“I don’t know if any rational person ever became a billionaire running a disruptive company.” — Keith Teare



Is capitalism by permission of democracy, or is democracy by permission of capitalism? That’s the question Keith Teare and I have been circling for a while on our weekly tech roundup, and this week it triggered a full-blown discussion of our 21st century economic and political fate.

Earlier this week, Vinod Khosla — one of Silicon Valley’s most successful venture capitalists — posted on X that “capitalism is by permission of democracy.” Keith agrees. I’m not so sure. My sense is that as AI start-ups approach valuations that rival the GDP of nation states, the old equation inverts. Governments no longer permit capitalism. Capitalism permits government. The Sam Altmans and Elon Musks of the future, running 10 or $15 trillion dollar startups, won’t lobby politicians. They’ll replace them. Dario Amodei’s confrontation with the US government, then, is a sneak preview of the future. Indeed, as what Om Malik [https://om.co] calls a “symbolic capitalist”, Amodei is a good example of the type of engaged capitalist who will usurp traditional politicians. That’s the good news. The bad news is that other examples of symbolic capitalists include Elon Musk and Peter Thiel.

 

Five Takeaways



•       Keith Says OpenAI Will Be Worth $10 Trillion in Five Years: I told him I’d take him to dinner if he’s right. He said I’d have to do more than that. His logic: NVIDIA promises $1 trillion in new revenue by the end of next year, Anthropic did $5 billion in new revenue in a single month, and the three expected IPOs — Anthropic, OpenAI, SpaceX — would together raise more money than the entire IPO market of the last decade. The Netscape moment, if it comes, won’t be a moment. It’ll be an earthquake.

•       Fundrise Is the Canary in the Coal Mine: A fund holding private shares in Anthropic, OpenAI, SpaceX, Databricks, and Anduril went public this week at $34 and closed above $100. Retail investors paying three times net asset value for companies that aren’t even public yet. Keith says that’s not irrational — it’s the market pricing the future. I’m less sure. History is littered with futures the market got catastrophically wrong.

•       Om Malik Reframes the Entire Debate: His essay on “neo-symbolic capitalism” argues that value in the 21st century derives from symbols, narratives, and reputation rather than products. In that framing, Amodei’s fight with the government isn’t a miscalculation — it’s brand-building. Musk is the master of it. Altman tries to wear every hat simultaneously. Peter Thiel is in Rome talking about the Antichrist. And the billionaires who signed the Giving Pledge now want out.

•       Keith and I Disagree on What $10 Trillion Means: Keith says the government retains power regardless of corporate size. Being big doesn’t give you political power unless governments are corrupt. I think that’s naïve. If AI companies approach valuations that rival the GDP of nation states, the old equation inverts. Government doesn’t permit capitalism. Capitalism permits government. The Amodeis and Musks of the future won’t lobby politicians. They’ll replace them.

•       Contrarianism Is at the Very Core of Innovation: The one thing Keith and I agree on this week. Every billionaire is irrational. Musk is on the spectrum. Thiel believes in the Antichrist. Amodei thinks he can fight the US government and win. Keith concedes: no rational person ever became a billionaire running a disruptive company. The question is whether that irrationality is a feature of capitalism or a threat to democracy. We disagree on the answer.

 

About the Guest



Keith Teare [https://thatwastheweek.substack.com/] is

Symbolic Capitalism vs. Symbolic Democracy: Will the $10 Trillion AI Startup Change Everything?

“We do not exist without nature — unless Silicon Valley figures something out in their bunkers.” — Natalie Kyriacou



Forget the Middle East for a moment. Or rather, don’t — because today’s petroleum war is an environmental catastrophe, perhaps even an ecocide. Militaries are the largest source of emissions on the planet. Trump uses Iran’s oil fields as a bargaining chip while assassinating its leaders, as if the price of petroleum is more important than human life (which it clearly is to him). Natalie Kyriacou [https://www.nataliekyriacou.com], an Australian environmentalist and author of Nature’s Last Dance [https://www.amazon.com/Natures-Last-Dance-wonder-extinction/dp/1923293613], isn’t surprised. Trump, she says, is the symptom rather than the disease. His rotten system of prioritising oil over human lives has been ruining the planet now for over a century. He’s just less polite about it.

