Seam №001 · Essay · July 13, 2026 · 13 minInteractive experience added July 13, 2026
You Can Just Build Things
On thinking rocks, forbidden skies, and the exhilarating responsibility of inheriting civilization.
scroll ↓
There is a dangerous way to get used to a miracle.
You put it in your pocket.
A black rectangle wakes under your thumb. Nothing explodes. No choir appears. Billions of microscopic switches agree to become your camera, map, library, studio, translator, calendar, and window into nearly every person you have ever loved.
We call this normal.
This morning, normal is the first thing worth distrusting.
Over the past week a few things snapped together in my head like magnets. A DeepMind documentary called The Thinking Game. A Marques Brownlee video that makes the scale of Apple’s M5 chip feel bodily real. Dr. STONE, which rebuilds civilization from nothing. And Orb: On the Movements of the Earth, about what people will risk to carry an idea past their own death.
They should not fit together. One goes inside a chip, one inside an AI laboratory, one begins after the end of the world, and one looks up at a fifteenth-century sky.
They landed on the same feeling anyway. Almost everything good around me is held up by a stack beneath it and a chain behind it.
The stack is what had to exist first: materials, tools, measurements, skills, institutions, agreements, and people doing hard work carefully. The chain is what had to survive long enough to reach me. Observations, diagrams, failed experiments, corrections, teachers, copies, courage.
Take the stack and the chain seriously and inheritance turns into a responsibility, and then, oddly, into a relief.
There is a slogan builders pass around: you can just build things. One honest look at a phone and the slogan collapses. Nothing was ever just built. It holds anyway, for one reason. The world was built. So you can build.
Civilization is the stack and the chain.
You inherited it. You are also a link in it.
01
The seam beneath the miracle
Start with the phone on the desk in front of you. The chip inside it was made on what the industry calls a 3-nanometer process.1 My brain can say that. My brain cannot feel it.
So I steal a trick from Marques Brownlee and scale the thing until my intuition has something to grab.
Shrink to the height of a keyboard key and a laptop is a skyline. Keep going, down past the width of a red blood cell, and a hair is a cliff, a raindrop is a lake, and light waves run about as long as you are tall. At the bottom of the ladder sits the transistor. If one were the size of a Toyota Prius, the phone holding 19 billion of them would be about the size of Earth.
Now watch one get made. Chips are printed, hundreds at a time, onto a wafer: a mirror-smooth disc of silicon about the width of a dinner plate, sliced from a single grown crystal. The patterns go on in layers, and at the end the wafer is cut apart into individual chips.
The printing is done by lithography machines, and the ones that handle the hardest layers come from a single company. ASML, in the Netherlands, is the only commercial supplier of extreme ultraviolet lithography systems. An EUV machine costs hundreds of millions of dollars. The newest High-NA generation runs to about $380 million and weighs as much as a hundred cars. The first installation took 250 engineers about six months.2
One company. On the whole planet. For the hardest layers in the most advanced chips, the road runs through a machine that leaves a warehouse in the Netherlands in 250 boxes.
Design to mass production takes hundreds of steps and up to four months, with as many as a hundred layers, each aligned to the others at nanometer precision.3
The machine makes its light by dropping molten tin. It shoots the falling droplet with a laser twice, once to flatten it, once to vaporise it. The tin becomes a plasma, and the plasma throws off extreme ultraviolet light at a wavelength of 13.5 nanometers. It does this up to 50,000 times a second. Glass would swallow light like that, so the whole optical path is mirrors, in a vacuum. Underneath it all, a stage holds the wafer in place to within a quarter of a nanometer, correcting itself 20,000 times a second while the exposure is happening.4
Somebody built that. On purpose. And then they built more of them.
Fifty thousand droplets a second.
Every single one of them gets hit.
None of that is a laboratory demonstration. It runs on a production line, every day.
A chip is a treaty between strangers. Almost nobody who uses one could build one, almost nobody who helped build it understands every layer, and the layers meet anyway.
In 1965 Gordon Moore had about five points on a graph. They showed the number of components on a chip roughly doubling every year, so he drew the line ten years forward and predicted that by 1975 you would fit 65,000 of them on a single chip. From five dots. He was right, and in 1975 he moved the rate to every two years, and the industry took a man’s guess and made a metronome out of it, and then a promise modernity made to itself: the impossible would arrive on schedule.5
Sixty years later, here is what the schedule delivered. The first general-purpose electronic computers ran on vacuum tubes and filled rooms. ENIAC used about 18,000 of them.6 Apple’s A17 Pro holds 19 billion transistors.7 Nineteen billion. In something you leave face down on a table.
I did not build any of this. I can barely explain it.
I am an inheritor, and so are you.
