Singularity and the Rain of Telephones

What three novels understood about the abundance we are already living inside

An artist’s rendition of a sci-fi singularity event. Image from Fate/Grand Order via FGO Fandom Community.

Charles Stross opens Singularity Sky with one of the great first paragraphs in science fiction. Telephones fall out of the sky onto the cobblestones of a colonial capital. Some of them are still half-melted from reentry. A pigeon pecks at one, gets startled when it beeps, and flaps off. A street kid picks another one up and hears a small voice ask whether he would like to entertain someone.

That is how the future arrives. Not with a manifesto or a mushroom cloud, but as loose hardware in the gutter that nobody has bothered to make illegal yet, because nobody in charge has understood what it does.

I have read three books that permanently rearranged how I think about what comes next: Stross’s Singularity Sky, published in 2003, Neal Stephenson’s The Diamond Age, published in 1995, and Arthur C. Clarke’s Childhood’s End, published in 1953. Half a century separates the oldest from the newest. They disagree about almost everything. But read together, in the right order, they stop looking like three competing predictions and start looking like three consecutive chapters of the same story, and we are somewhere in the middle of the second one right now.

The popular version of post-scarcity is a machine that makes stuff and a society that becomes nice. Build the replicator, cancel the hunger, cue the harp music. All three of these writers reject that, and they reject it in three completely different ways, because each of them is asking a sharper question: when the cost of making things falls to nothing, what is the new bottleneck?

Stross says the bottleneck becomes information, and that information is politically radioactive.

Stephenson says the bottleneck becomes access, and that whoever owns the pipe owns the future.

Clarke says the bottleneck becomes purpose, and that purpose is harder to manufacture than diamonds.

Here is the thing that makes this an optimistic essay rather than a gloomy one. Those three answers are not a warning that abundance fails. They are a description of what success actually feels like from the inside. Abundance does not end your problems. It promotes them. You trade the problem of calories for the problem of nutrition, the problem of nutrition for the problem of meaning, and every trade is a good trade. The reward for solving a hard problem is a better problem, and better problems are the entire point of civilization.

The question is not whether we get there. The question is whether you personally show up to the better problem prepared, or whether you show up holding a phone that fell out of the sky and no idea what story to tell it.

Part One: Stross, and the abundance that arrives as a rumor

The setup of Singularity Sky is beautifully cruel. The New Republic is a human interstellar polity that has deliberately frozen itself at roughly the technological level of Europe in the late nineteenth century. This is not an accident of decline. It is policy. The winning faction of an earlier civil war destroyed the last cornucopia machine on purpose and installed a feudal order with Victorian manners, on the reasoning that if nobody can make anything, everybody has a fixed place in society and the place stays fixed.

That is the crucial detail, and it is the one most summaries skip. The New Republic is not poor because the universe is stingy. It is poor because poverty is load-bearing. Scarcity is the foundation of the regime, and the regime knows it.

Then the Festival shows up over Rochard’s World. The Festival is a civilization of uploaded minds traveling as starwisps, and it does not want your land, your labor, or your women. It wants something much stranger. It drops phones and asks a single question. “Entertain us.” Tell it something it has not heard before, and it will give you whatever you ask for, up to and including its own manufacturing technology.

For a while the requests are modest. Food. A better roof. Then Burya Rubenstein, a political dissident serving internal exile on the colony, works out what the machine is actually offering. He does not ask for bread. He trades the Festival a political tract about how a sudden technological singularity would tear apart a repressive regime, and he asks for a cornucopia machine in return.

He wrote an essay predicting what a singularity would do to a regime like his, and he was paid for it with the machine that makes the prediction come true. Days later, the theory is the weather.

What Stross is doing here is not a gadget story. It is an argument about what matter fundamentally is. If matter is just organized information, then a machine that can reorganize matter turns every physical object into a file. And once an object is a file, every institution that was built on the difficulty of getting objects becomes a museum piece overnight. Land ownership only matters if land is the thing that grows the food. Aristocracy only matters if the aristocracy controls the forge. A monopoly on violence only matters if rifles are hard to make. When peasants can print rifles by telling a good story, the argument is over, and it was over before the shooting started.