Nature’s Last Dance is made up of what Kyriacou calls “tales of wonder” in our age of extinction. It tells the story, for example, of a 2000 oil spill off South Africa that threatened 90,000 African penguins and triggered the largest volunteer workforce ever assembled. Zoos, NGOs, school kids on bikes, Australians knitting sweaters all conspired to save the oiled penguins. It worked. At least in terms of those 90,000 penguins.

But did it change anything structurally? Perhaps not. But she’s arguing that the impulse to show up matters, that community is the unit of change, and that falling in love with the wonder of nature is the precondition for fighting for it. She presents forgiving Australian surfers who’ve been attacked by sharks now fighting to protect them. And she imagines birdwatching as a form of quiet rebellion.

But what does the world look like if this does, indeed, turn out to be nature’s last dance? Kyriacou’s answer is a kind of natural horror movie. A Hitchcockian David Attenborough movie: more pigeons, more rats and more “bin chickens” — Australia’s ibis, a bird that thrives in urban garbage. Nature’s revenge. So if we all don’t take up birdwatching, Kyriacou warns, we will all end up in The Birds.

 

Five Takeaways



•       Trump Is a Symptom, Not the Disease: Countries have prioritised oil over lives for centuries. Trump is just more abrasive about it. The US negotiated Kyoto and didn’t join it, designed Paris around its own preferences, then pulled out twice. Kyriacou argues we’ve been relying on a broken system long before Trump accelerated its collapse.

•       90,000 Oiled Penguins and the Largest Volunteer Workforce Ever Assembled: In 2000, an oil spill off South Africa threatened the largest colony of African penguins. What followed was extraordinary: zoos and NGOs from a dozen countries mobilised overnight, tens of thousands of volunteers arrived, Australians knitted sweaters. It didn’t stop oil. But it showed that the impulse to show up still exists, and that community is the unit of change.

•       AI Puts Our Destructive Relationship with the World on Steroids: Kyriacou’s sharpest point: the problem with AI isn’t water usage or compute power. It’s that AI amplifies every facet of humanity’s existing relationship with the planet. If we’re already this destructive, this divided, this extractive — AI makes all of it a million times more extreme. The same system that destroys nature destroys communities. It’s a systems failure.

•       No Country on Earth Is on Track: Not one country in the world is currently meeting its climate or nature targets. Not one. The UN has been stretched too thin, too bureaucratic, too afraid of self-criticism. World leaders set targets, shake hands, and go home to fail. Kyriacou wants to revive the UN, not destroy it — but she’s blunt about its limits.

•       

Nature's Last Dance: Natalie Kyriacou on Ecocide, Oiled Penguins, and Why We Need to Watch the Birds

“Nobody’s reality is more or less real.” — Kevin Ashton



It’s the chicken and egg question. What came first: stories or language? For Kevin Ashton [https://www.linkedin.com/in/kevinjashton/], the answer is stories. In his new book, The Story of Stories [https://www.amazon.com/Story-Stories-Million-Year-History-Uniquely/dp/0063438690], Ashton argues that rather than inventing stories with language, we invented language to tell stories. Stories, for Ashton, predate language. They are what makes us human.

300,000 years ago, Ashton argues, humans sat around night fires needing to talk about things they couldn’t point to — the past, the future, the Gods. So they created language. Grunts got grammatical. And the grammar had a structure that hasn’t changed since: character, chronology, consequence. Every sentence in every one of the world’s 7,000 languages is built upon the need to tell stories. Every conversation you’ve ever had contains a narrative. Even this one.

I asked Ashton whether this makes reality itself just another narrative and him just another postmodernist. Our brains construct reality, he explained, in the same way a graphic user interface constructs a desktop. Our dog sees a different rainbow to the one we see. But, in contrast with our dog, we tell stories about that rainbow.

Ashton is a technologist who first coined the term “Internet of Things”. But on AI, he is surprisingly critical. A large language model is a more complicated toaster, he says. It can produce language that fits the format of a story — character, chronology, consequence — because it’s digested millions of words. But it can’t produce meaning. We humans, in contrast, are made meaningful by our stories. That’s why you are reading this now.

