Warren Buffett paid more than $55 million for control of a furniture company without an audit.
The contract was one page.
Six years later, its 95-year-old founder walked out and opened a rival business across the street.
Her name was Rose Blumkin.
She arrived in America from Belarus in 1917. No formal education. No English. She learned the language from her daughter.
In 1937, she borrowed $500 from her brother and opened Nebraska Furniture Mart in the basement of her husband's shop in Omaha.
Her rule was simple. Sell cheap and tell the truth.
When manufacturers refused to sell to her for underpricing established retailers, she drove to Kansas City, Chicago, and New York. Bought furniture at full retail from department stores. Brought it back to Omaha. And still sold it cheaper than anyone in town.
By 1983, one Omaha location was generating more than $100 million in annual sales.
Buffett had watched that record for years. On August 30 of that year, his 53rd birthday, he made the cash offer. Mrs. B was 89.
Mrs. B gave Buffett her word. He wrote the check.
The footage attached to this post captures the acquisition as it happened.
Mrs. B remained chairman. Her son Louie and grandsons Ron and Irv continued managing Nebraska Furniture Mart with her.
In 1989, at 95, she felt frozen out as her grandsons took more control of the business.
Three months later, she opened Mrs. B's Clearance and Factory Outlet directly opposite Nebraska Furniture Mart.
Buffett never put a non-compete clause in the original deal.
Mrs. B had not broken her word. No agreement prevented her from leaving or competing.
In 1992, at 99, Berkshire bought her rival business and merged it back into Nebraska Furniture Mart. This time, she signed a non-compete agreement.
She worked until 103. Died at 104.
Trust closed the first deal. The missing clause brought Buffett back for the second.
NVIDIA nearly died because its first rendering technology was wrong.
Jensen Huang later admitted it on stage.
In the mid-1990s, NVIDIA bet on forward texture mapping, a way of drawing 3D images that put it outside the industry standard.
Silicon Graphics had the money. 3dfx had the hype.
NVIDIA had a chip built around the wrong idea.
Two years and millions of dollars later, the industry settled on inverse texture mapping and triangle-based graphics.
NVIDIA had to abandon the architecture it had built the company around.
It was competing against dozens of graphics companies, all trying to leapfrog one another every few months.
Then came the move that mattered.
NVIDIA stopped treating graphics as a chip problem.
It went directly to game developers.
In those rooms, developers did not talk about benchmarks. They talked about worlds, images, and what players would feel on screen.
NVIDIA learned that its product lived between engineering and art.
The company rebuilt around that idea.
The mistake did not make NVIDIA win.
The speed of abandoning it may have saved the company.
Milton Friedman held up a 10-cent pencil on television and made an impossible claim.
No single person on earth knew how to make one from scratch.
1980 PBS. Free to Choose. Episode one.
He holds the pencil up to the camera and starts pulling it apart.
Wood from Washington. Graphite from South America. Rubber from Malaya, off trees that arenāt even native there. Brass, yellow lacquer, black lines, glue.
It sounds like a geography lesson. It isnāt.
Thousands of people across four continents made that pencil. Different languages. Different countries.
Not one of them woke up that morning intending to make a pencil.
No ministry assigned them. No central office sent orders. No single blueprint exists.
Friedmanās argument was that prices coordinated it.
You walk into a store, pay a few cents, and receive a few seconds of work from thousands of strangers who will never know your name.
He did not invent the argument. Leonard Read wrote it first in a 1958 essay called āI, Pencil.ā
Friedman did something different. He turned the argument into a 10-cent object anyone could hold in one hand.
He could have used equations.
He chose a pencil.
Ten cents bought more coordination than any one person could design.
A UCLA professor raised 15 million Americans' savings rate from 3.5% to 13.6%. He never asked them to save a single dollar today.
His trick was one word: tomorrow.
Shlomo Benartzi stood on the TED stage and opened with a question. If I offer you a banana or chocolate for next week, what do you pick? 74% pick the banana. Healthy. Responsible. Smart.
Now what if I offer you a banana or chocolate right now?
70% pick the chocolate.
Same brain. Same person. Same choice. Different answer. Because next week you are a hero. Today you are a human.
That gap between who you plan to be and who you actually are is the reason you are not saving enough. Not discipline. Not income. Not knowledge. The gap.
Benartzi knew this. He is a behavioral economist at UCLA who spent 15 years studying why people fail with money. Not why they are stupid. Why they are human. And humans have one consistent flaw: we cannot sacrifice now for later. We know we should. We plan to. We swear we will start next month. And next month we pick the chocolate again.
So he stopped asking people to save today.