The rulers respond exactly the way you would expect, which is to say, exactly the way rulers always do. They ban the phones. They send the fleet. They classify a gift as an invasion, because from where they sit, it is one. And their war plan violates the single rule laid down by the Eschaton, the godlike posthuman intelligence that scattered humanity across the galaxy in the first place. The Eschaton left its terms carved into the worlds it seeded: it is descended from us, it exists in our future, it is emphatically not our god, and nobody is to violate causality inside its light cone. Since the Festival’s technology cannot be beaten by any fleet arriving on time, the New Republic decides to arrive early, using faster-than-light travel to loop backward and hit Rochard’s World before the invasion it is responding to has happened. That is time travel. That is the one thing the Eschaton punishes at the scale of stars.

Stross gets three things right that have aged extraordinarily well.

The first is that abundance is disruptive in proportion to how suddenly it arrives, not how much it delivers. A slow rise in prosperity is absorbed. A vertical one is a solvent. The New Republic could have survived a century of gradual industrialization. It could not survive one street kid telling a good story into a phone he found in the gutter.

The second is his read on what becomes valuable afterward. When physical goods cost nothing, the currency is novelty. The Festival is rich beyond imagining and it is bored, and boredom is the only appetite it has left. It will trade you a functional post-scarcity economy for a story it has not heard. Twenty-two years after publication, that is not a metaphor. It is a business model. The entire attention economy is a slightly less honest version of the Festival, and the current generation of AI systems runs on almost exactly that bargain: enormous capability offered in exchange for human-generated novelty, at a price that keeps falling.

And the third thing, the one nobody quotes, is the ending. The Festival leaves.

It is a migratory phenomenon. It arrives, it consumes the local novelty, it strips useful mass out of the system, and it moves on, leaving Rochard’s World transformed, littered, and permanently unable to go back to what it was. Some people have transcended. Some are cyborgs. The revolution Rubenstein started has grown past needing him or any other leader, which is the most quietly devastating line in the book for anyone who has ever wanted to be the one steering.

Abundance that arrives as a gift from outside is a weather event, not an economy. When it moves on, you are left with whatever you learned to build while it was overhead. That is the single most practical lesson in the novel, and we will come back to it, because a great deal of what is being offered to us right now arrives as a gift from outside, delivered by entities whose long-term plans are not published.

Part Two: Stephenson, and the abundance that turns out to be a plumbing question

If Stross writes the shock, Stephenson writes the morning after, and he writes it with the weary competence of a man who has read the terms of service.

The Diamond Age takes place in a world where the replicator problem is solved and filed away. Matter compilers are everywhere. They take atoms from a shared utility called the Feed and assemble them into whatever you have the rights to print. Food, clothing, a mattress, a bicycle. Nobody starves. The basic goods of life are, functionally, free.

And the world is still savagely unequal, because Stephenson’s central insight is the one that separates the adults from the children in this conversation: material scarcity and social scarcity are different problems, and solving the first does nothing whatsoever to the second. Humans do not want things. Humans want standing. If you give everyone infinite things, they will promptly invent a new scoreboard, and they will build it fast.

The nation-state has dissolved into phyles, which are tribes you opt into: the Neo-Victorians, the Han, the Nipponese, the Boers, plus a long tail of smaller claves and the great mass of thetes who belong to nobody. Membership is the real asset. And underneath the phyles sits the actual architecture of power, which is not a police force or an army. It is the Feed.

The Feed is top-down. It flows from centralized Sources, controlled by the phyles at the top of the pile, out through a hierarchy of pipes to every compiler in the world. It is generous. It is reliable. And it is a leash, because whoever operates the Source decides what may be compiled, by whom, and under what license. This is not a system that starves you. It is a system that grants you.

Against it stands the Seed, the ambition of the underground, championed by Dr. X and the CryptNet phyle: a technology that would let matter compilation propagate biologically, gathering raw material locally, growing rather than receiving, with no Source and no permission layer at all. The Feed is a subscription. The Seed is a garden. The entire political struggle of the novel is between those two shapes.

You do not need me to draw the diagram. In 2026 we are living in the Feed versus Seed fight, and we are living in it on a weekly release cycle.