 

Five Takeaways



•       We Invented Language to Tell Stories, Not the Other Way Around: Ashton’s central claim is that storytelling preceded and caused the evolution of language. A million years ago, humans around night fires needed to talk about things they couldn’t point to — the past, the future, the gods. Grunts became grammar. The structure hasn’t changed since: character, chronology, consequence. Every sentence in every one of the world’s 7,000 languages is built on this need to narrate.

•       Nobody’s Reality Is Real: Our brains construct reality the way a graphic user interface constructs a desktop — useful, not true. Your dog sees a different rainbow than you do. Whose is real? Both. Neither. Ashton isn’t a postmodernist — he’s arguing that our story-shaped brains are the lens through which all experience is filtered, and there is no stepping outside it.

•       The Bible Hitched a Ride on Writing: The world’s great religions spread because they were among the first stories to exploit writing as a distribution technology. The Bible is just a word for book. Scripture is a word for writing. Where those texts travelled, those religions still dominate today. Homer is an oral tradition frozen by the alphabet. The oldest surviving story in the world is Noah’s flood, and it comes from Southern Iraq, not Greece.

•       A Large Language Model Is a More Complicated Toaster: Ashton is brutally dismissive of AI. A machine can produce something that fits the format of a story because it’s digested millions of them. But it can’t produce meaning. Machines are inherently meaningless. We anthropomorphise them because that’s what our story-shaped brains do — we named our cars, now we’re naming our chatbots.

•       We Humans Are Made Meaningful by Our Stories: Ashton’s own life is the proof: a Birmingham DJ who learned Norwegian in nightclubs, fell for Ibsen, marketed lipstick for Procter & Gamble, and accidentally invented the Internet of Things because mascara kept going out of st

What Came First: Stories or Language? Kevin Ashton on the Story of Stories

“I don’t like the idea of losing out to a machine because I feel like I’m losing a part of myself in the process.” — Nelson Dellis, six-time USA Memory Champion



Most of us can’t remember our spouse’s phone number. We barely know our own. We haven’t read a physical map in years. Some of us don’t even know what a map is. Such is the impoverishment of mental life in our digital age.

Nelson Dellis [https://nelsondellis.com], unlike most of us, is a rich man — at least mentally. He can memorise a shuffled deck of 52 cards in under a minute. He stores every stranger’s phone number in his head for 24 hours before putting it in his phone — on principle. He’s a six-time USA Memory Champion, a computer science professor at Skidmore, and the author of a new book, Everyday Genius [https://www.amazon.com/s?k=Everyday+Genius+Nelson+Dellis], which suggests we can all be a lot smarter than our smart phones.

Dellis got into memory after watching his grandmother get lost in the fog of Alzheimer’s. And as a computer science professor, he’s equally terrified by what he now sees in the classroom. His students can’t craft an email without ChatGPT. They can’t focus. They can’t solve a problem without asking a machine. He warns that we’re outsourcing our cognitive agency to devices and mislabelling it as human productivity.

For Dellis, it’s the same mental atrophy that destroyed his grandmother. AI-generated mnemonics, he warns, feel “dead inside.” Our brains, like our language, are degenerating into slop. Thus the value of his hacks to restore our focus and boost our memories.

 

Five Takeaways



•       I Can’t Remember My Wife’s Phone Number: Neither can you. Neither can anyone under 50. We’ve outsourced our memories to devices and the consequences are only beginning to show. Nelson Dellis memorises every new phone number for 24 hours before putting it in his phone. Not because he needs to — because his brain needs him to.

•       His Grandmother Disappeared into Alzheimer’s and It Changed His Life: Dellis watched the woman who raised him become a shell of herself — unable to recognise her own grandson. He went down a rabbit hole into memory science, discovered a former champion’s audiobook, tried the techniques, and was hooked. He won his first US Memory Championship within two years. He’s won six.

•       If Everyone’s a Genius, Nobody Is: I pushed back on the book’s premise. Dellis conceded the point but held his ground: the techniques are learnable, the results are real, and the distinction between “genius” and “trained” matters less than the distinction between a brain that’s exercised and one that’s atrophying. The London cab driver study is his best evidence — hippocampi that grow with use and shrink without it.

•       AI Slop Is by Definition Forgettable: Dellis teaches computer science, so he’s no Luddite. But AI-generated mnemonics, he says, feel “dead inside.” The vivid, absurd, grotesque images that make memory techniques work are products of individual human imagination. A machine can’t generate weirdness. Not yet. Maybe not ever. His students can’t write an email without ChatGPT. That should terrify us more than it does.

