On ruin, nerve, and the quietest revolution you will ever live through
In March of 1930, with the American economy lying face down in the dirt, a crew of ironworkers began stacking steel on Fifth Avenue. They finished in 410 days. Four and a half stories a week. Lewis Hine photographed them eating lunch on the girders, a thousand feet above a city that was going broke underneath them.
Then it opened, and almost nobody came.
The Empire State Building stood mostly vacant through the 1930s. New Yorkers took to calling it the Empty State Building, because New Yorkers have never once passed up the chance to be cruel and correct in the same breath. The thing did not turn a real profit for the better part of two decades.
Sit with that for a second. In the worst economic collapse of the modern era, human beings built the tallest object on Earth, and then they were wrong about it for twenty years, and it is still standing there this morning with a line of tourists wrapped around the block.
That is the entire story of our species, compressed into one very tall building.
We have been here before, and we were worse off
I want to be careful here, because it is easy to wave at history like a tour guide and call it wisdom.
So let us be specific. Within a single human lifetime, roughly 1914 to 1945, the world did the following to itself. It fought a war that killed something like seventeen million people over a tangle of treaties almost nobody could explain. It then caught a flu that killed an estimated fifty million more, which is to say the disease outperformed the war by a factor of three and did not require a single general. It followed that with an economic collapse that put a quarter of American workers out of a job and turned the topsoil of an entire region into weather. And then, having learned very little, it fought a second war that killed somewhere between seventy and eighty-five million people and ended with two cities erased by a physics equation.
Thirty-one years. That is it. A person born in 1914 who lived to see V-J Day had watched the human race audition for extinction four separate times.
And here is the part nobody puts on the memorial. That same generation came home, invented the transistor, cracked the structure of DNA, put a polio vaccine into every schoolyard in the country, poured the interstate, and landed a man on the moon within twenty-four years of the last shot being fired.
They did not do this because they were noble. Plenty of them were not. They did it because there is something in us that cannot leave a broken thing alone. Give a human being rubble and enough time and they will start sorting it by size.
Nassim Taleb has a word for systems that behave this way. Not resilient, which only means you survive the hit unchanged. Antifragile, which means the hit is the ingredient. The stressor is not the thing you endure on the way to growth. It is the growth.
We are, on the evidence, the most antifragile thing this planet has produced.
Something new is on the table
Every generation gets handed a tool and gets told it changes everything, and mostly the tool changes one thing very well and leaves the rest of life exactly as annoying as it was.
The nail gun made carpentry faster. It did not make you a better carpenter. The automobile made you mobile. It did not make you interesting when you arrived. The airplane collapsed distance. The steamship collapsed oceans. Every one of these took a human limitation and pushed it back a few hundred yards.
What is sitting on the table now is not that.
For the first time, the thing being extended is not the arm or the leg or the range. It is the part of you that thinks. Not replaced. Extended. There is a meaningful difference, and most of the shouting on both sides comes from people who have not bothered to notice it.
A person with this thing can run twenty experiments before lunch that would have taken a research team a quarter to design. They can hold a strategy, a legal risk, a financial model, and a marketing narrative in the same conversation without any one of them falling out of the room. They can be wrong forty times before breakfast and cheaply, which is the only way anybody has ever been right about anything.
Taleb calls that convex tinkering. Low cost to fail, no ceiling on the win. Progress has almost never come from a smart plan executed cleanly. It comes from a thousand cheap attempts where one of them detonates. What has changed is not the logic. It is the price. The cost of an attempt just fell through the floor, and when the cost of trying falls far enough, the whole shape of a life changes.
The revolution nobody scheduled
Here is where it gets strange, and where I would ask you to actually look at the numbers rather than at the noise.
Gallup, surveying more than twenty-two thousand American employees in late 2025, found that 49% of workers say they never use AI in their job. Never. Not occasionally. Twelve percent use it daily. Twenty-six percent use it a few times a week or more.
Pew, polling 5,119 American adults in February 2026, found that 49% have ever used a chatbot, about 24% use one daily, and 4% describe their use as almost constant.
Four percent.
I have read that number probably thirty times and it still does something to my stomach. Four in a hundred adults are living inside this thing. The rest are standing in the yard, arms crossed, waiting for someone credible to tell them whether it is a fad.