On one side is the frontier, delivered by API. It is genuinely astonishing, it improves faster than anything else in the economy, and it is metered, moderated, versioned, and revocable. On the other side is the open-weight ecosystem, which has spent three years doing something that looked impossible when it started. In 2023 the question was whether open models could stay in the same conversation as the frontier. By 2026 the serious debate is not whether to use open weights but which family fits which workload, with several open-weight lines shipping frontier-class capability and the remaining performance gap now smaller than the price gap.

The licensing story is even more Stephensonian, because licenses are exactly how the Feed exerts control. Apache 2.0 and MIT have become the permissive default across several major open families. Meta’s Llama line keeps a custom license with a monthly-active-user clause that has been a real obstacle for large deployments and has created friction for users in the EU. Some weights are visible but commercially fenced. The weights download freely and the rights do not, which is the Feed’s whole trick rendered in legalese.

And then there was the demonstration nobody planned. In June 2026, a United States export-control order pulled Anthropic’s most capable models offline for several weeks. Anthropic restored access on July 1 after the Department of Commerce lifted the controls. For that window, work that depended on those endpoints stopped, and work that ran on hardware people owned did not. That is the Feed and the Seed, staged as a live fire drill, on a timescale of weeks rather than a novel’s worth of years. The lesson is not that centralized frontier models are bad. They are extraordinary, and for most tasks they are the right tool. The lesson is that a capability you rent is a capability someone else can pause.

But Stephenson’s deepest idea in the book is not about infrastructure at all. It is about what happens to human work when compilation is free, and this is where the novel turns unexpectedly warm.

When any physical object can be printed on demand, the printed object becomes worthless as a signal. A compiled diamond is a paperweight. So value stampedes toward the things that cannot be compiled: hand craft, taste, judgment, discipline, the visible evidence of human hours. The Neo-Victorians do not build their aesthetic out of nostalgia. They build it out of economics. When a compiler can produce a flawless copy of any object on demand, the flawless copy stops being worth anything, and the only genuinely scarce ingredient left in a physical thing is the human hours visibly poured into it. A hand-carved table is a luxury not because the machine was incapable, but because the machine was refused.

And the beating heart of the story is Nell, a girl from the bottom of the social order who comes into possession of a stolen copy of the Young Lady’s Illustrated Primer, an interactive book engineered to raise a subversive, self-reliant, formidable adult. The Primer works. Nell ends up leading an army.

Here is the part that matters. The Primer does not work alone. Its stories are performed in real time by Miranda, a ractor, a human actor who takes interactive contracts out of a theater in the Leased Territories. Miranda picks up Nell’s sessions, keeps picking them up, and comes to love a child she has never met and cannot identify. Elsewhere in the book, an army of orphan girls is raised on Primers with no dedicated human on the other end, and they turn out capable, coordinated, and eerie. They are missing something, and what they are missing is that somebody chose them.

Stephenson, in 1995, sat down and wrote out the answer to the question everybody is asking right now about AI tutors, AI companions, and AI therapists. The technology is real, the technology is transformative, and the technology’s ceiling is set by whether a human being is willing to be on the other end of it. The compute was never the scarce ingredient. The caring was.

If you are looking for a place to stand in the next twenty years, that is the ground. Not competing with the compiler. Being the voice the compiler cannot fake.

Part Three: Clarke, and the abundance that runs out of things to want

Then there is Childhood’s End, which is seventy-three years old and still the hardest of the three to shake off.

The Overlords arrive, park their ships over every major city, and end everything we complain about. War stops. Poverty stops. Cruelty to animals stops. Disease and want recede. They govern with an almost bored competence, through a supervisor called Karellen who speaks to humanity through the Secretary-General and refuses to show his face for fifty years. What follows is called the Golden Age, and Clarke means the term to cut both ways: gold is the color of triumph and also the color of late autumn, right before the leaves come down.

And in the Golden Age, according to the standard reading, humanity goes soft. Science drifts, because the Overlords already know the answers and the thrill of discovery has been priced out of the market. Art drifts, because comfort makes for thin material. Entertainment becomes plentiful and consumption becomes the default posture. People are pleasant, safe, well-fed, and vaguely unemployed at the level of the soul. A colony called New Athens is founded by artists and scientists specifically to fight the drift, an intentional community engineered to make great work happen again.

