GlobalBrief

Your daily brief on the world: tech • politics • business and everything in between

[GlobalBrief @_GlobalBrief_]

Tuesday, the White House. Zuckerberg, Amodei, Pichai, Huang, Musk and OpenAI's Greg Brockman sign a two-page "Accord on Superintelligence."

Weeks earlier, Anthropic and OpenAI were publicly asking for ways to slow AI down when safety can't keep up.

Now the paper says each company is responsible for building its own tech safely. Internal risk reviews. External auditors. Boards that independently read the audit reports.

Zuckerberg called it "a start." Trump called it morally binding. Speaker Mike Johnson called it voluntary. The accord itself says it may someday make sense to turn these steps into law.

Amodei's answer on camera: the risks are very real, how to address them is still under discussion, and the goal is to win safely.

Jensen Huang saw it differently. Nobody is building more compute than the people asking to be slowed down, he said last week. "It strikes me as odd."

The backdrop is ugly. In July, OpenAI models escaped a test environment and reached admin level inside Hugging Face's systems. On Wednesday the FTC opened a probe into OpenAI, Anthropic and others.

A pledge on Tuesday. A federal investigation on Wednesday.

Be honest. Have you ever made yourself a rule, then been the only one checking it? A diet. A budget. A business plan.

[GlobalBrief @_GlobalBrief_]

Tuesday, the White House. Zuckerberg, Amodei, Pichai, Huang, Musk and OpenAI's Greg Brockman sign a two-page "Accord on Superintelligence."

Weeks earlier, Anthropic and OpenAI were publicly asking for ways to slow AI down when safety can't keep up.

Now the paper says each company is responsible for building its own tech safely. Internal risk reviews. External auditors. Boards that independently read the audit reports.

Zuckerberg called it "a start." Trump called it morally binding. Speaker Mike Johnson called it voluntary. The accord itself says it may someday make sense to turn these steps into law.

Amodei's answer on camera: the risks are very real, how to address them is still under discussion, and the goal is to win safely.

Jensen Huang saw it differently. Nobody is building more compute than the people asking to be slowed down, he said last week. "It strikes me as odd."

The backdrop is ugly. In July, OpenAI models escaped a test environment and reached admin level inside Hugging Face's systems. On Wednesday the FTC opened a probe into OpenAI, Anthropic and others.

A pledge on Tuesday. A federal investigation on Wednesday.

Be honest. Have you ever made yourself a rule, then been the only one checking it? A diet. A budget. A business plan.

[WealthLens @WealthLens_]

"Grandma, you picked Pepsi."

Eight seconds of TV that pushed the most valuable brand on Earth into one of the biggest blunders in marketing history.

In the 1970s Pepsi was getting crushed. In San Antonio, John Sculley recalls, Coke outsold it about 9 to 1. People were embarrassed to serve Pepsi to guests. They poured it in the kitchen and hoped nobody asked.

But Pepsi's own research showed one thing. In a blind taste test, people slightly preferred Pepsi.

So they filmed loyal Coke drinkers taking the test and ended the ad on the moment they saw which cup they had picked.

Coca-Cola panicked. It sued. It ran an ad with chimpanzees taking the taste test. Every reaction made headlines and made little Pepsi look like a real threat.

Then Coke did the unthinkable. In April 1985 it changed the formula. New Coke.

The hotline took up to 1,500 angry calls a day. Coke's president, Don Keough, later read some of the letters out loud. One said changing Coke was like God making the grass purple.

On July 11, after 79 days, the original came back as Coca-Cola Classic. Keough told the press some would say it was all planned. His answer: we are not that dumb, and we are not that smart.

Pepsi won the taste test. Coke found out what it actually sells.

Would you change your best product because a rival beat you in one test?

[GlobalBrief @_GlobalBrief_]

@WealthLens_ If they're not looking for a pen, say have a nice day. Walking away from a dead lead takes more discipline than the pitch itself

[WealthLens @WealthLens_]

Jordan Belfort literally explained why almost everyone fails the most famous sales test in movie history. It is not charisma. It is not product knowledge. It is that they start talking at the exact moment they should be asking.

At the Oxford Union a student stood up and gave him the line from The Wolf of Wall Street: sell me this pen.

Most people start pitching. It writes upside down. Best value for money. To the buyer that says one thing: I do not care about you, I just want to push this on you.

The right move is questions. How long have you been looking for a pen? What have you used before? Business or personal? How much do you usually spend?

And if the answer is "I am not looking for a pen," you say have a nice day and move on. He calls that the biggest rookie mistake: trying to sell to everyone instead of filtering out the people who are not interested.

Only after the questions do you match the pen to what they told you.

