Illustration: Justin Metz

Julia Wise, a mother of three in Boston, published an essay last February that articulated a new kind of parenting anxiety. Titled “Raising Children on the Eve of AI,” it set out to answer a pressing question that no best-selling child-rearing gurus had asked. “I feel a bit like we’re preparing children to be good blacksmiths or shoemakers in 1750 when the factory is coming,” she wrote. “The families around us are still very much focused on the track of do well in school get into a good college have a career have a nice life.” That reality, she suggested, would expire within her children’s lifetimes.

Wise and her husband, Jeff, think about this stuff a lot. They are part of a community, the Effective Altruists, that has spent years gaming out different AI scenarios, both the rosy and the highly destructive. For a long time, those scenarios didn’t feel applicable to their own lives. But as AI development has sped up, she and Jeff, who works in biosecurity and pandemic detection, have become more concerned about how their children (ages 4, 9, and 11) will fare. They worry about everything from AI making it easier for a bad actor to unleash a world-ravaging pathogen to their kids getting attached to an emotionally expressive superintelligence. “We and some other parents we know have been thinking, Okay, it looks like there may be big changes in the next decade or two. What does that look like for how we prepare our children for the world?” she told me.

When we spoke recently, Wise laid out several visions of the future awaiting her children:

1. They won’t need careers because, well, “I take seriously the possibility there will be a disaster and my children may not live to adulthood.”

2. The world becomes a glorious post-scarcity utopia where no one needs to work and we all receive a universal basic income.

3. AI takes over most jobs, conventional careers cease to exist, and humans’ work is marginalized into limited roles.

4. There’s a disaster, but only some die and survivors need to navigate a post-apocalyptic landscape, perhaps as hunter-gatherers. (Wise is not optimistic about this possibility. Her two eldest children do Girl Scouts, but “we went on a camping trip, and we all had trouble building a fire,” she told me.)

Cataclysmic stakes aside, Wise and her husband could pass for relatably angsty parents conflicted about what to do or even whether to do anything. In her essay, she acknowledged that it would have been premature in 1750 to stop training blacksmiths and shoemakers. “I don’t want to be a crackpot who fails to prepare my children for the fairly normal future ahead of them because I wrongly believe something weird is about to happen,” she wrote.

“But thinking more analytically,” Wise said, “I do not expect normalcy.”

Every society trains its children for the demands of the moment. Elites from Alexander the Great to Regency Era lordlings on their grand tours had private tutors to mold them into world-beating polymaths, or at least clubbable men of leisure. When America was mostly agricultural, spinsters in one-room schoolhouses provided moral instruction using McGuffey’s Eclectic Reader and taught arithmetic using story problems about butter-churning and the measurements of an ox’s hindquarters. With the rise of industry in the 19th century, schools with regimented rows of desks and hourly bells and a foremanlike teacher cranked out punctual, docile workers to staff the new factories. The meritocratic turn of the millennium, with its résumé-padded achievement track, yielded David Brooks’s “Organization Kid,” a striving, overscheduled conformist engineered to assume their rightful place in the credentialed elite.

The challenge of this moment, though, is the utter lack of consensus as to what may happen and which traits or skills will help when it does. Even the class of parents who spend their days thinking about how technology will change the future is casting about for an answer to the ancient but newly urgent question of how best to raise a child. “We’re all thinking about this stuff,” said Rachael Horwitz, a Silicon Valley veteran (Twitter, Google, Facebook, Coinbase, Andreessen Horowitz) who’s CMO at a crypto venture-capital firm and a mother of two. Should kids all — as Geoffrey Hinton, the “godfather of AI,” advised — go to vocational school? Or should they be going to charm school? Horwitz has joked that, to AI-proof kids, parents should “no duh: Raise ENFJs,” as in the “Extroverted Intuitive Feeling Judging” Myers-Briggs personality type. “You don’t have to be technical,” offered Horwitz, who has spent her career in the nontechnical minority at engineer-dominated companies, “and the less of a bureaucrat you are, the better. I think we’re going to enter a world that’s going to triple-value people like that.” Do children even need to know things anymore? If it were ten or 20 years ago, Anthropic co-founder Ben Mann recently mused in an interview, he might have loaded his daughter up with extracurriculars to game her admission into a top-tier school. “But at this point,” he said, “I don’t think any of it’s going to matter … The facts are going to fade into the background.”