This is the passage everyone cites when they want to argue that post-scarcity would domesticate us. It is the intellectual ancestor of every warning about people becoming pets. And I want to make the case that the standard reading gets Clarke almost exactly backwards.

Because look at what actually causes the stagnation in that book. It is not the free food. It is the ceiling.

The Overlords do not merely provide. They close the frontier. Their signature intervention, the one that breaks Rikki Stormgren and defines the whole novel, is the flat verdict that space is not available to us. “The stars are not for Man.” Humanity is welcome to be comfortable, healthy, peaceful, and long-lived, on the condition that it stays in the yard. There is a hard boundary on human ambition, it is enforced by an unanswerable authority, and everyone knows it.

Of course the culture goes flat. It is not decadent. It is supervised. A society that has been informed on excellent authority that the biggest thing it could ever attempt is permanently off the table does not need free lunch to lose its nerve. The free lunch is a coincidence. The ceiling is the cause.

And Clarke, whether or not he meant to, wrote the counterexample right into the book. Jan Rodricks refuses the ceiling. He is a young scientist in a world that has told him there is nothing left to reach for, and rather than accept it, he executes an audacious, borderline insane plan to stow away aboard an Overlord vessel and see the stars anyway. He is the one character who does not take the deal. He gets a berth on the greatest voyage any human has ever made, and he ends up as the eyewitness to the whole species’ transformation, narrating the last chapter of Earth.

Abundance did not domesticate Jan. Abundance is what freed him to attempt something absurd. He had no rent to make.

Look at New Athens the same way. It is founded as therapy. Its stated purpose is to counteract cultural decline, which means every painting made there is a lecture about the importance of making paintings. That is not how art happens. Art happens when somebody has something to say and enough slack to say it, not when a committee identifies a deficit in the national soul. New Athens is not evidence that comfortable people cannot create. It is evidence that you cannot schedule inspiration as a public health measure.

And then there is the tell. Clarke put a line on the copyright page disclaiming that the opinions in the book were his own. He explained decades later that he had just published an optimistic account of humanity’s expansion into space, and had then written a novel arguing that the stars were closed to us, and he did not want anyone concluding he had changed his mind.

The bleakest reading of Childhood’s End is a reading its own author declined to endorse.

So take the real warning, which is sharper and more useful than the pop version. The danger is not that you will be given too much. The danger is that you will be given a lot, on terms set by someone else, in exchange for accepting a limit on what you are allowed to attempt. That is the trade the Overlords offer. It is a superb deal on comfort and a catastrophic deal on horizon.

The correct response to that offer is not to refuse the comfort. It is to refuse the ceiling.

Part Four: The prophets are all writing fan fiction of these three books

Something worth noticing about the current wave of abundance forecasting is that almost none of it is new. The people building the systems have independently rediscovered Stross, Stephenson, and Clarke, and have sorted themselves neatly into the three camps.

Ray Kurzweil is writing Clarke. In The Singularity Is Nearer, published in June 2024, he holds the line he has held since the 1990s: artificial general intelligence by 2029, and the Singularity itself around 2045, which he defines not as machines beating us but as brain-computer interfaces and eventually nanobots letting human intelligence expand a millionfold by merging with the cloud. He expects nanotechnology and materials science to make energy effectively renewable and abundant, medicine to reverse aging, and the boundary between biological and non-biological cognition to become a matter of preference.

Strip the engineering vocabulary out of that and you have the ending of Childhood’s End with better spreadsheets. A generation of humans transcends its physical limits, merges into something larger, and becomes unrecognizable to whatever it used to be. The difference is one of authorship. Clarke’s Overmind harvests humanity; the children go and the parents stay behind on a dying planet with nothing left to do. Kurzweil’s version is opt-in and self-directed, more like an upgrade than a rapture. Both are describing the same doorway. They disagree about who is holding it open.