•       Eat Your Blueberries: Four pillars of brain health: mental exercise, physical fitness, diet, and — the one that surprises people — social interaction. Dellis trains a 90-year-old and a five-year-old using the same techniques. Both can do things their peers cannot. The brain doesn’t expire at 70. But it does atrophy if you let your iPhone do the thinking.

 

About the Guest



Nelson Dellis [https://nelsondellis.com] is a six-time USA Memory Champion (2011, 2012, 2014, 2015, 2021, 2024), certified mountaineer and Everest summite

Have our iPhones Eaten our Brains? Nelson Dellis on Hacks to Restore our Focus and Boost our Memory

“If we don’t fight, then what are we doing?” — Jeff Boyd



How do you write fiction about contemporary America when reality itself is stranger than fiction? A country in which “alternative facts” is policy rather than satire. Where “truth” has been nationalized.

Jeff Boyd [https://www.jeffsboyd.com], an acclaimed young American novelist, sees fiction as refuge. For both writer and reader, it gets us inside the heads of people who both inflict and endure pain. And it enables the senseless to make sense. The news cycle can’t do that. A novel can.

Boyd’s second novel, Hard Times [https://www.amazon.com/Hard-Times-Novel-Jeff-Boyd/dp/1250348455], out today, is his latest attempt to make sense of the senseless. No, the title isn’t Dickensian — it’s from Curtis Mayfield. The song on the 1975 “There’s No Place Like America Today” album, with its cover juxtaposing some happy Americans in a car with others waiting miserably in the unemployment line. America might be great — but for whom, exactly? That dichotomy shapes Hard Times, which is set in a school on the South Side of Chicago where an innocent student gets shot and nobody can agree on what happened or why.

Is the American Dream over? Boyd isn’t quite sure. “As much as it feels impossible,” he says, “some part of me always wants to believe.” His characters fight — backs against the wall, cards stacked against them, but they don’t give in. That’s what Curtis Mayfield was singing about in 1975 and it’s what Jeff Boyd is writing about in 2026. The times are hard. A time, once again, for novelists to seize back reality.

 

Five Takeaways



•       How Do You Make Stuff Up When Reality Is Already Unbelievable? Boyd admits he sometimes wonders what the point of being a novelist is when the headlines are stranger than fiction. His answer: fiction is a refuge. It lets you get inside the heads of people who inflict pain or endure it, and try to make sense of what in reality remains senseless. The novelist can provide an answer. The news cycle can’t.

•       Not Dickens — Curtis Mayfield: The title comes not from the 1854 novel but from the 1975 song on There’s No Place Like America Today. The album cover says it all: happy people in the car, desperate people in the unemployment line. America is great — but great for whom? That dichotomy drives the book.

•       A Policeman’s Son on George Floyd: One of the officers who stood by while George Floyd died was black — a man whose family had been proud of him for getting the job, who went in wanting to do good. Boyd can’t write off an entire category of people. His black cop character in Hard Times exists to show the complexity of wanting to do right and getting caught up in wrong.

•       Fate vs. Agency on the South Side: Boyd’s grad school friend — not religious but deterministic — argued you could draw a line from where someone starts to where they’ll end up. Boyd’s characters fight against that line. A kid from a broken home on food stamps doesn’t have to end where you think. The novel asks whether the line holds or breaks.

•       The Fight Goes On: Is the American Dream over? Boyd isn’t quite sure. His characters have their backs against the wall and the cards stacked against them, but they don’t give in. That’s what Curtis Mayfield was singing about in 1975. It’s what Boyd is writing about in 2026. The times are hard. The fight goes on.

