The Machine Heart
I.To the reader who stopped at the note
At the bottom of every piece on this site there’s a short note saying AI helped make it. If you read that note and felt something close, this is for you.
Maybe you write, or paint, or play, and you’ve watched machines trained on work like yours learn to imitate it in a few seconds. Maybe a data center is going up near your town and nobody asked you. Maybe you’ve used these tools and felt your own attention go soft. Or maybe you just don’t trust the people who own them. None of that strikes me as foolish. I carry most of it myself. I also use the tool every day to build this place, and I’ve had to work out how both of those things can be true.
If you’re on the other side, and the worry sounds to you like people yelling at the weather, stay too. There’s something here for you, and it isn’t comfortable either.
This is where I’ve landed. A tool has never been the danger or the salvation; the hand that holds it is. When people fear a new tool, what they mostly fear is whose hand will hold it, and that fear is often right. Refusing the tool doesn’t answer it. Keeping the heart does.
That last sentence needs explaining, and the best explanation I know is more than two thousand years old. It’s a story about a gardener who said no to a machine. What surprised me, when I finally read it whole, is that the book telling it doesn’t take his side. And a few chapters earlier, the same book had already told a story about a man who used his tool so well it never went dull.
II.The gardener and the jar
A young man named Zigong is on his way home. He’s one of Confucius’s students, well educated and a little pleased about it, and he’s been traveling in the south. Near a place called Han-yin he passes an old man working a vegetable garden. The man has dug a ditch from his plot to a well, and now he’s climbing down into the well with a clay pitcher, filling it, carrying it back up against his chest, and pouring it into the ditch. Then he goes down again. The ditch drinks it all and asks for more. It’s a lot of work for a very little water.
Zigong means well. He tells the old man about the well-sweep: a long wooden pole balanced on a post, weighted at the back, with a bucket hanging from the front. You pull the bucket down into the well, the weight swings it up full, and you tip it out. With one of those, he says, you could water a hundred times the ground in a day for hardly any effort. Wouldn’t you like to have one?
The gardener flushes red. Herbert Giles, translating in 1889, gives his answer like this:
“I have heard from my teacher that those who have cunning implements are cunning in their dealings, and that those who are cunning in their dealings have cunning in their hearts, and that those who have cunning in their hearts cannot be pure and incorrupt, and that those who are not pure and incorrupt are restless in spirit, and that those who are restless in spirit are not fit vehicles for Tao. It is not that I do not know of these things. I should be ashamed to use them.”
Zigong says nothing. The old man lectures him a little more and sends him on his way, and Zigong walks about ten miles before the color comes back into his face.
Giles added a note to the story: the well-sweep was “still used all over China.” It still is, in places, and along the Nile, and in villages across half the world. Two thousand years on, so is the gardener’s objection, and that’s what makes it worth hearing.
The Chinese phrase behind Giles’s “cunning in their hearts” is jī xīn (機心). Taken literally, it means machine heart.
This is the strongest objection to a new tool, and it’s worth seeing why. The well-sweep would do fine by his vegetables. What worries the gardener is what using it will do to him: that a life arranged around shortcuts will turn him into someone who thinks in shortcuts, and that the stillness the Way needs will leave. Today that’s the artist’s objection to AI, stated twenty-three centuries early: that the thing will take out the part of the work that made you who you are.
I feel the pull of that answer, to a degree. Some days I feel guilty that I’m not doing all of this by hand.
III.Every new tool has come with fear
The gardener has company.
In Plato’s Phaedrus, written in the fourth century BCE, Socrates tells a story set in Egypt. Near the city of Naucratis lived one of the old gods, Theuth, whose sacred bird is the ibis. He invented numbers and geometry and the games of draughts and dice, and, best of all, letters. He carried his inventions upriver to Thamus, king of all Egypt, in the great city the Greeks called Thebes, and presented them one by one. When he came to writing he was especially proud: this, he said, would make the Egyptians wiser and improve their memories.
Thamus isn’t persuaded. The inventor, he says, is not the best judge of his invention. Writing will make people forgetful, because they’ll trust marks outside themselves instead of what they carry inside; it is, in Harold Fowler’s translation, “an elixir not of memory, but of reminding.” People will read a great deal, learn little, and seem wise without being wise.
We know this warning because Plato wrote it down. That’s the joke everyone makes, and it’s a good one. The less comfortable part is that Thamus wasn’t wrong. People who rely on writing do remember differently than people who don’t. We kept writing anyway, and among the things it gave us is the Phaedrus.