It is likely beyond the capacity of humans to realistically imagine how, and how fast, a superintelligence — a rapidly, recursively self-improving alien on earth — might reshape the course of events. The World Economic Forum published a Future of Jobs Report projecting that 65 percent of kindergartners will eventually work in “completely new job types that don’t yet exist,” while Yuval Noah Harari, the author of Sapiens, has suggested that the world may increasingly separate into a cognitive elite and an underclass of “useless” people. Even techno optimists acknowledge that the skills children will need are evolving faster than schools can, and that the most utopic AI outcome will still be completely unrecognizable in its particulars and its problems. (If none of us has to work anymore, will meaninglessness and overindulgence lead to a specieswide crisis of ennui and gout?) How, then, do you futureproof your child when the future suddenly appears exponentially more uncertain?

Just before the pandemic, Edward Nevraumont, an executive consultant in Seattle, became obsessed with generative AI. The second GPT model had recently been released; it behaved like your dimmest uncle — fact-challenged, meandering, self-repeating — but showed glimmers of vast potential. “I saw the writing on the wall,” Nevraumont told me. “This is going to happen.” Almost immediately, it became clear to the father of four that the technology would be an existential crisis for conventional school. “A lot of the world is proof of work — college applications, reference letters, thank-you notes — where the only value is time and effort,” he said. “Traditional education models have taught that it is the effort that matters,” but that’s rendered meaningless in a world where anyone with a free ChatGPT account can effortlessly perform effort. “AI,” he said, “kills the proof-of-work stuff.”

As a consultant, Nevraumont started helping private-equity firms learn how to operationalize AI. And last year, disillusioned with the education his kids were getting in Seattle, he moved his family to Texas so they could attend an unusual school.

Founded in 2014 as a tiny K–8 private school in Austin, the Alpha School has opened 15 additional campuses, from Scottsdale to San Francisco, on the strength of a tantalizing pitch. Using its AI-driven digital platform, Alpha asserts students learn “2.6” times faster on average than in regular schools while doing only two hours of schoolwork per day. The school has named this platform, which knits together proprietary and third-party apps, “TimeBack.” With those newly liberated hours, students focus on learning the life skills that Alpha’s co-founder MacKenzie Price believes standardized education neglects — things like entrepreneurship and developing “a growth mind-set.” At the flagship campus, a second-grader, in order to ascend to third, must complete a checklist that includes running five kilometers in 35 minutes or less; delivering a two-minute TED-style talk with “zero filler words, 120–170 [wpm] pace, and 90% confidence,” as judged by an AI speech coach named Yoodli; and calling “a peer’s parent” to “independently plan and schedule a playdate.” At Alpha’s middle school, projects have included starting and running an Airbnb and sailing a boat from Florida to the Bahamas.

Alpha has developed a fandom among parents who are disenchanted with the achievement track — the well-trodden path to a life of office dronedom — but are all in on spectacular, world-bestriding achievement. Its answer to the question of how to prepare children for the future (do school better) resonates with an influential class in Silicon Valley that has long viewed normie education as effectively running legacy code. The critique, roughly: Ability is more important than degrees, creativity trumps rote knowledge, traditional schools extinguish individuality and foster learned helplessness.

On a recent Thursday morning when I visited Alpha Austin, the drop-off line — Tesla, Hummer, Porsche Cayenne, another Tesla, Mercedes G-Class — felt of a piece with the city’s influx of heterodox podcasters and Silicon Valley refugees. (Students at one or another of Alpha’s constellation of schools in and around the city include children of tech investor and Substacker Byrne Hobart and founder of Lambda School and, now, an AI bootcamp Austen Allred.) In Level 2, which groups about three dozen fourth- and fifth-graders, the day began with a “morning launch” exercise, which today was a competition to slingshot balls at targets of varying difficulty. Then they fetched their laptops and dispersed — most to desks facing a wall or window, a few to pay-phone-like “Zen booths” for concentration, and others into beanbag chairs. On each of their screens was a dashboard displaying different-colored rings for each core subject, which would fill in as they completed the day’s allotment of readings, exercises, and tests.