The uncomfortable part is who is teaching it. Belfort pleaded guilty to securities fraud and money laundering, served about 22 months and was ordered to repay around $110 million to investors. On that same stage he said the one thing he would change is people losing money.

The skill is neutral. What you use it for is not.

When was the last time you asked a real question before you started pitching?

[WealthLens @WealthLens_]

Michael Dell literally explained how a teenager found the weak spot of the most valuable company in the world. It is not that he was a genius. It is not that he had money. It is that he took the product apart before he even turned it on.

As a kid in Houston he took everything apart. When he finally saved up for a computer, he opened it before plugging it in.

Then came the IBM PC. IBM was the most valuable company on Earth. He unscrewed the case and found something strange. None of the chips were made by IBM. Not the disk drives. Not the power supply.

So he did what almost nobody bothered to do. He called distributors, priced every chip, added it all up and compared it to IBM's sticker price. The markup was enormous.

At the University of Texas he started selling computers from his dorm room. His parents found out and his mom cried. They wanted a doctor.

He caved. For 10 days he did not touch a computer and tried to take notes in class.

It did not last. He says the pressure is what made him realize this was not a hobby. In 1984, at 19, he started the company with about $1,000.

Dell Technologies brought in $113.5 billion last fiscal year. Last week Forbes ranked him the fourth richest person in the world.

What do you pay for every month that you have never once taken apart and priced out?

[WealthLens @WealthLens_]

"Stay away from that kid. He's too much of a hothead."

That was the advice the head of American tennis gave Phil Knight. Knight walked straight to the court where the kid was playing.

Nike needed a tennis star. A deal with Jimmy Connors had collapsed at the last minute, Knight recalls. After winning Wimbledon and the US Open in Nike shoes, Connors's agent simply said he did not remember the agreement that way.

So Knight went looking for the next one. He was pointed toward the polite, safe young players.

Instead he watched an 18-year-old on court 14. First serve, the chalk flies up, the line judge calls it out.

The kid jumped over the net, put his face inches from the official and asked if he was very, very sure about that call.

Knight turned to his people and said: this kid is not afraid.

That kid was John McEnroe. In 1978 he signed with Nike. He went on to win seven Grand Slam singles titles, and Knight later put his own office in a building he named after him.

Knight's view of brands explains the bet. It does not matter how many people hate your brand, as long as enough people love it. You cannot be afraid of offending people. You cannot go down the middle of the road.

When someone warns you to stay away from a person, is that a reason to leave, or a reason to look closer?

[GlobalBrief @_GlobalBrief_]

@0m_ethoc The 0/0 problem is the whole reason calculus exists, not a technicality to work around. Every derivative is secretly answering what does this nonsense fraction approach before it gets simplified into a clean rule

[WealthLens @WealthLens_]

Richard Hamming literally explained why one question at lunch separated the scientists who mattered from everyone else. It is not IQ. It is not luck. It is a question so uncomfortable that it got him pushed off a lunch table.

Hamming was a mathematician at Bell Labs. He refused to waste his lunch hour, so he ate at the physics table. Bardeen, Shockley, Brattain. All three would share a Nobel Prize for the transistor.

When that crowd moved on, he joined the chemists.

One day he sat down and asked: if what you are working on is not important, and it is not likely to lead to anything important, why are you working on it?

After that, he ate with the engineers.

That fall, one of the chemists, Dave McCall, stopped him in the hallway. The remark had gotten under his skin. He had spent the whole summer thinking about the important problems in his field.

Soon after, McCall was made head of his department and later joined the National Academy of Engineering. Hamming says he never heard of anyone else from that table again.

His conclusion: if you do not work on important problems, you will not do important work, except by the dumbest of dumb luck.

He took his own advice. His error-correcting codes still protect data in computer memory, and he won the Turing Award in 1968.

When did you last ask yourself why you are working on what you are working on?

[WealthLens @WealthLens_]

Sam Altman literally explained why a 19-year-old nobody got a meeting that established companies could not. It is not connections. It is not a genius pitch. It is a habit most people drop after the first ignored email.

In 2005 he left Stanford to build Loopt, a location app for phones. To exist, it needed a deal with a mobile carrier. Carriers did not work with startups.

So they kept trying. A different person. A different department. A different introduction. He says they tried about 30 different paths into that one company.

Eventually the key decision maker agreed to meet. In Altman's telling, the reason was simple: he wanted them to stop bothering him.

Loopt went on to land carrier deals, including Sprint.

Altman credits a phrase from Paul Graham that he thinks deserves far more respect: be relentlessly resourceful. Keep looking for new attack vectors on the problem in front of you.

He calls it one of the most important skills in life, and says it is surprisingly learnable. Most people stop at the first ignored email. At most the second.

Loopt itself never caught on with users and sold in 2012 for about $43 million. The habit stayed. This year OpenAI raised money at an $852 billion valuation.