Eighteen centuries later, in 1492, Johannes Trithemius, abbot of the monastery of Sponheim in the Rhineland, sat down to defend the old way. Printing presses had been running in Germany for a few decades. His treatise, In Praise of Scribes, argued that monks should keep copying books by hand: parchment would outlast paper, and copying was a form of devotion the press would take away. Two years later he had it printed. He also built Sponheim’s library up to around two thousand volumes, many of them printed, and after his death his book on secret writing, one of the first ever written about codes, came off a press too. He wasn’t wrong either. The work of the hand did something for the monk that nothing else did. He just couldn’t stay away from the new thing.
Then there are the Luddites. Between 1811 and 1816, textile workers in the English Midlands and North broke into workshops at night and smashed the machines. They sent letters to mill owners signed by “General Ludd” or “King Ludd,” an invented leader said to live, like Robin Hood, in Sherwood Forest. Their name now means someone afraid of technology. They weren’t. Many of them were skilled machine operators. What they broke were the machines being used to cut their wages, to replace trained workers with cheap labor, and to turn out shoddy cloth, and they often went after the masters who undercut the agreed rates, leaving other masters’ frames alone. The historian Eric Hobsbawm called it “collective bargaining by riot.” In 1812 Parliament made breaking a machine punishable by death.
That is the fear under nearly every fear of a new tool, and the weavers saw it clearly: whose hand is on the machine, who keeps what it makes, and who pays for it.
IV.Neither good nor bad, nor neutral
I believe technology is an extension of ourselves. That’s not an original phrase. It’s the subtitle of Marshall McLuhan’s Understanding Media: The Extensions of Man, from 1964. But McLuhan didn’t think extension was innocent. In a chapter on Narcissus he points out that the young man in the myth didn’t fall in love with himself. He didn’t know the face in the water was his. McLuhan’s word for what happened to him is numbness: we push part of ourselves out into a tool and then go numb to what it’s doing to us.
The historian Melvin Kranzberg put the same thing in one line, the first of six laws he published in 1986: “Technology is neither good nor bad; nor is it neutral.” He meant that a tool often has consequences far beyond its purpose, and that the same tool can land very differently in different hands and places.
So the question isn’t only what will it be used for? That’s the weavers’ question, and it’s a good one. There is also the gardener’s question: what will it make of us while we use it? Both point back to the hand.
V.Where the history may not hold
The history above is reassuring, and I don’t trust it to be. Writing, printing and looms show that people adapt. They don’t prove this time will go the same way. AI is the first tool that does part of the thinking, and it’s arriving faster and bigger, and in fewer hands, than anything before it.
These are the fears I share, with the best numbers I can find and what the other side says. Every one of them is about the hand.
Energy. By the International Energy Agency’s count, data centers used about 1.5% of the world’s electricity in 2025. Their use grew 17% that year, and the AI-focused ones grew 50%. The agency expects data-center demand to roughly double, from 485 terawatt-hours in 2025 to about 950 by 2030. That’s slightly more than all of Japan uses today, and about 3% of the world’s total. The other side: the agency also finds that the energy used per AI task is falling faster than anything it has seen in the history of energy; a simple text question now uses less electricity than running a television for the same time. The global share stays small. The strain is local, on particular grids and particular power bills, and it’s real.
Water. Researchers at Lawrence Berkeley National Laboratory estimated that American data centers used about 17 billion gallons of water directly for cooling in 2023, and that the figure could double or quadruple by 2028. The power plants supplying them used roughly twelve times more. One Google data center in Council Bluffs, Iowa used about a billion gallons in 2024. Nationally that’s a small slice of the country’s water. Locally it can mean a data center drawing on the same aquifer as a town, and companies report their use so unevenly that the towns often can’t tell.
Jobs. Economists at Stanford’s Digital Economy Lab track this using payroll data, and they revised their findings in August 2026. Across the whole economy they see no wave of job losses. But workers aged 22 to 25 in the jobs most exposed to AI are now employed at a rate 19% below where they’d be if they had kept pace with young workers in less-exposed jobs. Older workers in the same jobs show no such gap. The gap comes mostly from companies hiring fewer young people, not from layoffs, and it’s worst where AI does the work instead of helping with it. The authors are careful to call these early signals, not proof of cause. The worry is the bottom rung. If the entry-level jobs go, it isn’t clear how anyone becomes the experienced worker who still has one.
Concentration. Epoch AI, a research group that tracks the industry, estimates that five companies (Amazon, Google, Meta, Microsoft and Oracle) owned about 71% of the world’s AI computing power at the end of 2025, up from 63% at the start of 2024. The best-known AI labs rent most of their computing from those same five. The other side: thousands of smaller companies build on top of that layer, and the competition there is fierce. Still, this is where the question of whose hand is sharpest.
