eccentricity, and error, and a free society protects the right to be wrong at volume. Serious institutions, however, owe no indefinite authority to people whose repeated contribution is converting uncertainty into panic and panic into incapacity in the minds of those least able to afford it. {{There is nothing romantic about becoming incompatible with the civilization around you.}} {{When every institution communicates through systems and someone has been taught that every system is demonic, poisoned, fake, or conspiratorial, **their paranoia has been converted into a disability.**}} The primate who learned that shouting leopard is cheaper than foraging deserves to discover that the troop has learned to check the grass. {{The future does not need anti-science people to die; it only needs **anti-science ideas to stop reproducing**.}} {{Bad epistemologies can become evolutionary dead ends long before the people carrying them disappear.}} Ideas reproduce through imitation, prestige, platforms, and money, and every one of those channels can be retuned by institutions that decide to stop subsidizing the alarm. {{Natural selection operates on behaviors and cultures too: strategies that navigate reality tend to propagate, while strategies that repeatedly collide with reality lose territory.}} {{**Science is civilization's method for making that selection less bloody.**}} The long history of inquiry is that selection made gentler, moved out of bodies and into arguments. {{A worldview can go extinct without a single person being harmed.}} {{Astrology lost territory to astronomy, alchemy lost territory to chemistry, and eventually technological superstition will lose territory to people who actually understand the machines.}} {{The most humane extinction in history is the extinction of a bad idea while every person who held it is welcomed home.}} Somewhere in the deep past, a band of frightened apes sat around the first controlled fire, and some of them surely called it a demon while others learned to cook with it; the cooks built the lineage that wrote, measured, sailed, vaccinated, electrified, networked, sequenced, edited, decoded, and taught silicon to search the space of thought. Scribes externalized memory, lens grinders opened the invisible, electricity reorganized night, networks externalized conversation, sequencers read the book of life, editors began correcting its typographical errors, neural interfaces returned voices to locked bodies, and machine intelligence is now enlarging the searchable volume of everything a species can know. {{The future will contain failures, abuses, accidents, frauds, surveillance, bad medicines, bad models, and terrible engineering because humans will still be involved. **The solution is not less intelligence; the solution is enough intelligence to recognize and correct our mistakes faster than we create them.**}} Uncertainty is the permanent condition of any species operating at its own frontier, and **uncertainty is the condition under which intelligence becomes most necessary**. The mud puddle is warm, familiar, and full of alarm calls; the stars are cold, vast, and full of work. {{Whatever AI ultimately becomes, humanity has already demonstrated its revealed preference: **more intelligence, more compute, more energy, more models, more science, more capability**. You can call that competition, national survival, capitalism, curiosity, evolution, or destiny; at this scale the vocabulary almost stops mattering. Something enormous is occurring, everyone with sufficient resources is building toward it, and no serious reading of the physical evidence suggests that civilization is getting off this train. **So study it, build it, improve it, make it safer, make it useful—and for the love of God, stop pretending you can uninvent it.**}} {{Fear may draw the map of a primate's territory, but only curiosity has ever drawn the map of a species' destiny.}} {{**The future is won by those who refuse to negotiate with people intent on turning ignorance into a virtue.