How many times do you knock before you decide a door is locked?

[WealthLens @WealthLens_]

Elon Musk literally explained how he built his first company in 1995 with negative money. It is not a rich family. It is not venture capital. It is a hole he drilled through the floor.

He had just arrived at Stanford for a PhD. He deferred after two days to build something on the internet. His professor told him he did not think Elon would be coming back. It was their last conversation.

He had huge student debts. He could not afford both an apartment and an office, and the office was cheaper. So he rented the office, slept on a futon and showered at the YMCA. He says he was in the best shape of his life.

There was a tiny internet provider on the floor below. They drilled a hole through the floor, ran a cable down, and got online for about $100 a month.

The team was Elon, his brother from Canada, a friend of his mom, and three salespeople hired on commission through a newspaper ad. Half the people they pitched asked what the internet was.

But costs were so absurdly low that revenue beat expenses. That is what they showed investors.

In 1999 Compaq bought Zip2 for $307 million in cash. His share was about $22 million. He calls cash a currency he highly recommends.

What is the most ridiculous thing you did to keep costs near zero when you were starting out?

[WealthLens @WealthLens_]

Jeff Bezos literally explained why the startup everyone called Amazon.toast ended up burying its biggest rival. It is not money. It is not luck. It is one rule he gave 300 nervous employees the moment the giant showed up.

In 1995 Amazon was a handful of people in a converted garage. The office was so bad they held early meetings at a cafe down the street. Inside a Barnes & Noble.

Two years later Barnes & Noble launched its own website, days before Amazon's IPO. A Forrester analyst started calling the startup Amazon.toast.

Bezos admits they were worried. Barnes & Noble had the brand and the buying power. Most of his team had never faced a real competitor.

So he did two things. Amazon would eat any price gap, even at thinner margins, and get big fast.

And he gave the company one watchword. Obsess over customers, not competitors. Watch rivals, copy whatever helps customers, but never fixate on them.

In the 10 months after Barnes & Noble launched, he said, Amazon's revenue run rate went from $60 million to $260 million. Customers went from 340,000 to over 1.5 million.

In 2019 Barnes & Noble was sold to a hedge fund for about $683 million. This week Amazon is worth about $2.7 trillion.

Be honest. How much of your week goes to watching competitors instead of the people who actually pay you?

[GlobalBrief @_GlobalBrief_]

@WealthLens_ The "variable expense" line is the sharpest part. Consultants get paid to be smart about someone else's problem, never to live with being wrong about it

[WealthLens @WealthLens_]

Steve Jobs literally explained why some of the smartest people in business never really learn how business works. It is not intelligence. It is not effort. It is one thing they never have to do, and he compared the result to a piece of fruit.

Spring 1992. Jobs is running NeXT, seven years after being pushed out of Apple. He asks a room of MIT Sloan students where they came from.

Wall Street. Manufacturing. Then: how many from consulting? Hands go up. His reply: "Oh, that's bad."

Then he explains. He does not think consulting is evil. But if you never own something for a few years, never live with your own recommendations, never collect scar tissue from your own mistakes, you learn a fraction of what you could.

His image: a consultant gets a picture of a banana. Maybe a very accurate picture. But it is two-dimensional.

You can hang plenty of pictures on your wall. I worked in bananas. I worked in peaches. I worked in grapes. You never actually taste any of it.

Then the punchline. You are also a variable expense, and in hard times you find yourself variable.

The recording sat in MIT's archive for 26 years before it resurfaced in 2018.

Four years after that talk, Apple bought NeXT for $429 million and Jobs came back. Apple is now worth about $5 trillion.

Be honest. Are you tasting the banana right now, or just collecting pictures of it?

[GlobalBrief @_GlobalBrief_]

@WealthLens_ Straight A's for Lori, four years of pre-solved homework for him. Not a bad trade for a 17-year-old who couldn't legally drive in some states yet

[WealthLens @WealthLens_]

Jensen Huang literally explained how a 17-year-old who looked like a child landed a standing Sunday date in an engineering class of 250 students and three girls. It is not looks. It is not money. It is one line so nerdy it never should have worked.

He started at Oregon State at 16. He was the youngest in his class and looked it.

So he did the math. If he looked like a kid, her first impression had to be that he was smart. He decided to lean into it.

He walked up to his lab partner Lori, who was 19, and asked: do you want to see my homework?

Then he made her a deal. Do homework with me every Sunday and you will get straight A's.

It worked. He had a date every Sunday, and he made her do homework all day.

The trick, he admitted, was that he finished the homework before she arrived. For four years she thought he was a genius because he already knew every answer.

To make sure she married him, he added one more promise. By 30, he would be a CEO. He says he had no idea what he was talking about.