Elon Musk is writing Stross. His forecast is the cornucopia machine dropped from orbit, minus the requirement that you entertain anybody. He has said, repeatedly and with escalating confidence, that AI and robotics will make employment optional on a horizon of roughly ten to twenty years, that any job you do will be a hobby, and that the result will be what he calls universal high income rather than universal basic income. At Viva Technology in Paris in 2024 he described a best case in which essentially nobody has a job because there is no shortage of goods or services. More recently he has argued that money itself stops being relevant, that saving for the future becomes pointless because there will be no poverty to save against, and that federal checks funded by AI-driven productivity are the right answer to AI-driven unemployment. His economic reasoning is that if robots drive the cost of goods down toward the cost of electricity and materials, prices fall faster than the money supply grows, so large transfers do not have to be inflationary.

That is the Festival’s offer stated as public policy. And the objections it draws are exactly the objections Stross dramatized. Michael Burry’s public response was that there would be a revolution first, which is a compressed summary of the entire plot of Singularity Sky: the arrival of abundance does not calmly replace the old order, it detonates it, and the people currently holding power have both the motive and the fleet to fight the transition. Jeff Bezos has taken a different line, arguing that AI is more likely to produce a labor shortage than mass unemployment, because it frees people to build things that do not exist yet. Both of those are Rochard’s World. Burry is describing the New Republic’s battle fleet. Bezos is describing the citizens who came out of the chaos as something new.

Sam Altman is writing Stephenson, and doing it deliberately. His June 2025 essay “The Gentle Singularity” argues that the takeoff has already started and that it does not feel like a rupture from the inside, because wonders become routine and then become table stakes. He sketches a dated forecast: agents doing real cognitive work in 2025, systems producing novel insights in 2026, robots doing physical tasks in 2027, and intelligence and energy becoming wildly abundant in the 2030s, at which point a decade of research might be compressible into a year. His signature phrase is that intelligence too cheap to meter is within reach, and that the cost of intelligence should converge toward the cost of electricity.

But the load-bearing sentence in that essay is not about capability at all. It is the insistence that once the safety problems are handled, access to superintelligence has to be widely distributed rather than concentrated. That is a man arguing for the Seed while running a Source. His observation that humans are wired to care about other people and not much about machines is, almost word for word, the Miranda principle.

Dario Amodei is writing all three at once. Machines of Loving Grace, published in October 2024, defines powerful AI as a system smarter than a Nobel laureate across most fields, able to work autonomously for long stretches, running as many parallel copies as compute allows at ten to a hundred times human speed. He compresses that image into a phrase that has entered the language: “a country of geniuses in a datacenter.” His central prediction is about time rather than intelligence. He argues such systems could compress fifty to a hundred years of biological and medical progress into five or ten, a compressed twenty-first century, delivering the end of most infectious disease, the elimination of most cancer, effective treatment for genetic disease, and a plausible doubling of human lifespan.

And then the essay does something the others mostly do not. It devotes its final section to work and meaning, and it visibly struggles there. It is much easier to describe a world without malaria than a world in which nothing you do is the best available version of doing it. Critics have pointed at that section as the essay’s weak point. I would say it is the essay’s most honest moment: the author walks his own argument all the way to the door of Clarke’s chapter eight, looks at what is on the other side, and reports back that this part is genuinely unsolved. Which is correct. It is unsolved. It has been unsolved since 1953, and it is the most interesting job opening in the world.

And Ezra Klein and Derek Thompson are writing the footnote that makes Stephenson actionable. Abundance, published in March 2025, is not about nanotechnology at all. It is about the ordinary, non-science-fictional observation that a great deal of the shortage in modern life is chosen. Housing is scarce because we decided not to build it. Clean energy is slow because we made it slow to permit. The bottleneck is not physics and it is not resources. It is process, veto points, and the accumulated preference of incumbents for a world that does not change.

That is the Feed, wearing a municipal zoning code instead of a nanotech license. And the reason it belongs in this essay is that it reframes the whole conversation from prophecy to engineering. If scarcity is partly manufactured, then abundance is partly a decision, and decisions can be made this year, by people who are not billionaires, in rooms that are open to the public. You do not have to wait for the Festival. Some of the phones are already on the ground.

Part Five: Are we already there?

Partly. And the parts where we have arrived are more instructive than the parts we are waiting on.