Over the next hours (closer to three than two for some students), the kids in Level 2 worked quietly, moving from subject to subject in whatever order they chose. Each student’s work is set at a personalized level of difficulty, and they can move forward to the next grade in a subject only once they’ve achieved 90 percent “mastery” — in other words, proof of ability rather than proof of work. This means a child may be ahead in one subject but behind in another. Smith, an upbeat 9-year-old with a skateboarder’s fringe of hair tucked under a backward baseball cap, had been homeschooled until about four weeks earlier. In reading, he was at a kindergarten level, he told me, but he had already completed some fourth-grade checklist items, including climbing 110 flights of stairs.

Meanwhile, Alpha’s AI proctor was recording both what was happening on the students’ screens and, through their laptop cameras, their eye movements and facial expressions. All the data were being piped into Alpha’s “vision model.” If a student gets distracted, the software might display a message (“Rushing through questions is not a good habit”) or alert a “guide,” Alpha’s name for the adult in the room who monitors each student’s progress on their own laptop dashboard.

Despite its near-future sheen, Alpha’s platform is informed by old pedagogical concepts. Both the bespoke lesson plans and the idea of learning to the point of 90 percent mastery are derived from the work of the educational psychologist Benjamin Bloom, whose findings have experienced a recent resurgence in the edtech world. In 1984, Bloom published research showing that, when given one-on-one tutoring to mastery, “the average tutored student was above 98% of the students in the control class.” This became known as Bloom’s Two Sigma Problem — the “problem” referring to the fact that one-on-one tutoring was never going to be available to a lot of kids. Until, perhaps, AI came along.

Other principles of pedagogy are at work too. Sarah Cone, a venture capitalist in Manhattan, told me that when looking for a school for her 7-year-old daughter, “I toured all the best ones, and there’s not a single hot private school that uses spaced repetition” — reviewing learned material at increasingly long intervals — even though many studies have suggested that it’s one of the best ways to retain new information. Ten years ago, Cone did due diligence on a digital tutor that DARPA had developed for the Navy to train IT workers and found the results “mind-blowing.” She’s been “AI-tutor-pilled” since and had been tracking Alpha for years. Last month, when the NYC campus opened, she became one of its two dozen “founding families.” Two of her daughter’s new schoolmates had left Horace Mann to attend Alpha.

The approach wouldn’t work for everyone, even if money were no issue. (Alpha Austin’s yearly tuition is $40,000.) Not every 8-year-old is willing to stare intently at a screen to do academic work for hours a day. School boards across the country have responded coolly, if not with hostility, to Price’s past attempts to start Alpha-style charter schools. It’s too early to know to what extent selection effects — Alpha parents skew smart, motivated, and prosperous — are responsible for the promising early academic results reported by the school, including average SAT scores of 1530 for the class of 2025 and 1420 for the class of 2026.

That morning in Level 2, when a fidgety boy at his desk slid off his chair onto his knees, a guide named Noah Rodriguez gently coaxed him: “Hey dude, can we sit in the chair correctly? Thanks, buddy.” Rodriguez, who joined Alpha Austin as a guide in May, had spent two years as a math teacher at a public middle school in San Antonio. He loved it, but “there were just so many things during the day that felt like we were doing it just to do it,” he told me. Still, at Alpha, he has had to adjust to the fact that it’s no longer his job to impart knowledge. If a kid’s struggling with math, Rodriguez’s role is not to teach but to coach: Does he need a break? Maybe he should take a drink of water? If the kid gets three problems in a row wrong, rereviews all the relevant material, and is still stuck, they can schedule a virtual coaching call with a subject specialist. “It’s hard not to want to help, but I’m getting him to think like a self-driven learner,” Rodriguez told me. “It’s really cool to see them figure it out on their own because then they get the excitement of, Oh, I don’t need teachers to tell me what to do.”

Matt Bateman, a father of two who works for the school as what he calls its “in-house philosopher,” argued that the “1,000 percent personalized” learning, whereby a student drives their own progress, works in tandem with the life-skills exercises, which are “designed to get the student to believe by demonstration, by living it, that they can do incredible things, push themselves, and that it’s worth it to do so.” Second- and third-graders are expected to “look in three places for a lost item” and “attempt a solution to a problem” before asking a peer for help. After lunch that day, the fourth- and fifth-graders split into groups. One did a Chopped-style competition to practice collaboration and giving and receiving feedback; another did a workshop called “Fear Factor” in which they learn how to coach peers through their fears and face their own.