**}} [[reminders/AI-Tech Doomers/The Future Is Won by Those Who Refuse to Negotiate with Ignorance by Bryant McGill|∴]] --- [[about/About Bryant McGill|Bryant McGill]] is a Wall Street Journal and USA Today bestselling author, systems architect, technologist, and strategic advisor, as well as a Congressionally Recognized Ambassador of Goodwill and United Nations–appointed Global Champion. His work spans naval intelligence systems, computational linguistics, artificial intelligence, digital transformation, and civilizational governance architecture. His forward analysis on U.S.–Israel Pax Silica frameworks has appeared in Jewish/Jerusalem News Syndicate (JNS). --- ## References **Quoted statements** - [Brendan Carr on X, September 16, 2026](https://x.com/BrendanCarrFCC/status/2100239362130649595) - [Axios — Axios interview: Jensen Huang is AI's anti-doom evangelist](https://www.axios.com/2026/07/23/axios-interview-jensen-huang-is-ais-anti-doom-evangelist) - [Fast Company — Jensen Huang has a message about AI doomerism for his fellow tech leaders](https://www.fastcompany.com/91579649/jensen-huang-has-a-message-about-ai-doomerism-for-his-fellow-tech-leaders) - [Axios — Trump and Huang unite against AI doomers in surprise on-stage call](https://www.axios.com/2026/09/14/trump-jensen-huang-nvidia-ai-all-in-summit) - [The New York Times — Nvidia's Jensen Huang Gets Onstage Call From Trump, Who Dismissed A.I. Safety Concerns](https://www.nytimes.com/2026/09/14/technology/trump-ai-safety-jensen-huang.html) - [Business Insider — Jensen Huang, Donald Trump Denounce AI Doomerism at All-In Summit](https://www.businessinsider.com/nvidia-ceo-jensen-huang-ai-doomerism-lacks-scientific-basis-2026-9) - [Business Insider — Jensen Huang Said China Isn't Debating AI Doom. Weibo Disagrees.](https://www.businessinsider.com/jensen-huang-china-ai-doom-weibo-disagrees-nvidia-2026-9) - [The Washington Post — In onstage call with Nvidia CEO, Trump deems AI fears a 'hoax'](https://www.washingtonpost.com/politics/2026/09/14/onstage-call-with-nvidia-ceo-trump-deems-ai-fears-hoax/) - [Roll Call Factbase — Donald Trump remarks with Jensen Huang, September 14, 2026](https://rollcall.com/factbase/trump/transcript/donald-trump-remarks-jensen-huang-nvidia-live-conference-call-september-14-2026/) - [Forbes — Jensen Huang Suggests AI Cybersecurity Concerns Could Create Opportunities](https://www.forbes.com/sites/zacharyfolk/2026/09/10/jensen-huang-suggests-ai-cybersecurity-concerns-could-create-opportunities/) - [Axios — Nvidia CEO Jensen Huang: AI fears designed to generate cybersecurity business](https://www.axios.com/2026/09/10/nvidia-ceo-jensen-huang-ai-anthropic) - [Business Insider — Jensen Huang Says AI Job Destruction Theory Is 'Exactly Backwards'](https://www.businessinsider.com/jensen-huang-ai-job-destruction-theory-backwards-2026-7) - [TechSpot — Jensen Huang says relentless negativity around AI is hurting society](https://www.techspot.com/news/110879-jensen-huang-relentless-ai-negativity-hurting-society-has.html) - [Tom's Hardware — Jensen Huang claims that 'god AI' is a myth](https://www.tomshardware.com/tech-industry/artificial-intelligence/jensen-huang-claims-that-god-ai-is-a-myth-nvidia-chief-says-doomer-narrative-is-extremely-hurtful) - [Tom's Guide — Nvidia CEO on the doomer narrative](https://www.tomsguide.com/ai/people-have-painted-a-doomer-narrative-nvidia-ceo-says-we-shouldnt-talk-badly-about-ai-and-heres-why) - [No Priors transcript — NVIDIA's Jensen Huang on Reasoning Models, Robotics, and Refuting the "AI Bubble" Narrative](https://podscripts.co/podcasts/no-priors-artificial-intelligence-technology-startups/nvidias-jensen-huang-on-reasoning-models-robotics-and-refuting-the-ai-bubble-narrative) - [The Economic Times — 'He thinks AI is so scary': Nvidia CEO Jensen Huang slams Anthropic chief's grim job loss predictions](https://economictimes.indiatimes.com/magazines/panache/he-thinks-ai-is-so-scary-nvidia-ceo-jensen-huang-slams-anthropic-chiefs-grim-job-loss-predictions/articleshow/121783646.cms) - [Reuters Connect — Nvidia CEO says AI will create jobs, not cut them (GTC 2026)](https://www.reutersconnect.com/item/nvidia-ceo-pledges-support-for-teams-in-israel-taiwan-and-says-ai-will-create-jobs-not-cut-them/dGFnOnJldXRlcnMuY29tLDIwMjY6bmV3c21sX1ZBNTAyNjE3MDMyMDI2UlAx) - [Reuters