He built a program with Nobel laureate Richard Thaler called Save More Tomorrow. The idea was almost insultingly simple. Instead of asking employees to save more right now, he asked them to commit to saving more with their next raise. Not today. Tomorrow.
You earn $5,000 a month and save 3%. Fine. Keep spending exactly what you spend. But the next time you get a raise, half of it goes to savings automatically. You never see it. You never touch it. You never feel the loss.
The results broke everything behavioral economists thought they knew.
Savings rates went from 3.5% to 13.6%. Nearly four times higher. Not because people earned more. Not because they learned more. Because someone finally designed a system that worked with human nature instead of against it.
Then he showed the slide that made the room go quiet.
Germany and Austria. Same culture. Same continent. Same values. Germany: 12% organ donors. Austria: 99%. The difference? In Germany, you have to opt in. In Austria, you have to opt out.
The default wins. Every time. In every domain. Organ donation. Retirement savings. Your spending habits. Whatever the default is, that is what you will do. And right now, your default is to spend everything that hits your account.
That is why willpower fails. That is why budgets fail. That is why "I'll start saving next month" has a 100% failure rate across 300 months. You are fighting the default. And the default always wins.
Benartzi did not fight it. He changed it.
Save More Tomorrow is now used by the majority of large retirement plans in the United States. It has been written into the Pension Protection Act of 2006. 15 million Americans are saving more because of one behavioral insight: do not ask people to change today. Ask them to change tomorrow, and automate it so they never have to decide again.
The most powerful financial advice on earth is not "save more." It is "set up a system where saving happens without you."
One automated transfer. One decision. Made once. That is it. That is the whole program that changed 15 million lives.
You already know you should save more. You have known for years. And you have not done it. Not because you are lazy. Because you are human. And humans pick the chocolate every single time.
Stop asking yourself to pick the banana. Set up the transfer and take yourself out of the equation.
In 2009, Elon Musk called a New York Times reporter "a huge douchebag" on camera. Then repaid the government $465 million nine years early.
Tesla had sold one product at that point. A $109,000 sports car. Sarah Lacy asks about the Randall Stross piece, the one that framed taxpayer money going to a toy for the wealthy.
Musk doesn't soften it.
"Randy Stross is a huge douchebag and an idiot."
Then he does the math on camera. The ATVM program is $25 billion. Tesla is asking for 1.5% of it. The other 98% goes to Ford, GM, and Chrysler. He looks at the room and says "why doesn't he bug the guys that are getting the 98%?"
Lacy pushes. Tesla pulled a $100 million round the summer before. Fisker just closed $85 million. Model S needs $400 million to exist. Where does the money come from?
"99%, damn close to 100%."
Then the part nobody clipped. Buried in the same answer. No investor, no employee, no stockholder sees a dollar until the government is paid back in full.
The loan lands June 2009. Tesla clears it May 2013. Nine years ahead of schedule.
GM files Chapter 11 weeks after that interview. Taxpayers eat roughly $11 billion. Fisker draws $192 million from the same fund and dies the same year Tesla pays off its debt.
The Times prints a correction. Everyone remembers the douchebag line.
The 1.5% is the part that mattered.
Fred Trump made $300 million and never put his name on a building. His son put his name on everything. It almost killed him. Then it saved him.
Fred built apartments in Queens, Brooklyn, and Staten Island for 50 years. Thousands of units. Never did an interview he didn't have to.
Donald joined the family business in 1968 and immediately wanted out of the outer boroughs. Manhattan. Glass. Steel. His name on the front.
His father thought it was reckless. He gave him the loan anyway. $1 million in cash and a $70 million bank guarantee for a crumbling hotel near Grand Central. It reopened as the Grand Hyatt in 1980. Donald was 34.
Then Trump Tower. Gold letters on Fifth Avenue. Then three casinos in Atlantic City. A yacht. An airline. A football team. A ghostwritten book that sat on the bestseller list for 40 weeks.
By 1989 he was on every magazine cover in New York. He hadn't just built buildings. He'd played the tabloids off each other until his name was the most recognized in the city.
His father never understood why any of it was necessary.
Then 1990 hit. The debt was $3.4 billion. Four properties filed for bankruptcy. The banks owned him.
They sat down to decide whether to bury him or restructure. And they calculated something Fred never would have predicted. The name on the buildings was worth more than the buildings. The covers, the bestseller, the celebrity he'd manufactured for a decade. That was the asset the balance sheet couldn't measure.
No one restructures debt for a man nobody's heard of. Fred's strategy would have killed him.
Fred Trump died in 1999. Worth $300 million. Never famous. Never bankrupt. Never president.
Same last name. Opposite playbook. Both worked. Only one of them almost didn't.
Jeff Bezos told his parents there was a 70% chance they'd lose everything. They wrote him a check for $300,000. It was their life savings.