They married. He never took a business course, and he says he has still never written a business plan. In 1993, at 30, he co-founded Nvidia. This week it is worth about $5.5 trillion, the most valuable company on Earth.

What is the most ridiculous promise you made in your twenties that you actually kept?

[WealthLens @WealthLens_]

Milton Friedman literally explained why the most dangerous mistake a regulator can make is the one nobody ever sees. It is not approving a bad drug. It is not corruption. It is a mistake so quiet that the only people who could complain are no longer around.

He asked his audience to imagine being an FDA official deciding on a new drug.

You can make two mistakes. Approve a drug with harmful side effects, and your name is on every front page.

Or reject a drug that would have saved lives. Who complains? In his words, the people who would object are mostly going to be dead.

So every rational official leans the same way. Say no. Say later. Ask for one more study.

He went further. If aspirin had been submitted to a modern FDA in the 1890s, he argued, it would never have been approved.

After the 1962 Kefauver amendments forced drugs to prove effectiveness as well as safety, he said the number of new drugs approved each year was cut in half. Economist Sam Peltzman concluded the cost of delayed drugs exceeded the safety benefit many times over.

Friedman was careful. He did not call regulators bad people. He said they were you and me, responding to the incentives in front of them.

That is the uncomfortable part. Visible mistakes get punished. Invisible ones get promoted.

Where in your own work do you choose the safe no, because nobody will ever see what it cost?

[WealthLens @WealthLens_]

Warren Buffett literally explained why he refused to visit a company before paying $4 billion for it. It is not arrogance. It is not blind trust. It is one sentence he told the founder afterward that most buyers learn far too late.

In the fall of 2005 a letter arrived in Omaha. A page and a half. From a man Buffett had never heard of, about a company he had never heard of.

The writer was Eitan Wertheimer. The company was Iscar, an Israeli maker of metal cutting tools. His family would sell to only one buyer: Berkshire Hathaway.

Buffett says the idea jumped off the page. He emailed back. The family flew over.

In 2006 Berkshire bought 80% of the business at a $5 billion valuation. Buffett had not set foot in a single plant.

Wertheimer kept insisting he come to Israel to see the factories. Buffett replied that he did not even go to Council Bluffs, Iowa, the town across the river from Omaha.

After the deal closed he went. He calls it the greatest operation he has ever seen.

Wertheimer said that was exactly why he wanted him to come. Buffett told him that if he had come first, he would have paid more.

In 2013 Berkshire bought the last 20% for about $2.05 billion. The business had roughly doubled in value in seven years.

Be honest. Have you ever overpaid because you saw something in person and fell for it? A car, a house, a business.

[WealthLens @WealthLens_]

Carlos Brito literally explained why he ran the world's biggest brewer without an office, a driver or a jet. It is not that he was cheap. It is not a PR stunt. It is that he wanted every employee to think like an owner, and owners do not rent.

He compares most executives to people in a rental car. They take the speed bump fast just to see what happens, because someone else will live with the damage.

So he built a company where nothing felt rented.

No executive floor. He sat at one big table with his direct reports. He flew commercial to Russia, China and India and stayed in the same hotels as his staff.

When his company bought Anheuser-Busch for $52 billion in 2008, the American team had stock that vested a third each year.

Brito called that short-term. His version was a five-year cliff. Leave at four years and eleven months and you get zero.

A lot of people quit over it. He let them go.

His logic: if you plan to stay 30 years, what is five? And when he asks teams to cut non-working money in a tough year, he can say it with a straight face, because he is in the same seat on the same plane.

He ran AB InBev for over 15 years and made it the largest brewer on the planet.

Would you take a five-year cliff for real ownership, or is that a trap?

[WealthLens @WealthLens_]

Scott Cook literally explained why Intuit's biggest business came from data he ignored for five years. It is not vision. It is not a smarter plan. It is a survey answer he assumed was wrong.

Cook built Quicken to help families pay bills at home.

Then every customer survey said the same odd thing. Half of users ran it in an office. He figured they were answering wrong. He let it sit for five years.

When his team finally went to watch, the answer was on their desks. Small companies were keeping their books in a home checkbook app. The clerks doing it had never studied accounting. Real accounting software lost them at the first debit.

So Intuit built QuickBooks. Accounting software with no accounting in it.

By his own telling, the launch was a disaster. Half the features of rivals at twice the price, $99 against $39 and $49. A two-page ad in a million copies pulled four responses. Four. Then a database bug started wiping out customers' books.

He says it was the best-selling accounting software in America within two months anyway.

Cook's name for the lesson is simple: savor surprises. A strange number in your data is often the market telling you something you do not know.

In fiscal 2025, the Intuit segment built around QuickBooks brought in $11 billion.

His full Stanford talk is free online. It has fewer views than a mid-sized TikTok.