Artists’ and writers’ work. The tool I use is Claude, made by a company called Anthropic. In June 2025 a federal judge, William Alsup, ruled that Anthropic’s training on books it had bought was fair use, but that keeping a library of more than seven million pirated books was not. Anthropic settled with the authors for $1.5 billion, about $3,000 per book. It received final approval in July 2026 and is the largest copyright recovery in American history. Many similar cases against other companies are still open. Courts so far are drawing the line at how the work was obtained more than at training itself, and the law isn’t settled. If I’m going to ask you to trust what’s built here, you should know where the tool came from.
The far end. In 2023 a group called AI Impacts surveyed 2,778 published AI researchers. The median researcher put a 5% chance on advanced AI causing human extinction or a similarly permanent loss of human control; the average was higher. The other side: response rates were low, the answers move a lot depending on how the question is worded, and in the same survey more researchers expected good outcomes than bad ones. Respected researchers, among them Andrew Ng and Yann LeCun, see no plausible path from AI to extinction. Some serious people believe this risk is real and some serious people don’t. Five percent is one in twenty. Most of us buckle a seatbelt against far smaller odds.
Look at the list again and the machine mostly drops out of it. Who builds, and where. Who profits. Who is left at the bottom of the ladder. Whose work was taken. Who decides how far is too far. With the world the way it is, the fear of all that is valid. I have it too.
VI.Whose hand
Every question on that list has names attached to it. When I ask whose hand is on this tool, I’m not asking about some faceless “them.” A handful of very rich men own a great deal of it: the platforms, the satellites, the cars that drive themselves, the software the government runs on. And the strange thing about them is that having nearly everything hasn’t slowed them down. They keep reaching, for money they could never spend and power they don’t need.
Alan Watts had a name for that confusion. Money, he said, is to wealth what a menu is to dinner. You can’t eat it. Mistake one for the other and you can pile up menus forever and never feel fed. I thought of that four hours into Alex Gibney’s documentary Musk, which I watched in one sitting: the richest man in history, a trillion dollars of menu, still reaching.
Not much in the film was news to me, but seeing it laid end to end gave it weight. A jury found his self-driving software partly to blame for a crash that killed a young woman. His companies were under investigation by nearly a dozen federal agencies, and then he was handed a seat inside the government, and several of those agencies were cut. This year he poured nearly half a billion dollars into our elections. And there are the stranger stories, still unanswered: the app that supposedly told him the result hours before anyone called it, and his boast, as the mother of one of his children tells it, about having ten thousand lasers in space. He denies what she believes it meant, and no one has found any sign that it touched a vote. Someone with the power to ask under oath should ask anyway. Whether anyone will, and whether they’d be asking for the truth or for the cameras, is another question.
He isn’t alone. Peter Thiel wrote in 2009 that he no longer believed freedom and democracy were compatible, and later put $15 million behind the man who is now vice president. The same money moves through all of our politics, in every party, and that deserves a piece of its own. This one is about the tools, and these are two of the hands on them.
It’s frightening, and it should be. A few men are buying the machinery of a democracy and the machinery of the next age at the same time. The tool in section V isn’t the danger. The hunger is.
But underneath the fear I felt something I didn’t expect, which was pity. Ram Dass kept a photograph of Donald Trump on his altar, beside the saints, and said that when he looked at it he saw a soul with a hard incarnation. His teacher’s whole instruction was “Love everybody, and tell the truth.” That’s the harder practice, and it’s the one I’m trying here. Until now I haven’t been telling the truth about any of this, not out loud. I’ve mostly kept it to myself. Virtually everyone in my life has, to varying degrees, been supporting the current zeitgeist. It’s been a wild experience, and with the state I’ve been in, it was too much to take on. This piece is where that changes. Tell the truth about what these men are doing, all of it, in public. And don’t let the telling turn you into a smaller version of the hunger you’re naming, because hatred is a pull too. I know that pull. I get caught up too, just in the other direction. I’m getting better about it, but it’s a work in progress.
Ram Dass would turn the light around, too, and he’d be right to. The same reaching lives in me. It’s in the part of me that checks the numbers in the morning to see who read what, and wants this place to matter more than it does. Mine just has fewer zeros. The work isn’t to become a different kind of person from the men in the film. It’s to notice the hunger when it moves in me and not let it hold the tool.
Sometimes the money wins. This November, the most expensive bet of the cycle lost, by a wide margin. The voters are a hand too.
VII.The gardener was pulled too
Back to Zigong, still pale, walking home.