1997. A camera crew finds him outside a building in Seattle. He's 33. The sign says https://t.co/6oJGDQ4bEj. The office inside shares a block with a pawn shop and a needle exchange.
He walks them in and there are no desks. Every surface in the building is a door from Home Depot with four-by-four wooden legs screwed into the bottom. Someone later asked him why. He said he looked at the price of desks and the price of doors and the doors were cheaper.
He'd quit a hedge fund job in New York the year before because he found one number: web usage was growing 2,300% a year. He made a list of 20 products to sell online. Books won. Not because people love reading. Because there are 3 million titles in print and no building on earth could stock them all.
The company had just gone public. $54 million raised. He was a millionaire on paper. But he tells the camera something that sounds insane for a CEO weeks after an IPO.
"This is Day One. This is the Kitty Hawk stage of electronic commerce."
Not early. Not young. The very first flight. Barely off the ground.
Three years later the dot-com bubble burst. Amazon's stock dropped from $107 to $6. A 94% fall. Analysts called it Amazon.bomb. Barron's ran a cover story titled "Amazon.toast." His parents' $300,000 was almost gone.
He didn't sell. They didn't sell.
That $300,000 check is now worth over $30 billion.
The desks are still doors.
Donald Trump went $3.4 billion into debt building things. He came back by putting his name on things other people built. That one pivot made him president.
1990. Four bankruptcies in three years. Taj Mahal, Trump Castle, Trump Plaza, the Plaza Hotel. Yacht sold. Airline shut down. Banks put him on a personal spending allowance. The New York Post runs "UH-OWE!" on the front page. Forbes drops his net worth to $500 million, then questions if even that's real.
By 1992 he throws himself a comeback party in Atlantic City. Most of the debt is still there. But the cameras show up anyway.
That's when he sees it.
The buildings lose money. The name doesn't. The name fills rooms. The name gets cameras. The name gets the party thrown even when the math says you shouldn't be throwing parties.
So he stops building.
Trump Tower Toronto. Trump SoHo. Trump International Hotel Las Vegas. He didn't build them. He didn't own them. He didn't finance them. Developers paid him millions just to put his name on the glass. If the building failed, they ate the loss. He kept the licensing fee.
By 1997 he publishes The Art of the Comeback and tells NBC the brand is the business. Not real estate. The name.
He licensed it onto steaks. Vodka. A university. A board game. A reality show. Most of it failed. Didn't matter. Every failure kept the name in circulation and every dollar of risk belonged to someone else.
The man who nearly lost everything because he personally guaranteed $832 million in debt never personally guaranteed anything again. He made other people bet on his name while he collected a percentage for showing up.
By 2004 The Apprentice made him the most recognized businessman in America. By 2016 he used that recognition to win the presidency.
The buildings bankrupted him. The name on the buildings made him the most powerful man in the world.
Michael Gerber coached 25,000 businesses. 80% of them had the same problem. Not money. Not clients. The owner wouldn't leave the room.
1977. Palo Alto, California. Gerber walks into a pie shop. The owner, Sarah, makes the best lemon meringue in the Bay Area. Lines at 7 AM. Sells out by 2 PM. Revenue $380,000 a year. Profit? Negative. She hasn't taken a day off in three years. Sleeps four hours. Gained 30 pounds. Tells Gerber she needs a $50,000 loan or she's done.
He doesn't look at her books. He asks one question. "What happens to the pies if you break your arm tomorrow?"
She starts crying. Because she already knows.
There is no business. There's Sarah. Sarah opens. Sarah bakes. Sarah sells. Sarah closes. Sarah IS the pie shop. And a business that is one person isn't a business. It's a hostage situation where the hostage also pays rent.
Gerber spent the next 40 years proving the same pattern in every industry. Dentists. Contractors. Developers. Lawyers. Gym owners. The best barber opens a barbershop. Works 12 hours cutting hair. Goes home. Comes back. Cuts more hair. Makes $9,000 a month and thinks the answer is $12,000 a month.
It's never the answer. More money through a system built for one person just burns faster.
Ninety-six percent of small businesses never reach $1 million in revenue. Not because they're bad at the work. Because they never stop doing it long enough to build the machine that does it for them.
Gerber wrote one book about it in 1986. The E-Myth. Five million copies sold. Forty years later, still the same book on every failing business owner's nightstand. Not because nobody reads it. Because everybody reads it and then walks right back into the shop and starts baking pies again.
The best pies in the Bay Area almost bankrupted the only person who could make them.
In 2008, Elon Musk was borrowing money from friends to pay rent. In 2026, he became the first trillionaire in human history.