The tolerances. The standards. The night shifts. The prototypes that failed and the people who came back the next morning anyway. We inherit the finished verbs:call, render, search, translate, and forget the thousands of nouns underneath.
When the seams disappear, gratitude goes with them, and so does agency. The world starts to feel like a set of sealed products delivered by forces beyond you.
The world isn’t sealed.
It has seams everywhere.
02
Civilization, run backward
Dr. STONE, a series by Riichiro Inagaki and Boichi, opens with a flash of green light that turns every human on Earth to stone. Thousands of years later two boys wake up, Senku and Taiju, in a world where the cities have crumbled and the forests have come back. Everyone else is still out there, standing exactly where they were, as statues in the grass. So the first thing Senku has to invent is the fluid that brings a person back.8
It is an adventure story where the enemy is entropy and the fights are chemistry. Senku is a young prodigy, and he wants to make sulfa drugs, so he has to get to sulfuric acid, so he has to find the right rock. He wants a phone line, so first there has to be copper, and before copper there has to be a way to smelt it, and before that a furnace, and before that charcoal, and someone strong enough to work the bellows. Slowly but surely, he builds the tech tree of the human race back up out of the mud, and I could not stop reading.
Senku can just build things. He is also drawn. He has perfect recall of the entire scientific corpus, infinite patience, and a plot that wants him to succeed.
The rest of us cannot just make a light bulb. You need glass of sufficient quality and a furnace hot enough to work it. You need tungsten drawn into a wire thinner than a hair, which means you also need metallurgy. You need a filament that survives being white hot over and over, a vacuum sealed inside the bulb, a power source, a grid to carry the power to the room, tools that survive the heat, instruments to tell you what you are doing, and people who can make every one of those parts again tomorrow. Every prerequisite opens into more prerequisites.
Senku’s superpower is dependency. He looks at a finished thing and sees the road running backward from it. What has to be true before this can be true? What can we make with what we already have? Which missing piece unlocks the most, and who knows something I do not?
The series hands you a genius and spends the next hundred chapters proving he is useless alone. Knowledge does not hammer metal. A formula does not carry lumber. The Kingdom of Science works because it stops being one boy and becomes a crowd of people who are each good at something else.
You do not need the whole stack before you can add to it.
You need the next dependency.
The first version of anything can be embarrassingly small. A rough script, a crooked shelf, a paragraph that finally says what you meant, a process written down before it disappears. Small work is the unit civilization is made of.
Once you hold the formula that brings the petrified back, you meet the first political question that ever mattered.
The early conflict is a fork. One road revives everyone and accepts the mess of being human. The other keeps the strong and lets power be the filter.
Both roads run on the same chemistry. Which one you walk is not a chemistry question.
That is why Dr. STONE sits so well beside the argument we are having about AI. It keeps technology entangled with governance, where it lives.
Are you using your tools to narrow the world, or to widen it?
03
A fair way to be wrong
The Thinking Game, a documentary directed by Greg Kohs, follows Demis Hassabis from chess and game design through neuroscience to the founding of DeepMind. Its subject is a research culture pointing learning systems at scientific problems, above all protein structure prediction with AlphaFold.9
A protein starts life as a chain, a long string of amino acids in a known order. Then it folds itself into a shape, and the shape is what does the work. Get the shape and you understand the machine. Work out that shape in a laboratory and it can take you months, sometimes years, for one protein.
DeepMind went at it twice: a first system in 2018, then a rebuilt one two years later.
CASP is a competition rigged so you cannot cheat. Research groups get the amino acid sequences of proteins whose real structures have been solved but not published. The answers exist but they are locked in a drawer. You submit your predictions and independent assessors score you against reality. In CASP #14, close to a hundred groups submitted more than 67,000 models across ninety targets.10
The rebuilt system scored a median of 92.4.11 Its predictions came back so close to the real shapes that the assessors began comparing them against the experiments themselves, and it finished miles clear of every other group. It did not put the laboratories out of work. The paper is candid about the limits: accuracy falls away when there are few related sequences to learn from, and when a protein’s shape depends on the other chains it is pressed against.12
And then they gave it away. Imagine how easily they could have done the other thing. Keep it in-house, license it to the big pharmaceutical companies, turn biology into a subscription, and nobody would have raised an eyebrow.
DeepMind and EMBL-EBI put the system and the database out in the open instead, and by 2022 it held more than 200 million predicted structures, free to anyone doing research.13 Researchers without the equipment or time to determine a particular structure experimentally now had immediate access to a useful prediction, with confidence estimates and limits they still had to understand.