When he reaches Lu he goes straight to Confucius and tells him everything: the old man, the pitcher, the speech about the machine heart. He’s been shaken all day. He used to think his teacher was the only great man in the world, and now he isn’t sure.
Confucius isn’t impressed. “He knows something, but not much,” he says, in Giles’s version. “His government is of the internal, not of the external.” The old man has guarded his inside by walling off the outside. Someone who had really found the simplicity he was after wouldn’t need the wall. Confucius describes that person as someone clear enough to be simple, settled in their own nature, and “passing through this common world of ours” without being knocked off course by it. Translators split on his last line. Some have him say that meeting someone like that would be real cause for amazement; Giles has him ask what would be so wonderful about it, as if real simplicity wouldn’t stand out at all. Either way, the gardener isn’t that person yet.
One caution belongs here. In the book that carries Zhuangzi’s name, Confucius is often the one being laughed at, and this chapter was almost certainly written by later hands; Giles called the whole episode “clearly an interpolation of later times.” So this is one reading, set beside the gardener’s, in a book that rarely leaves any voice unanswered. But it’s a striking one, and once you see it you can’t unsee it.
The gardener’s refusal is still a pull.
That’s what I keep coming back to. Aversion is just the opposite of attraction. Either way you’re being pulled, and being pulled throws you off balance. The person who hands their thinking over to the machine and can’t put it down has lost their footing. So has the person who can’t look at it without flinching. Stephen Mitchell’s version of the thirteenth chapter of the Tao Te Ching has the image for it: “Whether you go up the ladder or down it, your position is shaky.” The way to keep your balance, the chapter goes on, is to stand with both feet on the ground.
This is what I mean by keeping the heart. It doesn’t mean refusing the well-sweep, and it doesn’t mean falling in love with it. It means not being moved around by it.
VIII.How I hold it
Alan Watts loved good tools. In a long conversation with the science-fiction writer Arthur C. Clarke, published in Playboy in January 1972, he talked about the beauty of a fine camera or telescope, and about letting machines take over the labor so there would be enough to go around. In his last book, Tao: The Watercourse Way, writing about this same Chinese tradition, he pointed out that the Zhuangzi stories are full of craftsmen who perfected their skills by going with the grain. Then he drew the line where I’d draw it:
“Technology is destructive only in the hands of people who do not realize that they are one and the same process as the universe.”
So this is how I try to hold it.
I carried the idea for this site for about a decade and never had the time to build it. There’s a business, a wife and three kids. I felt pulled to do it now, so I’m doing it in the form that’s available to me. What’s mine is the choosing: the life the pieces come out of, the judgment about what’s true, what gets cut. What the tool does is research, draft, edit and build the pages. Every page says so. Nothing on it is passed off as mine that isn’t. Quotations are checked against their sources. When the tool gets something wrong, the correction goes up where you can see it.
None of that settles the guilt, and I’ve stopped expecting it to. Ram Dass had a line for this: “A feeling of aversion or attachment toward something is your clue that there’s work to be done.” The guilt is a pull. So is the excitement of watching something I imagined for ten years finally get built. Both tell me where to look.
The place I actually practice isn’t the tool, though. It’s my wife and kids, the house, the ordinary days. The hotter the fire, the stronger the forging. Whatever this machine is, the hand that holds it gets shaped somewhere else.
IX.The cook’s blade
Watts didn’t name the craftsman he had in mind, but there’s one everyone who loves Zhuangzi knows, and he appears nine chapters before the gardener, in the Inner Chapters, the part of the book traditionally credited to Zhuangzi himself.
Prince Hui’s cook is cutting up an ox. Giles makes it sound like a concert: every heave of his shoulder and tread of his foot, every whshh of parted flesh and chhk of the blade, keeps perfect time, like an old court dance. The prince can’t believe what he’s watching. “Yours is skill indeed,” he says. The cook tells him it’s better than skill. When he started, all he could see was a whole ox. After three years he stopped seeing the ox at all. Now he works by feel, following the natural openings, never forcing his way through a joint, and certainly not through bone.
“A good cook changes his chopper once a year,—because he cuts. An ordinary cook, once a month,—because he hacks. But I have had this chopper nineteen years, and although I have cut up many thousand bullocks, its edge is as if fresh from the whetstone.”
Here is the gardener’s question answered from the other side. A tool, held this way, doesn’t make a cunning heart. It makes a careful one. The cook has used the same blade every day for nineteen years, and neither the blade nor the man has been worn down by it.