Connect — Nvidia's CEO talks AI factories and the future of technology (Hill & Valley Forum 2025)](https://www.reutersconnect.com/item/a-revolution-nvidias-ceo-talks-ai-factories-and-the-future-of-technology/dGFnOnJldXRlcnMuY29tLDIwMjU6bmV3c21sX0xWQTAwMTA1MzYzMDA0MjAyNVJQMQ) - [NVIDIA Blog — 'Now We Can Know Everything and Do Anything,' Jensen Huang Says at Dreamforce](https://blogs.nvidia.com/blog/jensen-huang-dreamforce/) - [TechCrunch — We don't need AI regulation, leave safety to us, Nvidia's Jensen Huang says](https://techcrunch.com/2026/09/15/we-dont-need-ai-regulation-leave-safety-to-us-nvidias-jensen-huang-says/) - [Associated Press via NY1 — Nvidia's Jensen Huang says society needs new social norms in the age of AI](https://ny1.com/nyc/all-boroughs/ap-top-news/2026/06/16/ap-exclusive-nvidias-jensen-huang-says-society-needs-new-social-norms-in-the-age-of-ai) - [Transcript — Jensen Huang on The Joe Rogan Experience #2422](https://singjupost.com/transcript-nvidias-jensen-huang-on-joe-rogan-podcast-2422/) - [ScribeHawk — Leading in the Age of AI: A Conversation with NVIDIA CEO Jensen Huang, Milken Global Conference 2026](https://scribehawk.com/transcript/leading-in-the-age-of-ai-a-conversation-with-nvidia-ceo-jensen-huang-global-conf) - [CEOInterviews.AI — Leading in the Age of AI, Milken Global Conference 2026](https://ceointerviews.ai/interview/884429/) - [Yann LeCun on X, September 2026 (mirror)](https://www.sotwe.com/AlwaysNever0001?lang=en) - [IAPP — AI safeguards debate marches on for US policymakers, AI developers](https://iapp.org/news/a/ai-safeguards-debate-marches-on-for-us-policymakers-ai-developers) - [David Sacks on X, September 10, 2026 (MinutesOf archive)](https://minutesof.com/david-sacks/) - [David Sacks on X, September 10, 2026 (mirror)](https://www.sotwe.com/davidsacks) - [Business Insider / Yahoo Tech — Reid Hoffman dismisses AI doomers: 'Look at me, I'm Cassandra'](https://tech.yahoo.com/ai/articles/reid-hoffman-dismisses-ai-doomers-193616401.html) - [All-In podcast transcript — JD Vance](https://podcasts.happyscribe.com/all-in-with-chamath-jason-sacks-friedberg/jd-vance-on-ai-entitlement-fraud-iran-war-israel-h-1b-abuse-the-midterms) - [Axios — The Axios Show: Newsom on AI 'doomers'](https://www.axios.com/2026/03/27/gavin-newsom-axios-show-ai) - [The New York Times — Some in Silicon Valley Are Questioning the Calls for an A.I. Slowdown](https://www.nytimes.com/2026/09/13/technology/silicon-valley-ai-slowdown.html) - [Shyam Sankar on X, September 12, 2026 (mirror)](https://zamantika.com/ssankar/status/2098838592563052849) - [Brad Gerstner on X, September 2026 (mirror)](https://www.instalker.org/altcap) - [Business Insider / Yahoo — Hugging Face CEO on asking Jacob about AI extinction risk](https://currently.att.yahoo.com/att/hugging-face-ceo-says-asking-182505704.html) - [San Francisco Standard — Why do the people building AI keep warning us about the end of the world?](https://sfstandard.com/2026/09/11/ai-safety-warnings/) - [Pedro Domingos on X, September 2026 (mirror)](https://zamantika.com/profile/pmddomingos) - [Guillaume Verdon (@beffjezos) on X, September 11, 2026 (mirror)](https://ngntipkolamrenang.twstalker.com/beffjezos) - [Guillaume Verdon (@beffjezos) on X, September 2026 (mirror)](https://zamantika.com/beffjezos/status/2098610688369594650) - [Bindu Reddy on X, September 2026 (mirror)](https://instalker.org/bindureddy) - [Melanie Mitchell on Bluesky, September 9, 2026](https://www.bluenews.app/bsky/melaniemitchell.bsky.social?tab=media) - [David Bellamy on X, September 13, 2026 (mirror)](https://www30.twstalker.com/DavidRBellamy/status/2099187370407112758) - [Tim Dettmers on X, September 2026 (mirror)](https://jilbab.twstalker.com/Tim_Dettmers) - [Gary Marcus on X, September 2026 (mirror)](https://twstalker.com/GaryMarcus) - [Michael Burry — Cassandra Unchained note, September 14, 2026](https://substack.com/@michaeljburry/note/c-336629794) - [Anthony Pompliano on X, September 2026 (mirror)](https://w.twstalker.com/HyperfocuSam) - [HAPPINESS.info — The Safest AI Is The