 

About the Guest



Jeff Boyd [https://www.jeffsboyd.com] is the author of The Weight [https://www.amazon.com/Weight-Jeff-Boyd/dp/1668007258] (Simon & Schuster, 2023) and Hard Times [https://www.amazon.com/Hard-Times-Novel-Jeff-Boyd/dp/1250348455] (Flatiron Books, 2026). A former Chicago

Hard Times Again? Jeff Boyd on Chicago, Charles Dickens and Curtis Mayfield

“Had another nation done this, we would regard this as an act of war.” — Arthur Levine, President of Brandeis University



Forget Iran for a moment. I asked Brandeis [https://www.brandeis.edu/] President Arthur Levine [https://www.brandeis.edu/president/] whether the Trump administration has gone to war with the American university. He paused diplomatically. “Going to war is a very restrictive term,” he answered. Then added: “Had another nation done this, we would regard this as an act of war.” From the president of Brandeis, that’s not a metaphorical dodge. He is, of course, referring to the singling out and bullying of Harvard, Columbia, the University of Pennsylvania and other universities by executive order. Levine trusts nothing like this will happen again. But he also trusted it wouldn’t and shouldn’t have happened in the first place.

Levine is back on the show with a new book, From Upheaval to Action: What Works in Changing Higher Ed [https://www.amazon.com/Upheaval-Action-Works-Changing-Higher/dp/1421454033], co-authored with Scott Van Pelt. Last time we talked [https://keenon.substack.com/p/is-that-320000-college-degree-really], we argued about whether the $320,000 degree is worth it. This time our conversation wasn’t so much about whether the degree is worth the exorbitant price tag, but whether the institution that grants it will survive. Indeed Brandeis is about to announce guaranteed transparent pricing — a necessary revolution in an industry that has, for too long, thrived on financial opacity.

A more existential threat to universities like Brandeis is AI. In this week’s That Was The Week [https://keenon.substack.com/] tech roundup, Keith Teare noted that even engineers at major tech companies are being told to stop coding and run AI instead. I tell the story of a UC Berkeley student who told his professor he didn’t need to read anymore because AI could do the reading for him. For Levine, this represents a failure of education, not a triumph of technology. Reading and writing are muscles, he says. You don’t build intellectual heft by outsourcing thinking to smart machines.

Levine draws the Luddite parallel. He argues the early 19th century craftsmen got better-paid work in factories. Every technological revolution produces fear, displacement, and eventually adaptation, he warns. So are university faculty the modern-day craftsmen? Their work will change, Levine explains. AI will take the routine parts with new more creative jobs emerging. But anyone who tells you they know what those jobs are is making it up, he says.

I pushed him on Epstein and the ethical rot of the American elite. He deflected — “we’re talking about a very small number of people” — but eventually conceded that ethics should be woven into every undergraduate subject, not taught as a single standalone course. I’m not sure that goes far enough. When university presidents are resigning because they took money from a child trafficker, it suggests that something is really rotten.

On DEI, Levine is surprisingly blunt: drop the term. It’s become a target for both left and right. Replace it with full access to higher education for those who can benefit from it. He sold this full access program to Democrats as equity and to Republicans as workforce development. Both bought it. The label was the problem, he explains, not the policy.

Henry Adams went to Harvard in 1850 and said he received an 18th century education for a world preparing for the 20th century. The worst mistake, Levine says, is not adapting to change. On that, Luddite university faculty, and perhaps even Donald Trump, might agree.

 

Five Takeaways



•       “Had Another Nation Done This, We Would Regard It as an Act of War”: Brandeis President Arthur Levine chose his words with the care you’d expect

An Act of War? Brandeis President Arthur Levine on Trump’s University Policy

“You would not want to be me.” — Elon Musk



Yesterday I argued that Dario Amodei is the most interesting man in America because he’s doing something nobody else has the balls to do: acting like a human being in public. Elon Musk is the opposite. He has the balls — nobody would deny that — but what’s missing is the human-being. Or perhaps Elon is all-too-human, which explains why so many of us — including myself — loathe him.

Charles Steel [https://charlessteel.com/], a London investor, doesn’t loathe Elon. In fact, he’s self-published a book about him: The Curious Mind of Elon Musk: Nine Ways He Thinks Differently [https://www.amazon.com/Curious-Mind-Elon-Musk-Differently/dp/B0FPZTT534]. Rather than an Elon hagiography, Steel insists, it’s an attempt to explain why Musk admirers don’t fully understand him, and the Hate-Elon crowd would probably loathe him for different reasons even if they had full navigation rights to his mind.