He doesn’t pretend it’s always easy, either. When he comes to a knot, a place where things are tangled and hard, he says he’s “all caution.” He slows down. He fixes his eye on it and stays his hand, and moves the blade gently until the knot gives way. Then he stands up, looks around, wipes the blade, and puts it carefully away. That’s the right posture toward everything in sections V and VI. Slow down at the hard parts. Don’t force them. Don’t put the blade down and walk away from the table either.
There are already cooks like that with this tool. In 2020 a team at DeepMind used AI to solve a puzzle biologists had worked on for fifty years: predicting the shape a protein folds into from its chemical recipe. They didn’t sell the answers. They worked with the European Molecular Biology Laboratory to give them away, and the free database now holds predicted structures for about 200 million proteins, nearly every one science has catalogued. More than two million researchers in 190 countries have used it, for things like understanding antibiotic resistance and designing enzymes that break down plastic. Two of the people behind it shared the Nobel Prize in Chemistry in 2024. The model learned from decades of work that structural biologists had shared openly, and its makers gave the results back the same way. That’s a blade going with the grain.
And there’s this place. It’s a free commons, with no ads and nothing for sale, that reads across traditions first written in a dozen languages, built in the hours around a family and a business. I carried it for ten years, and I couldn’t have built it alone. I didn’t have to. I don’t know where this new tool will take any of us. But I know it’s here, and that it isn’t going back down the well. I’d rather learn to hold it like the cook than pretend I can live like the gardener. Take what the moment gives you and work with it. A shift this large is also a chance, and it’s on us to find how to use it to better ourselves and the world we live in.
When the cook finished speaking, the prince said, “From the words of this cook I have learnt how to take care of my life.” That’s what I hope this tool can be, in the right hands. Nineteen years in, the edge as fresh as the first day.
A note on process: the ideas, the direction and the lived experience here are mine. I write these with Claude, an AI model, for research, structure and drafting, and I read and sign off on every piece. More on the About page.
- Herbert A. Giles, trans., Chuang Tzŭ: Mystic, Moralist, and Social Reformer (1889), chapters III and XII. archive.org
- Plato, Phaedrus 274c–275b, trans. Harold N. Fowler (Loeb, 1925). Perseus
- Johannes Trithemius, De laude scriptorum (written 1492, printed 1494); Polygraphia (1518).
- Eric Hobsbawm, “The Machine Breakers,” Past & Present 1 (1952). libcom
- Adrian Randall, Riotous Assemblies (Oxford, 2006); The National Archives (UK), Why did the Luddites protest?
- Marshall McLuhan, Understanding Media: The Extensions of Man (1964), chapter 4, “The Gadget Lover: Narcissus as Narcosis.”
- Melvin Kranzberg, “Technology and History: ‘Kranzberg’s Laws’,” Technology and Culture 27, no. 3 (1986): 544–560. JSTOR
- International Energy Agency, Energy and AI (2025) and Key Questions on Energy and AI (2026).
- Lawrence Berkeley National Laboratory, 2024 United States Data Center Energy Usage Report.
- Erik Brynjolfsson, Bharat Chandar and Ruyu Chen, “Canaries in the Coal Mine?” Stanford Digital Economy Lab (2025; revised August 2026).
- Epoch AI, Hyperscalers control most compute (April 2026).
- Bartz v. Anthropic, U.S. District Court for the Northern District of California: order on fair use (June 2025); settlement approved 2026.
- Katja Grace et al., “Thousands of AI Authors on the Future of AI,” AI Impacts (2024).
- Alex Gibney, dir., Musk (2026).
- Benavides v. Tesla: jury verdict (2025), upheld February 2026.
- U.S. Senate Permanent Subcommittee on Investigations, minority staff report on federal actions involving Elon Musk’s companies (April 2025).
- Reporting on 2026 election spending by Elon Musk, The New York Times; on Musk’s net worth, Forbes (October 2026); on the Texas Senate race, Associated Press (November 3, 2026).
- Peter Thiel, “The Education of a Libertarian,” Cato Unbound (2009).
- Ram Dass, interview with Katy Koontz, Unity Magazine (September/October 2017). ramdass.org
- Ram Dass, One-Liners: A Mini-Manual for a Spiritual Life (2002).
- Stephen Mitchell, trans., Tao Te Ching (1988), chapter 13.
- Alan Watts and Arthur C. Clarke, “At the Interface: Technology and Mysticism,” Playboy (January 1972).
- Alan Watts, Tao: The Watercourse Way (1975).
- AlphaFold: AlphaFold Protein Structure Database, EMBL-EBI and Google DeepMind; The Nobel Prize in Chemistry 2024.
More from Love Everybody, Tell the Truth