Revolutions almost never announce themselves. They happen in garages, in spare bedrooms, in the ugly hours between eleven at night and two in the morning, conducted by people who look, to their neighbours, like they are wasting their lives on the computer again. Nobody sends a card. There is no ribbon cutting for a shift in what a single person is capable of.
Gallup found one more thing worth chewing on. Among leaders, frequent AI use has climbed from 17% to 44% since 2023. Among individual contributors it sits at 23%. The gap is not closing. It is widening. Read that however you like, but do not read it as nothing.
It is a skill, and skills take years
I have been at this for three years, most days, seriously.
I say that not as a credential but as a warning, because the single most common mistake I see is a person opening one of these tools, typing a lazy question, receiving a mediocre answer, and concluding that the technology is overrated. They have run the equivalent of sitting down at a piano, striking one key, and announcing that music is a scam.
It is a skill. It has a learning curve with the same miserable shape as every other skill, which is to say it is flat and humiliating for longer than you would like. Carol Dweck spent a career demonstrating that ability behaves less like a fixed quantity and more like a muscle, and Angela Duckworth spent hers showing that the thing separating high performers is rarely raw talent but the willingness to stay in a fight that has stopped being fun. Research on high achievers points at the same joint. Failure is not the divider. What separates people is whether the last failure gets carried into the next attempt or thrown away.
Low performers restart. High performers accumulate.
The tool does not fix this. The tool is a lever, and a lever with nobody leaning on it is a stick.
Which brings me to the part the loudest voices keep missing. This thing has no skin in the game. It does not lose the house. It does not have to look anyone in the eye. It cannot feel the specific, clarifying dread of a decision that is actually yours. That is what you bring, and it is not a small contribution. It is the entire reason the work means anything.
A prediction, with my name on it
I want this on the record so you can come back in 2030 and enjoy yourself at my expense.
Deezer reported in July 2026 that fully AI-generated tracks now exceed 50% of daily uploads to the platform, running near ninety thousand songs a day. Half the new music arriving on a major streaming service is machine made.
Now the part that should stop you cold. Those tracks account for between 1% and 3% of actual streams, and Deezer flags roughly 85% of even that as fraudulent.
The machines have taken the supply side completely, and the audience has not moved.
So here is my prediction, and I am labelling it clearly as a prediction rather than dressing it up as data. By 2030 a substantial share of charting music, mainstream film, corporate strategy, and government policy will have been built with this thing in the room. Not instead of people. With them. I think the number will be large enough to embarrass everyone currently insisting it will never happen.
But I no longer believe production is the interesting variable, and the Deezer numbers are why. Ninety thousand songs a day and nobody is listening. Infinite supply, unchanged demand. The bottleneck was never the making. It was ever and always the taste, the timing, the nerve, the particular human who knows which of the ninety thousand is the one.
Abundance does not make judgment obsolete. It makes judgment the only thing left worth having.
The unshackling
There is a version of the next decade where people get soft. I will not pretend otherwise. Some will let the muscle go and never notice it happening, and there is no clever way around that, because there has never been a tool in history that did not make somebody lazier.
But hold that next to what is now possible for people who were locked out.
For most of recorded history, the ability to build something meaningful was gated behind capital, credentials, pedigree, geography, and a certain kind of luck that looks a lot like being born near the right people. You could be the sharpest mind in a small town and die anonymous because the gate was somewhere else and you had no key.
That gate is coming off the hinges. Not fully. Not fairly. Not all at once. But it is moving, and what is replacing it is not IQ. It is not a degree. It is not money.
It is whether you are curious enough to keep asking. Whether you are stubborn enough to be bad at something in public for eighteen months. Whether you can hold a hard idea in your head at midnight when nobody is checking.
That is not intelligence. That is character, and character has never been distributed according to income.
The ironworkers on Fifth Avenue in 1930 had no idea whether anyone would ever rent the thing. The building sat empty for years and they were, by every reasonable short-term measure, wrong.
They built it anyway. Four and a half stories a week, over a city that was falling apart, because the alternative was to stand in the street and wait for conditions to improve.
Conditions never improve first. That is not how any of this has ever worked.
Somebody is in a spare bedroom tonight, three hours into something nobody asked them to do, teaching themselves a skill their entire circle would find faintly embarrassing if they tried to explain it.
I would keep an eye on that person.
Notes Beyond the Map is a blog about mind, world, and enterprise, written for people who refuse to follow the herd. New articles published at joshnash.ca.
Notes Beyond the Map 
Leave a comment