As I said, I’m in the second camp. My dislike of Musk is political — the cosying up to Trump, the DOGE fiasco, the embrace of far-right groups, the transformation of Twitter into a safe space for misanthropes. But Steel makes a case that, in our therapeutic culture, might be harder for some to dismiss: Musk’s “curious mind” is the product of childhood bullying, high-functioning autism, an abusive father, and an existential crisis resolved not by philosophy but by The Hitchhiker’s Guide to the Galaxy. Apparently Elon read Nietzsche and that, of course, only compounded his existential crisis. Probably because Nietzsche was warning us about a future dominated by philistines like Elon Musk.

In navigating the Musk mind, Steel discovers three traits: hyper-rationality, existential angst, and belligerence. Lots of Silicon Valley founders have the first. Some have the second. Almost none have the third. The combination produces a man who genuinely believes that the scientific method — the right of anyone to criticize anything — is a secular religion, and that “wokeness” is a competing religion that must be destroyed. Whether or not you buy this self-serving argument, Steel might be right to stress a Musk worldview — even if that worldview is often childishly indefensible.

I suggested to Steel that Musk is trapped in a Hobbesian state of nature — frozen alone, unable to read other people, incapable of separating himself from himself. A kind of naturally narcissistic state. This is what I most dislike about Elon. That he’s normalizing this state of nature. Nietzsche might (like his contemporary disciple Peter Thiel) have called him the Anti-Christ. He’s certainly the anti-Dario.

 

Five Takeaways



•       Musk Is the Anti-Dario: Amodei acts like a human being in public. Musk has the balls but what’s missing is the human-being. Or perhaps he’s all-too-human, which explains why so many of us loathe him. The contrast between them is the story of Silicon Valley in 2026.

•       Steel’s Case Is Harder to Dismiss Than You’d Think: Musk’s “curious mind” is the product of childhood bullying, high-functioning autism, an abusive father, and an existential crisis resolved not by philosophy but by The Hitchhiker’s Guide to the Galaxy. He read Nietzsche and it made things worse. Probably because Nietzsche was warning us about philistines like Musk.

•       Three Traits: Hyper-Rationality, Angst, and Belligerence: Lots of Silicon Valley founders have the first. Some have the second. Almost none have the third. The combination produces a man who believes the scientific method is a secular religion and wokeness is a competing one that must be destroyed. Whether or not you buy this self-serving argument, Steel might be right to stress a Musk worldview — even if it’s often childishly indefensible.

•       Trapped in a Hobbes

Is Elon Human? Charles Steel on the Curious Mind of Elon Musk

“Whether you like Amodei or not, at least he’s a leader.” — Andrew Keen



Dario Amodei is the most interesting man in America right now. Not because he runs a $500 billion company or because he’s suing the Trump administration or because Anthropic’s Claude topped the iPhone charts. But because he’s doing something nobody else in Silicon Valley has the balls to do: he’s acting like a human being in public. He has principles, he states them, and he accepts the consequences. That’s leadership. It shouldn’t be remarkable. In 2026, it is.

This week’s That Was The Week [https://thatwastheweek.substack.com/] is about how America both loves and hates AI. An NBC poll found 60–70% of Americans are concerned about AI — making it even less popular than the Democratic Party (quite an achievement). A hundred planned data centers have been cancelled because of local protests. 10,000 authors published an anti AI manifesto at the London Book Fair this week. Each week, in contrast, a billion people used ChatGPT, but these users often seem oblivious to its weaknesses. So Keith’s AI-generated video for the show was, by universal agreement (including his own), not going to win an Oscar tomorrow. Except for Most Sloppy AI generated video.

Every road this week led back to Amodei who is anything but sloppy. He’s become a Rorschach test for the entire industry. Tech progressives Kara Swisher and Scott Galloway are lauding him. The MAGA crowd — including David Sacks, Trump’s AI czar — on the All In podcast are doing the opposite. Keith thinks Dario is a naive CEO making bad business decisions — comparing him to his own doomed battle in the late Nineties against Microsoft’s Steve Ballmer. It’s a fair point. Should a tech CEO really be setting AI policy? Keith’s answer is no — that’s for people like David Sacks appointed by executive, legislative, and judicial branches. I’m not so sure. In an America defined by its dysfunctional political system, we need leaders like Amodei to take ethical stands. If not, then who?