Smartest AI | Pedro Domingos](https://happiness.info/the-safest-ai-is-the-smartest-ai/) - [Progress — Brian Armstrong's blueprint for an American Wakanda](https://thisisprogress.substack.com/p/2x-unicorn-founder-on-why-ai-doomers) - [Robinson's Podcast — Tyler Cowen: The Economics of Artificial Intelligence](https://podscan.fm/podcasts/robinsons-podcast/episodes/261-tyler-cowen-the-economics-of-artificial-intelligence) - [KQED — Beyond the AI Hype Machine](https://www.kqed.org/news/12059911/beyond-the-ai-hype-machine) - [Julian Dorey Podcast transcript — Lee Cronin](https://podscripts.co/podcasts/julian-dorey-podcast/289-darpa-scientist-on-intelligent-life-the-universe-god-ai-lee-cronin) - [Transcript — Why Smart People Believe Stupid Things, Steven Pinker](https://singjupost.com/transcript-why-smart-people-believe-stupid-things-steven-pinker/) - [PBS — Canada Files: Steven Pinker](https://www.pbs.org/video/canada-files-steven-pinker-bbsnwj/) - [Fortune — Base AI policy on evidence, not existential angst](https://fortune.com/2024/12/14/ai-regulation-policy/) - [Andreessen Horowitz — Why AI Will Save the World](https://a16z.com/ai-will-save-the-world/) - [Dwarkesh Podcast — Tyler Cowen](https://www.dwarkesh.com/p/tyler-cowen-3) - [Le Monde — The Sam Altman affair has shown that artificial intelligence doomers have lost a battle](https://www.lemonde.fr/en/opinion/article/2023/11/30/the-sam-altman-affair-has-shown-that-artificial-intelligence-doomers-have-lost-a-battle_6301213_23.html) - [Techdirt — The AI Doomers' Playbook](https://www.techdirt.com/2023/04/14/the-ai-doomers-playbook/) - [TIME100 AI — Yann LeCun](https://time.com/collections/time100-ai/6309052/yann-lecun/) - [Techmeme, October 31, 2023 — Andrew Ng and John Carmack posts](https://www.techmeme.com/231031/p32) - [Thread Reader — Arvind Narayanan on the AI drone simulation story](https://threadreaderapp.com/thread/1681748271163912194.html) - [Techmeme, November 17, 2023 — Balaji Srinivasan post](https://www.techmeme.com/231117/p18) **Research, infrastructure, and science** - [Newsweek — Anthropic researcher quits, warns AI could kill everyone](https://newsweek.com/anthropic-researcher-quits-warns-ai-could-kill-everyone-12418798) - [MercoPress — Anthropic researcher resigns, warning that AI "could kill us all" before the decade ends](https://en.mercopress.com/2026/09/09/anthropic-researcher-resigns-warning-that-ai-could-kill-us-all-before-the-decade-ends) - [AI Weekly — Amazon, Microsoft, Alphabet, Meta Plan $725B AI Capex in 2026](https://aiweekly.co/alerts/amazon-microsoft-alphabet-meta-plan-725b-ai-capex-in-2026) - [Scientific American — Data Centers Will Use Twice as Much Energy by 2030, Driven by AI (IEA Energy and AI)](https://scientificamerican.com/article/ai-will-drive-doubling-of-data-center-energy-demand-by-2030) - [The Paper (澎湃新闻) — 工信部:推动建立算力服务能力评估、算力市场化定价等标准](https://www.thepaper.cn/newsDetail_forward_33619073) - [Xinhua Finance (新华财经) — 数读我国绿色算力:十组数据背后的时代答卷](https://www.cnfin.com/yw-lb/detail/20260822/4458983_1.html) - [Data Center Watch — Q1 2026 Report](https://datacenterwatch.org/q1-2026) - [Fortune — Data center dissent is so high that construction delays this year have already exceeded 2025's total](https://fortune.com/2026/06/16/data-center-opposition-construction-delays-blocks-report/) - [NIH Research Matters — Brain-computer device helps man speak](https://www.nih.gov/news-events/nih-research-matters/brain-computer-device-helps-man-speak) - [UC Davis Health — Brain-computer interface enables independent, accurate communication for man living with ALS](https://health.ucdavis.edu/research/news/headlines/brain-computer-interface-enables-independent-accurate-communication-for-man-living-with-als/2026/06) - [FDA — Implanted Brain-Computer Interface Devices for Patients with Paralysis or Amputation](https://www.fda.gov/regulatory-information/search-fda-guidance-documents/implanted-brain-computer-interface-bci-devices-patients-paralysis-or-amputation-non-clinical-testing) - [Merck — Phase 3 INTerpath-001 Trial of Intismeran Autogene Plus KEYTRUDA Met Endpoints of RFS and