The self-directing quality that Bateman calls “agency” first drew his interest when he was getting a Ph.D. in philosophy. He gave the “small example of ordering off-menu at a restaurant. You’re not stuck with the options given to you. When you develop this way of thinking — what are all the options? What’s in my power? — it’s very morally powerful.” It’s the opposite of what he calls “patiency”: “The universe is doing things to you.” Bateman has mounted pegboards around his house in Austin, the lower sections of which are hung with art supplies, tools, and batteries (when a battery dies, his kids know where the replacements are), and he recently let his diaper-wearing 2-year-old climb an eight-foot ladder.

Lately, on X and LinkedIn and the other places where Silicon Valley thought leaders dwell, “agency” has become a way to describe what sets founders apart from civilians: a bias toward action, a willingness to ignore the haters, a steadfast belief that “you can just do things.” Credit for turning the philosophical concept into a buzzword has sometimes gone to Eric Weinstein, a conspiracy-minded mathematics Ph.D. (he believes the truth has been suppressed about UFOs and his own theory of physics, among other things) who used to work for Thiel Capital. In 2016, he appeared on The Tim Ferriss Show and described “high agency” as the essential quality of innovators. “When you’re told that something is impossible,” he elaborated, “is that the end of the conversation, or does that start a second dialogue in your mind — how to get around whoever it is that’s just told you that you can’t do something? So how am I going to get past this bouncer who told me that I can’t come into this nightclub? How am I going to start a business when my credit is terrible and I have no experience?”

In the past three years, the concept has acquired a new aura as agency and agentic became technical vocabulary, referring to the next generation of AI that can act with autonomy, not merely planning your itinerary but also booking your plane, hotel, and restaurant reservations and adapting to problems encountered along the way. In this new context, the human capacity for agency has come to encapsulate what some believe will mark the cleft, in an AI-suffused future, between winners and losers. Even after raw intelligence has been commoditized by smart machines, the role of choosing what to do will remain a scarce resource controlled by people. Those who’ve trained themselves to decide what matters and to pursue goals, the thinking goes, will have a stark edge over more passive people.

OpenAI CEO Sam Altman: “Agency, willfulness, and determination will likely be extremely valuable.”

Naval Ravikant, founder-investor: “If you’re a high-agency person, there’s never been a better time to be alive.”

Y Combinator CEO Garry Tan: “*agency* is now the most important trait to teach our kids and will be a mega multiplier on any given person’s life outcome.”

As “high agency” has emerged as body armor against ascendant AI, it has also become a character ideal that certain parents seek when raising children. “Agentic” kids are the savvy parent’s hedge against their offspring growing up to be dreaded NPCs — unquestioning accepters of the world’s default configurations — or “sheeple,” mindless herd followers. It’s an automation-era reinterpretation of what drove Jeff Bezos to give his children access to sharp knives at age 4 and power tools by age 8. (He has quoted his then-wife, MacKenzie Scott, as having said “I’d rather have a kid with nine fingers than a resourceless one.”) “The ability to develop knowledge combined with agency — if you have those, you’re probably going to do fine,” said Nevraumont.

But until its students come into their full agentic powers, Alpha is extremely focused on figuring out how to motivate them and, to that end, has launched separate schools that program their afternoons differently. At NextGen Academy, in a different part of Austin, the payoff for completing the morning’s work is an afternoon focused on video games with life skills smuggled in through the likes of team play and learning how to become a gaming YouTuber. (“It’s hard to get moms onboard,” an Alpha source acknowledged.) At Alpha Austin, completing checklist items, among other things, wins a student Alpha Bucks — one part of the school’s elaborate systems of gamified rewards. These can be spent on toys like a Labubu or a Pokémon box, and they can be lost by committing offenses such as “unkind language” or “non-constructive criticism” (a fine of 15 Alpha Bucks) and “bullying” or “hurting a peer due to recklessness” (a fine of 30).

Later that afternoon, the three dozen Level 2 kids sat in a circle for their weekly Town Hall, where they practice something like Athenian democracy. In advance of the meeting, any student may submit a proposal for consideration.

Many of the agenda items had to do with points of etiquette: Some kids had been spamming the Level 2 group chat with the nonsense meme “6 7”; someone had scratched words into the surface of a Zen booth. A lengthier discussion concerned “Italian brain rot,” AI-slop memes like a Nikes-wearing shark named Tralalero Tralala. Some boys had been running around yelling “Tung Tung Tung Sahur” — the name of another brain-rot character, a baseball-bat-wielding wooden cylinder with a face — and driving some girls crazy. After a student named Elaina described the yelling issue (she’d been the one who proposed it for the week’s agenda), other kids raised their hands.