The IPO race this year between Anthropic, OpenAI and xAI makes this particularly interesting. I wonder whether Amodei might use the IPO itself to force a public debate that nobody in government is willing to have. Not just about guardrails or weapons — but about what kind of society AI is building and who gets to decide what does and doesn’t get used. Musk, by publicly embracing white racists and other groups of hate, is making his politics clear. Sam Altman, as always, is wearing every hat simultaneously. Amodei, in contrast, knows his hat. Rather than MAGA, it should say: The Most Interesting Man in America. He’s got my vote. Even if he’s not running for office.

 

Five Takeaways



•       AI Is Less Popular Than the Democrats: An NBC poll found 60–70% of Americans are concerned about AI. A hundred data centres have been cancelled due to local protests. 10,000 authors published an anti-AI manifesto at the London Book Fair. Close to a billion people use ChatGPT each week — but the haters are the non-users, and they outnumber the lovers by a wide margin.

•       Amodei Is the 21st Century’s First Real Leader: He’s suing the Trump administration. He’s refusing to let Claude be used for autonomous weapons. He’s accepting the business consequences. Keith thinks he’s naive. I think he’s the only person in Silicon Valley acting like a human being in public. The debate between us is the show.

•       Keith Compares Amodei to His Own Doomed Battle Against Ballmer: In the late Nineties, Keith fought Microsoft with RealNames and lost. He sees Amodei on the same trajectory — noble, principled, already finished. I compared Keith to Pete Hegseth declaring the Iranian regime defeated. The MAGA crowd on All In, including Trump’s AI czar David Sacks, agree with Keith. That alone should give him

Why Dario Amodei Might Be the 21st Century’s First Real Leader

“I can point to things. But is that a systemic explanation? I think there the answer is a little less clear. I mean, surely people need love and all of that, but then there’s this risk of just devolving into platitude.” — David Sussillo



David Sussillo [https://www.amazon.com/s?k=emergent+david+sussillo] is a big time neural reverse engineer. The Stanford brain scientist worked at Google Brain with Geoffrey Hinton, and now is at Meta Reality Labs. What distinguishes Sussillo, however, is not his Silicon Valley good luck, but the bad luck of his origins. In his memoir, Emergent: A Memoir of Boyhood, Computation, and the Mysteries of the Mind [https://www.amazon.com/s?k=emergent+david+sussillo], Sussillo begins at the Albuquerque Christian Children’s Home — a modern-day orphanage — and the Milton Hershey School, the boarding school endowed by the chocolate magnate for kids with nowhere else to go. Both his parents were addicts. His mom died young. His dad spent his life as an untrained preacher ministering to homeless people on the streets of Albuquerque while managing a lifelong heroin habit.

The book’s thesis borrows from the science he studies: “emergence” — simple things interacting to produce complex behaviour that none of them could produce alone. His life is both proof of and a challenge to this concept. He made it out. Most of the kids he grew up with didn’t. He can point to moments — a gifted-and-talented test in third grade, an aunt and uncle’s intervention at nine, a first love in college — but he can’t build an explanatory system from these haphazard events. The Sussillo quilt doesn’t have an innate pattern. It just has patches.

What makes Sussillo unusual as a memoirist is his refusal to sentimentalise. Twenty years of psychotherapy, he confesses, has taught him something most authors never learn: that understanding your own story doesn’t mean you’ve explained it. His science can’t explain his childhood either. “The big dirty secret of neuroscience,” he says, “is that we don’t really understand much in the ways that people would love us to understand.” The man who reverse-engineers neural networks can’t reverse-engineer himself.

I asked him whether having children would have been harder than writing the book. Yes, he said. With the book, you can take a break. With kids, you relive things through a very specific way of relating. He and his wife chose not to. His mentors all told him he’d have been great at it. He’s not so sure. That honesty — the willingness to say “I don’t know” and mean it — runs through everything Sussillo does. He says he’s happy, claiming to have found peace with his past. But he still carries the baggage. Who wouldn’t? He’s just learned to manage it. Emergent, not emerged.

 

Five Takeaways



•       From Orphanage to Google Brain: Both parents were heroin addicts. Sussillo grew up in a modern-day orphanage in Albuquerque and then the Milton Hershey School. He went on to work at Google Brain with Geoffrey Hinton, now works at Meta Reality Labs, teaches at Stanford. Most of the kids he grew up with didn’t make it.