It doesn’t make DeepMind a saint and it settles nothing about AI. But a door did open, and a whole class of questions got cheaper to ask, for far more people. A capability inside one laboratory is an achievement. A capability the rest of us can inspect, test, attack and build on is the one that turns into civilization.14
04
The protagonist is transmission
Orb: On the Movements of the Earth, a series by Uoto, is historical fiction. In its imagined fifteenth-century Europe, saying that the Earth moves around the Sun will get you tortured and burned. So the people who work it out do it in secret, and when the church comes for them, they rush to hide their notes somewhere a stranger might find them.15
The story keeps changing main characters. Each time, the previous one has met their fate at the hands of the church’s inquisitors. The notes keep moving, the people carrying them keep dying, and I have never read anything with this much nerve.
Characters arrive, meet a question larger than themselves, change it, and hand it on.
The handoff is the movement that had me gripping the book. An idea goes from a dying man to a stranger, and nothing about it is guaranteed.
A truth can be correct and still vanish. A discovery can die with the person who made it. An insight nobody can copy, teach, translate, or recover is a private revelation and nothing more.
Knowledge has to be made survivable. It needs vessels: diagrams, books, code, demonstrations, archives, teachers, institutions, and people willing to carry a fragile thing through being misunderstood. Sometimes courage is the discovery. Sometimes courage is making a copy.
Orb refuses to make the persecution polite. It shows you the machinery. Being right is not safety. Being sincere is not safety. In a system like that the only safe move is silence, which is the entire purpose of the system.
Ours is softer and more “efficient”. It rarely needs a prison. It can drown you instead, flooding your attention until you lose the telescope you were born with.
Whoever holds your attention shapes what you can notice, and therefore what you can believe.
If you do not protect your attention, someone else will spend it for you.
This is the chain behind everything I know. No scientist starts from silence, no engineer starts from raw nature, no writer starts before language. We arrive in the middle of a conversation whose first speakers we cannot name.
Dr. STONE runs civilization backward, from an object to its prerequisites. Orb runs knowledge forward, from one person to the next. One asks what had to exist before this. The other asks who carries it after me.
Between those two questions sits nearly everything worth building.
05
Accurate awe
Everything I have just told you, I told you like a fan.
A fan’s mistake is to treat cleverness as innocence: to see a beautiful machine and assume the system around it is good, to see a breakthrough and assume the benefits will be shared, to see capability and call it wisdom.
Making something extraordinary tells you nothing on its own about who bore the cost, who shared the reward, who consented, or whether the thing should be used every way it can be used.
A chip can be a monument to global cooperation across supply chains. An open scientific database can be a real public good while the company behind it stays powerful enough to deserve suspicion. A tool can widen participation and widen the reach of carelessness.
You can love the thing and still ask who pays for it.
06
The tools that speak
In August 2025 I wrote an essay asking whether a piece of writing was made with AI. Its spine still holds: use the tools and stay answerable for what they make. You can ask a person what they meant and what they will do about the consequences. You cannot ask that of a system.
Since then, content got cheaper to make and the internet is filling with slop, fluent output with no care behind it. Calling it ugly would be a compliment. The danger is that it trains you out of thinking.
The same tools genuinely widened who gets to make things. Someone who spent thirty years believing they had nothing to make is making things, and I refuse to be sour about that. Great work stays a moving horizon. What the tools raise is the floor.
The obligation travels with the inheritance. Credit the source. Respect consent. Pay people where their work creates value. Stay answerable for what you ship.
Own what you make.
07
Leave a rung
No law of physics marks a beginning. A day becomes a door because you decide it is one.
So here is the practice.
Rung 01 / Follow one thread
Once a week, pick one ordinary object you touched. A phone, a spoon, a hinge, a streetlight, a book. Spend ten minutes tracing one part of its stack or its chain. What is it made of? What discovery made it possible? Who standardized it? What labor does the smooth surface hide? One honest thread is enough. You are looking for the seam.
Rung 02 / Build one thing
Make the smallest useful version of something that did not exist before you started. A paragraph. A function. A meal. A repair. A diagram. A shelf. A message you have been putting off. Do not wait for the project to feel important enough. Let it teach you what the next dependency is.
Rung 03 / Leave a map
Write down what worked, what failed, where the source is, and what the next person should know. Name the people whose work you used. Share the template. Comment the strange decision. Teach the shortcut without hiding its cost. Make your rung strong enough for a foot you will never see.
Notice. Build. Transmit.
Ten minutes to notice. One small act to take part.
None of this is easy, and the chip says so, and the laboratory says so, and so does every pane of glass and every printed page. The just in “you can just build things” gives you permission to begin before you feel ready. It makes no promise about difficulty.
We invent from inheritance, understand only pieces of the stack, and carry the chain for a while.