“Brain rot is mocking God,” a boy offered earnestly.

“It’s just really annoying,” a girl said.

Then it was time for a secret vote on whether to ban the shouting of brain-rotisms except during lunch or recess. The children leaned forward, brought their foreheads to the ground, and closed their eyes. As a boy with a bowl cut read the options aloud, kids raised a hand behind their backs to signal their choice.

Bowl Cut appeared vaguely to register the vote count, then decreed that disruptive brain-rot babble “is banned, and so it is done. One, two, three,” and the circle of children, in unison, clapped once.

Alpha school presents one possible answer to the dilemma of raising an agentic child. Other proposals are more radical. In January, Ravikant, who, besides being a prominent Silicon Valley investor, is an X aphorist (“‘Consensus’ is just another way of saying ‘average’”) with 3 million followers, posted an Amazon link to a self-published book called The Sovereign Child, which he called “the single most important book on parenting that I’ve read.” It was certainly the first parenting book to jack directly into the limbic system of a type of very online techno-futurist who would never think of reading Bringing Up Bébé or Dr. Becky’s Good Inside. Over the next 48 hours, the author, Aaron Stupple, sold more than 18,000 copies of the book. Later, Ravikant joined Stupple on The Tim Ferriss Show for an episode in which Stupple talked about how he and his wife put no limits on their kids when it comes to food, sleep, and screens and even let them hover-board inside the house.

Stupple is a hospital doctor in Springfield, Massachusetts, but he was previously a high-school and middle-school teacher, briefly followed the Effective Altruists, and then became a techno optimist with a deep interest in the philosophy of quantum-computing-theory founder David Deutsch. Stupple’s book presents his contemporary gloss on a child-rearing philosophy called Taking Children Seriously, which Deutsch co-created. In The Sovereign Child, Stupple argues that enforcing rules creates four specific harms: an adversarial relationship with the parent, damage to a child’s self-confidence, confusion about why to do something (brush your teeth to prevent cavities and bad breath or because you’ll get yelled at if you don’t?), and confusion about how to solve problems on their own.

TCS, which had a moment in the individualist, post–Margaret Thatcher U.K. of the early 1990s, altogether bypasses the nub of most parenting debates: What is the ideal point on the permissive-to-authoritarian spectrum? Instead, it holds that children are autonomous moral beings who should be treated as adults’ equals. This means, in TCS’s purest form, eschewing coercion of any sort. It is based in part on the epistemology of Karl Popper, arguing that a child should acquire knowledge only through self-motivated hypothesis testing — in other words, they should learn by trial and error.

In Stupple’s view, TCS’s first-principles, epistemologically rigorous approach is the only sensible way to raise a child for a completely unknowable future. Saying a kid should study a particular canon of knowledge when you have no idea what the next decades will bring is as wrongheaded, in his view, as Thomas Malthus extrapolating disaster from population trends when he had no idea that crop genetics would obviate the famine he predicted. As Stupple sees it, he’s preparing his children to navigate whatever world arrives — and he’s very much on Team AI Optimism — rather than the world as we already know it. “Agency is everything in the Sovereign Child parenting framework,” Stupple told me. “I think the older view, our parents’ generation, was you did one thing your whole life. Now, you’re on your own, surfing. The old strategy would drown in the new world.”

Although Stupple and his wife aren’t dogmatic about coercion (they made the kids get vaccines), they’re absolutists when it comes to rules, dispensing with them entirely in favor of joint creative problem-solving by parent and child. TCS appealed to Stupple’s wife, an infectious-disease doctor, because it suited her intuition that children could regulate themselves when it comes to food and sleep. (It also offered a pragmatic way to manage having five children.) At first, they had doubts, fretting about how much Peppa Pig their elder daughter was watching and how late she was staying up to do so. When their son showed no interest in school or other children, they worried they were stunting him socially and educationally and cognitively by letting him stay home. They worried, like any parents, about nutrition. Switching to “unschooling” — homeschooling minus a curriculum — made everything easier. It removed scheduling pressure, which in turn removed the frictions of regulating sleep hours.