•       Emergence as Autobiography: The book’s thesis borrows from the science he studies: simple pieces combining into complicated outcomes. His life is the proof of concept and the counter-example simultaneously. The quilt doesn’t have a pattern. It just has patches.

•       The Dirty Secret of Neuroscience: The man who reverse-engineers neural networks can’t reverse-engineer himself. “We don’t really understand much in the ways that people would love us to understand.” Twenty years of therapy taught him more than the science.

•       Would Kids Have Been Harder Than the Book? Yes. With the book, you can take a break. With kids, you relive trauma through a very specific way of

From Orphanage to Google Brain: David Sussillo on Heroin, Neural Networks and the Mysteries of the Heart

“They all did it. They’re all guilty.” — Amy Littlefield



Who killed Roe? Amy Littlefield [https://www.thenation.com/authors/amy-littlefield/], the abortion access correspondent at The Nation and big time Agatha Christie fan, has written a true crime book about it. Literally. Killers of Roe [https://www.amazon.com/s?k=killers+of+roe+amy+littlefield] treats the death of the constitutional right to abortion as a murder mystery in the Poirot or Miss Marple tradition, complete with suspects, motives, and a forensic reconstruction of the 50-year crime scene. The suspects have Christie-style names: the Racist (Jesse Helms), the Little Brother (James Buckley), the Devout Bureaucrat (Paul Herring), the Closeted Congressman (Bob Bauman), and of course Mr Hyde Amendment himself, Henry Hyde — six foot three, helmet of white hair, serial groper of women who ensured poor women lost access first.

The Hyde Amendment is where the crime begins: 1976, a ban on federal funding of abortion. If you’re poor, the Supreme Court ruled, that’s your problem. The constitutional right exists, but don’t expect anyone to pay for it. Surprise surprise. Black women, low-income women, women on Medicaid understood immediately. Democrats and mainstream pro-choice groups took longer to notice. By which time the damage was done — and the playbook established: chip away at access rather than try to ban it outright.

Littlefield is more Miss Marple than Poirot — unassuming, persistent, sitting with her suspects for hours until they tell her why they did it. The devout bureaucrat, Paul Herring, spent their interviews trying to convert her to Catholicism. Henry Hyde made a pass at the president of Planned Parenthood during a commercial break on the Phil Donahue show. Bob Bauman — closeted, adopted, alcoholic — confessed to her that his anti-abortion politics may have come from identifying with the unwanted fetus, because that could have been him. These are complicated people doing terrible things for reasons they believe are righteous.

And the ending? Littlefield steals it from Murder on the Orient Express. They all did it. Every suspect is guilty — including the Democrats who failed to defend poor women, and the pro-choice movement that didn’t fight hard enough for the most vulnerable. Since the Dobbs decision in 2022: 59 excess pregnancy-associated deaths, 500 additional infant deaths, 22,000 additional births. The numbers aren’t a Miss Marple mystery. The crime is ongoing. And Trump, who declared himself “very pro-choice” before he appointed the justices who drove the final nail in, is the ultimate opportunist — a fat, orange haired version of Hyde. Murder on the Abortion Express. They all did it. All the men, at least.

 

Five Takeaways



•       The Hyde Amendment Is Where the Crime Begins: 1976. A ban on federal funding of abortion. Poor women lost access first. Black women, women on Medicaid understood immediately. Democrats and mainstream pro-choice groups took longer to notice. By which time the playbook was established.

•       The Anti-Abortion Movement Stole the Language of Civil Rights: White conservatives who didn’t want to think about the harms of white supremacy found an escape valve: their own civil rights movement, with the fetus — almost always imagined as white — as the victim.

•       The Suspects Are Complicated. The Crime Is Not: Henry Hyde groped women during commercial breaks. Bob Bauman — closeted, adopted, alcoholic — identified with the unwanted fetus. Paul Herring tried to convert Littlefield to Catholicism. Complicated people, terrible consequences.

•       The Numbers Are Real: Since the Dobbs decision in 2022: 59 excess pregnancy-associated deaths. 500 additional infant deaths. 22,000 additional births. The crime is ongoing.

Murder on the Abortion Express: Amy Littlefield on Who Killed Roe

The daily show about the big ideas, new technologies, books and people changing the world. Hosted by Andrew Keen.

Keen On