Take hold of the part within reach, treat it honestly, and pass it farther than you found it.
It is exhilarating to be alive right now. Reality is bottomless and participation is possible.
You are alive on a moving rock that learned to look at stars, to carve sand into computation, to train systems inside games until they could help decode the machinery of life, and to pass forbidden ideas hand to hand through centuries of fear!
You are reading this. The lineage held.
Somewhere, a laser is striking a falling droplet of tin.
Somewhere, a researcher is comparing a prediction against a structure the team was not allowed to see.
Somewhere, a student is copying a diagram because an idea deserves another life.
Somewhere, a person is opening a blank document.
You are not outside civilization.
You are one of the places it continues.
Count this as your turn.
Look closely. Stay answerable. Leave a rung.
Notes and sources
- Apple, “Apple unleashes M5, the next big leap in AI performance for Apple silicon,” October 15, 2025. Apple identifies the M5 as a third-generation 3-nanometer design. The node name marks a manufacturing generation and does not correspond to any single dimension on the finished chip. Marques Brownlee, “I shrunk down into an M5 chip,” December 26, 2025. Brownlee ends the scale journey with the Prius-to-Earth comparison used in the text.↩
- ASML, “EUV lithography systems,” and “TWINSCAN EXE:5200B.” ASML is the only commercial supplier of EUV lithography systems. Paul Alcorn, “Intel shares biggest unboxing video ever as ASML’s $380 million High-NA lithography machine is installed in Oregon fab,” Tom’s Hardware, June 14, 2024. The installation report gives the approximate price, 150-tonne weight, 250-crate shipment, and six-month installation by 250 engineers.↩
- ASML, “How microchips are made.” ASML describes hundreds of fabrication steps, a design-to-mass-production cycle of up to four months, and chips with up to 100 precisely aligned layers.↩
- ASML, “EUV lithography systems.” The page explains the two-pulse tin-droplet light source, its rate of up to 50,000 cycles per second, the reflective optical column, the vacuum requirement, and the wafer-stage positioning and adjustment specifications.↩
- Gordon E. Moore, “Cramming more components onto integrated circuits,” Electronics 38, no. 8 (April 19, 1965), and Computer History Museum, “Moore’s Law predicts the future of integrated circuits.” Moore extrapolated from a handful of data points to a projection of 65,000 components per chip by 1975, revising the doubling period to roughly two years in 1975. Carver Mead later named it a law. The Prius-to-Earth comparison in the text is a scale translation, not a measurement.↩
- Smithsonian Institution, “ENIAC.” ENIAC is commonly described as using on the order of 18,000 vacuum tubes.↩
- Apple, “Apple unveils iPhone 15 Pro and iPhone 15 Pro Max,” September 12, 2023. Apple describes the A17 Pro as a 3-nanometer chip with 19 billion transistors.↩
- VIZ Media, “The Official Website for Dr. STONE.” VIZ credits Riichiro Inagaki and Boichi, and summarizes Senku and Taiju’s plan to restart civilization through science. The revival conflict described here is the series’ central early split in values.↩
- Tribeca Festival, “The Thinking Game.” The festival synopsis identifies director Greg Kohs and follows Demis Hassabis from chess and game development through neuroscience to the co-founding of DeepMind.↩
- CASP14, “Critical Assessment of Techniques for Protein Structure Prediction.” CASP describes its use of not-yet-public experimental structures, independent assessment, and the participation figures for CASP14.↩
- CASP13, “Critical Assessment of Techniques for Protein Structure Prediction,” 2018. DeepMind’s first AlphaFold placed first in CASP13, with a median score in the high fifties on the free-modelling targets, well short of the accuracy the CASP14 system reached two years later with a rebuilt architecture.↩
- John Jumper et al., “Highly accurate protein structure prediction with AlphaFold,” Nature 596 (2021): 583–589. See the abstract for CASP14 performance, and the section “MSA depth and cross-chain contacts” for the limitations.↩
- Google DeepMind, “AlphaFold reveals the structure of the protein universe,” July 28, 2022. EMBL-EBI and Google DeepMind, “AlphaFold Protein Structure Database.” The 2022 release expanded the resource beyond 200 million predicted structures and made them broadly accessible. The database publishes confidence estimates and cautions that its predictions have varying levels of confidence and must be interpreted accordingly.↩
- Leslie Felperin, “The Thinking Game review, DeepMind study offers wide-lens view of our tech lords and AGI,” The Guardian, March 19, 2025. Go watch the documentary.↩
- Seven Seas Entertainment, “Orb: On the Movements of the Earth (Omnibus).” The publisher credits Uoto, classifies the work as period science fiction and historical fiction, and summarizes its chain of hidden astronomical texts.↩