On a recent Saturday, I arrived for a visit at their home in a Rockwellian town in a part of Massachusetts dotted with American flags and tobacco-drying barns and memorials to native sons. Their firstborn daughter, who is 7, was in the yard cupping a toad in her hands, flanked by her 4-year-old sister and 6-year-old brother. They then surrounded their father and placed a stethoscope on his shaved head. Twin 1-and-a-half-year-old boys scooted around in the grass.

A contemporary parent of median uptightness who spent time with the Stupple family might experience sporadic wavelets of discomfort. There was the moment that day when the 4-year-old was standing perilously close to an extremely hot grill. Or the one when the 7-year-old, eyeing a spider, said, “Let’s burn him,” and proceeded to do so, dropping it on the 600-degree Weber lid and watching it sizzle. When she covered Stupple’s mouth with her hand, muting him mid-sentence, I thought he would pull it away, but he rolled with it for a surprisingly long time. At another point, the 4-, 6-, and 7-year-olds escaped the makeshift green plastic fencing Stupple had put up because they spied someone driving a backhoe next door and wanted to investigate, though this meant walking beside a well-trafficked country road. Later in the day, after we piled into a Honda Odyssey and drove to a roadside attraction where kids can mine fossils and pan a sandy streambed for precious items, one of the twins kept flinging clumps of wet sand or putting it down his shirt. At home, the same child opened the door of a below-counter kitchen cabinet, which any of the kids can access at their whim. Inside was a trove of chocolate-chip cookies, muffins, chocolate bars, mini-cupcakes, cinnamon rolls, popcorn, and honey buns. There was a footstool for the freezer, which was stocked with ice cream and Popsicles.

Rather than being laissez-faire, Sovereign Child–rearing is, in some ways, more labor- and attention-intensive than other styles of parenting. (It isn’t trivial that the Stupples have full-time coverage by caregivers who are familiar with homeschooling.) When the kids started walking toward the road, Stupple asked what they were doing in a tone of guileless curiosity, then followed them in a spirit of co-exploration. The spider’s demise was a prompt to discuss animal ethics: “Why do you want to fry the spider and not the toad?” And when the 4-year-old, pacifier in mouth, raised the lid of the flaming grill, Stupple was watching carefully. When he then coached her on turning the Weber’s burners down and how far to rotate each knob, it became an impromptu lesson in fractions. At another point, Stupple put his hand between the kids and the grill and, when they started playing near it, rolled it out of the way. “Saying, ‘Don’t go near the hot stove’ destroys all of it,” he said. “The richness of what there is to learn about the world is extraordinary.”

In the living room, the older kids had begun climbing on furniture, trampolining on a red sofa and clambering up a blue one. Then one of them had a better idea and started vaulting off the arm of the blue sofa onto the seat of the red one, sometimes with a heavy thud. “Can we do one jump?” Stupple tried. “One for everyone?” The kids laughed and ignored him. “How about a game of Trouble?” No buyers. They were happy to continue doing what they were doing.

Then one of the toddler twins showed up. “Oh boy,” Stupple said. He scanned the living-room area for hazards and moved a standing lamp out of the way, then rotated the blue sofa onto its side, making it harder to scale.

“Okay, so this guy’s trying to crawl up here and jump off,” Stupple said of the rogue twin. “It’s really cool that he’s trying to copy his siblings, and this is an independence thing.” Stupple quickly positioned some padded blocks and other wedges under the sofa so the toddler’s siblings couldn’t topple it over. Unable to climb it, the toddler instead played “fort” underneath, where his siblings soon joined him.

Critics say that Sovereign Child parenting risks — among many other things — raising children with major knowledge gaps and lack of boundaries and squandering a period of development when children are at peak neuroplasticity. Ravikant said he practices about 30 to 50 percent of Stupple’s philosophy in his own home but draws the line at certain foundational subjects, insisting his kids learn math and reading. “What is creativity?” Nevraumont asked. “The ability to make new connections between disparate things. That only works if you know what those disparate things are.”