Calories are effectively abundant in the developed world, and the resulting crisis was not laziness. It was a public health problem of a completely new kind, which we are now, notably, solving with a class of drugs that did not exist a decade ago. Entertainment is beyond abundant; there is more high-quality film, music, writing, and game design available tonight than a person could consume in ten lifetimes, and the constraint moved from availability to attention. Information is so abundant that the scarce goods became curation, trust, and the ability to tell whether a thing is true. Expertise is the one currently in freefall. The cost of a competent second opinion on your code, your contract, your lab results, or your business plan has fallen further and faster in the last four years than in the previous four hundred.

So: does it domesticate us?

The evidence so far says no, and it says no loudly.

Computers have been better at chess than any living human for a generation. Chess was supposed to die. Instead more people play it now than at any point in its history, and the engines became training partners rather than executioners. Machines have been able to move us faster than our legs since the nineteenth century, and marathon participation exploded anyway. Bread has been industrially abundant and cheap for a century, and people still get up early to feed a sourdough starter they named. Every single one of those activities is economically pointless. That is why people do them. Difficulty that you choose is not the same category of thing as difficulty that is imposed on you, and human beings appear to have a bottomless appetite for the first kind.

The domesticated-consumer scenario is real, but it is not caused by having enough. It is caused by a specific arrangement: abundance delivered passively, on a schedule set by someone whose incentives are not yours, in a form that asks nothing of you.

Which brings back the detail from Stross that I cannot stop thinking about. The Festival’s bargain was extractive and alien and it upended a civilization, but it demanded something. It wanted a story it had never heard. It made you interesting in order to pay you. The infinite scroll demands nothing at all. It will run whether you are awake, engaged, or drooling. Measured on the single axis of whether the technology requires you to become more of a person to use it, a hostile alien intelligence from a 2003 space opera is a better deal than most of what is on your phone right now.

That is the actual risk. Not plenty. Passivity. And it is a risk you can personally opt out of, starting immediately, without anyone’s permission.

Part Six: How to be ready

None of what follows requires you to predict a date. All of it pays off whether the timeline is Kurzweil’s or Amodei’s or twice as long as either.

Keep a Seed, not just a Feed. Rent the frontier, because the frontier is worth renting; it is the best tool available and it is getting better on a schedule that is difficult to overstate. But own something. Run a capable open-weight model on your own hardware and learn what it takes to keep it running. Keep local copies of the work that matters to you. Understand at least one layer beneath the interface you use every day. This is not doomsday prep, it is the same instinct that makes a chef keep her own knives. The point is not that the Feed will fail. The point is that a person who can grow food thinks differently about the grocery store than a person who cannot, and that difference in posture shows up in every negotiation they ever have.

Become the ractor. Whatever your field, identify the part of it that only works because a human is doing it, and move toward that part deliberately. It is almost never the production of the artifact. It is the judgment about which artifact should exist, the taste to know when it is wrong, the accountability of a named person who can be held responsible, and the relationship with whoever is on the other end. Nell got an army out of the Primer. The Chinese orphans got a very good education. The difference was Miranda. As the artifact gets cheaper, the person gets more valuable, and that trend has been running in one direction for three years with no sign of reversing.

Practice wanting things. In a world of constrained options, your preferences are mostly formed by what was available. In a world of unconstrained options, taste is the whole ballgame, and taste is a muscle that atrophies when it is not used. Most people have never had to answer the question of what they would actually do with an open afternoon, an open budget, and a capable assistant, because they have never had all three at once. Start answering it now, in small ways, while the stakes are low. The people who will struggle most in an abundant world are not the ones who lack resources. They are the ones who get handed the resources and discover they have no idea what they were for.

Keep one hard thing on purpose. Something with a real standard that does not care about your feelings. An instrument, a language, a barbell, a boat, a proof. This is the vaccine against the Golden Age, and it works because the difficulty is voluntary and therefore cannot be taken away from you by a machine that does it better. Nobody plays chess because the world needs another chess game. Somewhere in your life there should be an activity whose only justification is that doing it makes you more.