Somewhere between those two points, he stood in a room in Hollywood trying to sell a $109,000 electric car to people who thought he was insane.
His own Roadster was parked outside. Tesla had built 99 cars. Number 100 was coming off the line Tuesday. The company had days of cash left.
He didn't mention that part.
He talked about a $25 billion fund Congress had approved to help automakers build efficient vehicles. GM, Ford, and Chrysler flew to Washington and asked for that money to cover payroll on their existing gas-burning fleets. Congress said yes. Musk looked at the room and said "that gives them until about February."
Tesla applied from the same fund. But for what the fund was actually created for: a cheaper electric car.
Then he said something nobody in the room registered. Tesla doesn't pay dividends. Never will. Every dollar from every $109,000 Roadster goes directly into building the next car at a lower price. His salary: minimum legal.
What he didn't tell the room: if Tesla didn't close an emergency funding round before Christmas Eve, the company was dead. Not struggling. Dead. He later said he put in every penny he had. Didn't own a house. Borrowed money from friends for rent.
The round closed at 6pm on Christmas Eve. Last hour of the last possible day. Investors were leaving town that night.
Six months later, General Motors filed for bankruptcy. The largest industrial bankruptcy in American history. Chrysler followed.
The guy on minimum salary who was borrowing money for rent didn't.
Jimmy Fallon asked Bill Gates if weāll still need humans. Gates said ānot for most things.ā The audience laughed. He wasnāt joking.
February 2025. The Tonight Show. Gates is promoting his memoir. But the only clip anyone remembers is 12 seconds long.
He explains it simply. Right now intelligence is rare. A great doctor, a great teacher, a great mental health professional. There arenāt enough of them. With AI, within the next decade, that expertise becomes free.
Commonplace. Available to everyone.
Fallon asks the obvious follow-up. So will we still need people?
Not for most things, Gates says. Then adds: weāll decide what to keep for ourselves. Like baseball. Nervous laughter. He moves on.
Making things. Moving things. Growing food. He calls these āsolved problemsā within a decade. Not reduced. Solved.
Heās not warning you. Heās not worried. Heās describing what he considers progress. Free intelligence for everyone. The question heās not answering is what happens to the people whose only product was their intelligence.
12 seconds. 1 clip. The most honest thing a billionaire has said about your job on national television.
Steve Jobs walked into Xerox in 1979 and saw the future. He left 20 minutes too early. That mistake cost Apple a decade.
Xerox showed him three things. A graphical interface with windows you could move with a mouse. Object-oriented programming. And a network of over 100 computers all talking to each other, sending emails, sharing files. 1979. Fifteen years before the internet went mainstream.
Jobs saw the first one and lost his mind. Started jumping around the room. Went back to Apple and told his team to drop everything and build it. That demo became the Macintosh.
But he never saw the other two.
He admitted it years later. If he had stayed for another twenty minutes. If he had watched the networking demo. If he had seen those 100 computers working together. The Mac would have been built as a networked machine from day one.
Instead, Apple spent the next ten years making beautiful computers that couldnāt talk to each other.
And Xerox had it worse. They had all three. The interface. The programming language. The network. They invented the future of computing in one building and did nothing with it. Jobs said it himself: Xerox could have been as big as IBM plus Microsoft plus Xerox combined. They became a footnote.
The guy who visited grabbed one idea and built a $3 trillion company. The company that owned all three sold toner.
A GIRL FROM JAKARTA MAKES $17,000 A MONTH. THE BUYERS HAVE NEVER MET HER. SHE DOESNāT EXIST.
Not showbiz. Not a business. Not trading. AI models.
Buyers are foreigners. Canada, France, Switzerland. They pay for content. Thatās it.
She never shows her face. Lives in a country where the infrastructure barely holds up. Out-earns the president.
Hereās what actually gets you the detail work.
This isnāt the AI slop you clock in one glance. No extra fingers. No plastic skin. No blurred background where the camera shouldāve moved. Shadows fall like someone was actually standing by a window. Movement has the micro-delays that used to give generation away.
The gap between āyou can tell itās AIā and āyou canātā is the gap between zero followers and $17,000 a month.
And itās not one girl. Itās a system.
One operator runs 4 of these characters at once. The girls donāt exist. The money comes from donations, selling adult content to subscribers from all over the world, using nothing but AI.
The subscriber thinks heās talking to a real woman. Heās actually talking to a prompt file and a chatbot that remembers what it said yesterday.
The money isnāt coming from ads. Itās coming from trust an algorithm is building.
This stopped being āgenerated a picture.ā Itās a production studio on one laptop, manufacturing intimacy as a product.
The question isnāt āis this actually possible.ā Itās already running. The question is how long before this stops being an anomaly and becomes the default.