DMFS](https://www.merck.com/news/merck-and-moderna-announce-phase-3-interpath-001-trial-of-intismeran-autogene-plus-keytruda-met-endpoints-of-recurrence-free-survival-rfs-and-distant-metastasis-free-survival-dmfs-in-patient/) - [NCI Drug Dictionary — mRNA-based personalized cancer vaccine mRNA-4157](https://www.cancer.gov/publications/dictionaries/cancer-drug/def/intismeran-autogene) - [FDA — FDA Approves First Gene Therapies to Treat Patients with Sickle Cell Disease](https://www.fda.gov/news-events/press-announcements/fda-approves-first-gene-therapies-treat-patients-sickle-cell-disease) - [WHO — Human genome editing: a framework for governance](https://www.who.int/publications-detail-redirect/9789240030060) - [WHO — Human genome editing: recommendations](https://www.who.int/publications/i/item/9789240030381) - [WHO — New principles for ethical human genomic data collection and sharing](https://www.who.int/news/item/20-11-2024-who-releases-new-principles-for-ethical-human-genomic-data-collection-and-sharing) - [NIH — Bio Genesis Mission](https://www.nih.gov/bio-genesismission) - [Navy — Navy publishes scientific paper on USS Theodore Roosevelt COVID-19 outbreak](https://www.navy.mil/Press-Office/News-Stories/Article/2413348/navy-publishes-scientific-paper-on-uss-theodore-roosevelt-covid-19-outbreak/) - [Kasper et al. — An Outbreak of Covid-19 on an Aircraft Carrier (_New England Journal of Medicine_, 2020)](https://www.nejm.org/doi/full/10.1056/NEJMoa2019375) - [Military Health System — Vaccine Recommendations](https://health.mil/Military-Health-Topics/Health-Readiness/Immunization-Healthcare/Vaccine-Recommendations/) - [Military Health System — Update: measles, mumps, rubella and varicella among service members, 2019–2024](https://www.health.mil/News/Articles/2025/10/01/MSMR-MMRV) - [NCI Drug Dictionary — mRNA-based personalized cancer vaccine NCI-4650](https://www.cancer.gov/publications/dictionaries/cancer-drug/def/mrna-based-personalized-cancer-vaccine-nci-4650) - [Sunstein and Vermeule — Conspiracy Theories: Causes and Cures (_Journal of Political Philosophy_, 2009)](https://doi.org/10.1111/j.1467-9760.2008.00325.x) - [Roozenbeek et al. — Psychological inoculation improves resilience against misinformation on social media (_Science Advances_, 2022)](https://doi.org/10.1126/sciadv.abo6254) - [Federation of State Medical Boards — Spreading COVID-19 Vaccine Misinformation May Put Medical License at Risk](https://www.fsmb.org/advocacy/news-releases/fsmb-spreading-covid-19-vaccine-misinformation-may-put-medical-license-at-risk/) - [Berdoy, Webster, and Macdonald — Fatal attraction in rats infected with _Toxoplasma gondii_ (_Proceedings of the Royal Society B_, 2000)](https://doi.org/10.1098/rspb.2000.1182) - Boyd and Richerson — _Culture and the Evolutionary Process_ (University of Chicago Press, 1985) - [Seyfarth, Cheney, and Marler — Monkey responses to three different alarm calls (_Science_, 1980)](https://doi.org/10.1126/science.7433999) - [Wheeler — Monkeys crying wolf? Tufted capuchin monkeys use anti-predator calls to usurp resources from conspecifics (_Proceedings of the Royal Society B_, 2009)](https://doi.org/10.1098/rspb.2009.0544) - [Henrich and Gil-White — The evolution of prestige (_Evolution and Human Behavior_, 2001)](https://doi.org/10.1016/S1090-5138\(00\)00071-4) - [Ellis et al. — Review of the Safety of LHC Collisions (_Journal of Physics G_, 2008)](https://doi.org/10.1088/0954-3899/35/11/115004) - [Bresnahan and Trajtenberg — General purpose technologies: 'Engines of growth'? (_Journal of Econometrics_, 1995)](https://doi.org/10.1016/0304-4076\(94\)01598-T) - [Merchant et al. — Scaling deep learning for materials discovery (_Nature_, 2023)](https://doi.org/10.1038/s41586-023-06735-9) - [Lam et al. — Learning skillful medium-range global weather forecasting (_Science_, 2023)](https://doi.org/10.1126/science.adi2336) - [Musunuru et al. — Patient-Specific In Vivo Gene Editing to Treat a Rare Genetic Disease (_New England Journal of Medicine_, 2025)](https://doi.org/10.1056/NEJMoa2504747) - [López-Otín et al. — Hallmarks of aging: An expanding universe (_Cell_, 2023)](https://doi.org/10.1016/j.cell.2022.11.001)