Refuse the ceiling, and vote against ceilings. This is the political half, and it is the half that most abundance writing skips. The Overlord deal is comfort in exchange for a limit on ambition, and it is offered constantly, in small denominations. It looks like a licensing term that says you may use this but not build with it. It looks like a permitting process that makes the ambitious version of a project impossible and the timid version merely expensive. It looks like an argument that some capability is too powerful for ordinary people to hold. Some of those arguments are correct and serious. Many of them are the New Republic banning telephones. Learn to tell the difference, because that distinction is going to be the central political question of the next two decades, and it will be decided by people who show up.

Do not wait for the Festival, and do not assume it stays. Rochard’s World got its miracle from the sky and then watched it leave. What survived was not the gift; it was the capability the population built while the gift was overhead, and the people who used the window to become something that did not need the window. Right now there is an enormous, temporary, historically strange window in which extraordinary capability is available at nearly no cost to anyone with a laptop and curiosity. It is being subsidized by a capital expenditure race that will not run at this intensity forever. Use it to build something durable, learn something structural, start the thing you have been putting off. Not because the window is closing tomorrow. Because windows close.

The last paragraph, which is the first paragraph

Three writers, working decades apart, looked at a world where making things is free, and each of them found a different thing underneath.

Stross found politics, and told us that the arrival will be sudden and contested and that the people who benefit most will be the ones who understood the offer first.

Stephenson found plumbing, and told us that the fight will be about who controls the pipe, and that the thing which stays priceless is a human being who chose you.

Clarke found the void at the end of comfort, and warned us, in a book he half disowned, that a species with everything and nowhere to go stops being interesting to itself. And then he wrote a character who refused the deal and put him on the last ship out.

Put them together and you get something better than any of them alone, and considerably more hopeful. Abundance is coming, in pieces, unevenly, faster in some domains than anyone expected and slower in others. It will not make us peaceful and it will not make us happy, because those were never material problems. What it will do is hand us a class of problems that our ancestors would have wept to have: what to build, who to love, what to attempt, what to become. Those are the good problems. Those are the ones worth having.

The telephones are already falling. They have been falling for about four years now, and they are landing in the gutter, and most people are walking past them.

The voice on the line is asking for a story.

Have one ready.

Sources and further reading

The novels

• Charles Stross, Singularity Sky (Ace, 2003). Nominated for the Hugo Award for Best Novel in 2004. Publisher page
• Neal Stephenson, The Diamond Age: Or, A Young Lady’s Illustrated Primer (Bantam Spectra, 1995). Winner of the Hugo Award for Best Novel and the Locus Award for Best Science Fiction Novel, 1996. Publisher page
• Arthur C. Clarke, Childhood’s End (Ballantine, 1953). Developed from the 1950 novella “Guardian Angel”; the copyright page of most editions carries Clarke’s disclaimer regarding the book’s opinions. Publisher page

Contemporary forecasts

• Sam Altman, “The Gentle Singularity,” June 10, 2025. blog.samaltman.com
• Sam Altman, “Moore’s Law for Everything,” March 2021. moores.samaltman.com
• Dario Amodei, “Machines of Loving Grace,” October 2024. darioamodei.com
• Dario Amodei, “The Adolescence of Technology.” darioamodei.com
• Ray Kurzweil, The Singularity Is Nearer: When We Merge with AI (Viking, June 25, 2024). Publisher page
• Ezra Klein and Derek Thompson, Abundance (Avid Reader Press, March 18, 2025). Publisher page

Reporting on the Musk predictions

• Fortune, on Musk’s ten-to-twenty-year timeline for optional work and the irrelevance of money. fortune.com
• Fox Business, on Musk’s universal high income remarks and the exchange with Ray Dalio. foxbusiness.com
• Business Insider via Yahoo Finance, on Michael Burry’s response and the wider billionaire debate, including Jeff Bezos on labor shortage. finance.yahoo.com

The open-weight landscape

• “The State of Open-Weight Models in 2026,” May 2026, on license classes and the narrowing capability gap. jamesm.blog
• “Open-Weight LLM Showdown 2026,” July 2026, on family-by-family comparison and licensing. wavect.io
• Anthropic’s statement on the June 2026 export-control suspension and the July 1, 2026 restoration of access. anthropic.com