Like every parent, Stupple wants his children to be happy and confident and to flourish in the world. He’s all for math and reading, if and when a child wants it, but he rejects the view that certain subjects are more important than others. “I think math is fetishized. It’s a sorting mechanism for showing people you’re smart,” he said. He thinks knowledge only really sticks if it comes organically. “My kids know a ton of stuff,” Stupple said. More, he believes, than they would have learned from conventional school by this age, and much of it, he says unabashedly, from YouTube. The oldest reads Dog Man and Diary of a Wimpy Kid and can handle basic percentages. The 4-year-old knows letters and numbers. The 6-year-old is “uninterested in reading and math,” but “he understands how the solar system works, what DNA is, what viruses are, all of the distinctive features of a surprising range of animals, how electricity works,” among other things. (At various times during my visit, he was immersed in videos about piranhas and bedbugs.) And what if the neuroplasticity window, Stupple rebutted, is when you develop your passions, your epistemology, your autonomy, your sense of humor? His daughter enjoys writing for writing’s sake: “She wants to write an essay about her vacation,” he told me.

Even though Julia Wise thinks it’s great that there are kids getting a flexible education, guided by a tutor, with lots of outside time and experiential learning, she and her husband send theirs to a regular public school. Homeschooling wouldn’t be a great fit for them, and the pandemic showed her that her children wouldn’t necessarily be happier staying home anyway. But she does expect more people, especially those with children who don’t like school, to start pulling them out if a normal-jobs-die-out future starts to become more visible.

As for her own kids, the prospect of a superintelligence-transformed future “makes me more prioritize them having a happy childhood,” Wise told me, “and doing things that are enjoyable over the medium term.” Her eldest daughter, who is in elementary school, is a good fiddler and has started going with her dad to folk-music gigs he plays. She has missed some Fridays without her academics suffering. Wise and her husband figure they will be less lax in middle school, but it’s an ongoing conversation. “We don’t have the goal of her being an academic hotshot, and she really enjoys these music events, and she’s developing a part of herself. If we end up in a position where humans don’t have to make money, playing folk music seems like a noble vocation.” And it’s the kind of archaic, handmade product or service that, along with high-touch skills like child care and plumbing, may still have value in a hardscrabble world where a human proletariat must scratch out a living. Wise says she has become less stressed about her three kids’ academic success.

“What I’m picking up more than anything is this chillness related to achievement culture,” echoes Rachael Horwitz, speaking of parents in her Silicon Valley circles. “It’s totally wide open now about what it’s going to take to lead a good life. There’s this weird loosening up. They are relaxing their shoulders.” The venture capitalist Chamath Palihapitiya recently said in an interview that his son told him he’s looking only at big SEC state schools for college because of the culture; two years earlier, Palihapitiya said, “I would have panicked,” but “now I’m like, ‘That’s actually the best thing you could do’ — be socially well adjusted, psychologically deal with different kinds of failures. You know?”

Wise, for her part, has “a fuzzy picture” of which qualities will be important, but when she listed traits like resilience, a flexible skill set, and creative problem-solving and things she believes will continue to have value in most future scenarios — physical activities like cooking and building and gardening and exploring; emotional skills like self-regulation, empathy, friendship, and conflict management; creative pursuits like music and writing and game design — she sounded quite a bit like the high-agency touters. Still, much of this, she pointed out, is stuff she already thought was important.

In her early-2024 essay, Wise wrote, “I’m trying to lean toward more grasshopper, less ant. Live like life might be short. More travel even when it means missing school, more hugs, more things that are fun for them.” A friend of hers suggested that “it’s good if children’s self-image isn’t too built around the idea of a career.”

Something that has shaped her approach in the face of so much uncertainty was watching her mother-in-law die of cancer. Wise had a newborn and was living with her husband’s parents. “She was going to be such an awesome grandmother,” Wise says. It drove home the fleeting preciousness of life. She likes what C. S. Lewis wrote in 1948 when global nuclear annihilation seemed like a real possibility: “If we are all going to be destroyed by an atomic bomb, let that bomb when it comes find us doing sensible and human things … not huddled together like frightened sheep and thinking about bombs.”

Edward Nevraumont, who is more sanguine about AI’s potential, believes even the most universally enriching version of an AI-ified world won’t mean the marginalization of humans. He gave the example of art history, the study of which was “knowledge for knowledge’s sake” but could be made essential in an era when we’re making art with machines. Only art history would give you the vocabulary, the context and connections, a sense of what he calls the “possibility space.” Plus, “if society is extremely wealthy, there will be a status battle, and there will be jobs to do that are status-related,” he said. AI is “an order of magnitude” better at chess than Magnus Carlsen, he points out, but no one shows up to watch AI chess engines play each other, and more people are playing chess than ever before.

From Intelligencer - Daily News, Politics, Business, and Tech via this RSS feed