September 1998, greenwich, connecticut. a hedge fund with sixteen partners, two of them nobel laureates - merton and scholes - led by john meriwether, the legendary bond trader who built the team out of his old salomon brothers desk.
ltcm entered 1998 with $4.8B in capital. by late september it had $400M left. against that: over $100B in liabilities. leverage of 50 to 1. buffett would later say those 16 people probably had the highest combined IQ of any 16 people working together in one business in the country.
september 1: meriwether sends a letter to investors. capital has fallen to $2.3B. withdrawals capped at 12%, and not until december. september 22: capital is down to $600M. creditor banks are terrified - not of the losses, but of what liquidating the fund themselves would do to the market.
september 23, 1998, new york. the fed calls fourteen banks into its offices. buffett, goldman sachs, and AIG offer $250M for the partners and $3.75B into the fund. LTCM says no. hours later, the same fourteen banks put in $3.625B themselves, for 90% ownership.
weeks later, buffett explained to students in florida why the smartest people in finance had just gone broke:
“to make money they didn’t have and didn’t need, they risked what they did have and did need. that is just plain foolish.”
the fund was liquidated in 2000. a dollar invested in 1994 was worth pennies by the end of 1998.
a nobel prize doesn’t cover a margin call.
September 1998, greenwich, connecticut. a hedge fund with sixteen partners, two of them nobel laureates - merton and scholes - led by john meriwether, the legendary bond trader who built the team out of his old salomon brothers desk.
ltcm entered 1998 with $4.8B in capital. by late september it had $400M left. against that: over $100B in liabilities. leverage of 50 to 1. buffett would later say those 16 people probably had the highest combined IQ of any 16 people working together in one business in the country.
september 1: meriwether sends a letter to investors. capital has fallen to $2.3B. withdrawals capped at 12%, and not until december. september 22: capital is down to $600M. creditor banks are terrified - not of the losses, but of what liquidating the fund themselves would do to the market.
september 23, 1998, new york. the fed calls fourteen banks into its offices. buffett, goldman sachs, and AIG offer $250M for the partners and $3.75B into the fund. LTCM says no. hours later, the same fourteen banks put in $3.625B themselves, for 90% ownership.
weeks later, buffett explained to students in florida why the smartest people in finance had just gone broke:
“to make money they didn’t have and didn’t need, they risked what they did have and did need. that is just plain foolish.”
the fund was liquidated in 2000. a dollar invested in 1994 was worth pennies by the end of 1998.
a nobel prize doesn’t cover a margin call.
Jesse Livermore made $100 million in a single trade. then he lost it, and himself.
by october 1929 he had already made and lost three fortunes. the “boy plunger” started at 14, copying stock quotes onto a chalkboard at a boston brokerage, and by 22 he’d been banned from every bucket shop in new england for winning too consistently.
in the summer of 1929, while the Dow sat near 381 and the public borrowed $8.5B on margin to chase it higher, livermore built a short position. quietly. against the advice of nearly everyone around him.
black monday hits october 28th. the Dow drops 13% in a day. black tuesday follows - another 12%. livermore is on the other side of the panic. by the time the dust settles, his short nets him roughly $100 million, something north of $1.7 billion today.
he went home that night and told his wife, simply, they’d never have to worry about money again.
“there is nothing new in wall street,” he wrote. “there can’t be because speculation is as old as the hills.”
five years later, 1934: bankrupt for the fourth time. assets of $84,000 against debts of $2.5 million. the strategies that made him a legend in a panic didn’t survive a slow bleed.
on november 28, 1940, he checked into the sherry-netherland, wrote eight pages to his wife ending “i am a failure,” and ended his own life. he was 63.
livermore didn’t lose to the market. he lost to himself, repeatedly, and the market just kept score.
Jesse Livermore made $100 million in a single trade. then he lost it, and himself.
by october 1929 he had already made and lost three fortunes. the “boy plunger” started at 14, copying stock quotes onto a chalkboard at a boston brokerage, and by 22 he’d been banned from every bucket shop in new england for winning too consistently.
in the summer of 1929, while the Dow sat near 381 and the public borrowed $8.5B on margin to chase it higher, livermore built a short position. quietly. against the advice of nearly everyone around him.
black monday hits october 28th. the Dow drops 13% in a day. black tuesday follows - another 12%. livermore is on the other side of the panic. by the time the dust settles, his short nets him roughly $100 million, something north of $1.7 billion today.
he went home that night and told his wife, simply, they’d never have to worry about money again.
“there is nothing new in wall street,” he wrote. “there can’t be because speculation is as old as the hills.”
five years later, 1934: bankrupt for the fourth time. assets of $84,000 against debts of $2.5 million. the strategies that made him a legend in a panic didn’t survive a slow bleed.
on november 28, 1940, he checked into the sherry-netherland, wrote eight pages to his wife ending “i am a failure,” and ended his own life. he was 63.
livermore didn’t lose to the market. he lost to himself, repeatedly, and the market just kept score.
jim donovan stands at a lectern inside the university of virginia school of law and hands a room of students eleven rules for keeping a client for life. the video has pulled millions of views on youtube. business schools charge six figures to teach less.
november 6, 2015. donovan is vice chairman of global client coverage at goldman sachs, advising some of the firm’s largest corporate and individual clients. he is also, unpaid, an adjunct professor - 14 years of checks signed straight back to the law school foundation.
he joined goldman in 1993, the year he finished harvard law. seven years later he makes partner. he tells the room none of it was instinct. his first two years he carried one senior banker’s bags and ran his photocopies, watching how the man worked a room before he was ever allowed to speak in one.
the rules that follow are unglamorous: find the best relationship manager in the building and shadow him. put ego aside. be available - first impressions die hard. take a position, never equivocate.
“there’s no more powerful way to establish credibility with a client,” he says, than handing them advice that costs you the fee.
then the point most bankers and lawyers miss entirely: they treat signing day as the finish line. donovan’s read is the opposite - the deal is the audition, not the payoff. stay close after the ink dries and the next mandate finds you. walk away and someone else gets the call.
the fee closes on one transaction. the trust compounds across the next ten.
jim donovan stands at a lectern inside the university of virginia school of law and hands a room of students eleven rules for keeping a client for life. the video has pulled millions of views on youtube. business schools charge six figures to teach less.
november 6, 2015. donovan is vice chairman of global client coverage at goldman sachs, advising some of the firm’s largest corporate and individual clients. he is also, unpaid, an adjunct professor - 14 years of checks signed straight back to the law school foundation.
he joined goldman in 1993, the year he finished harvard law. seven years later he makes partner. he tells the room none of it was instinct. his first two years he carried one senior banker’s bags and ran his photocopies, watching how the man worked a room before he was ever allowed to speak in one.
the rules that follow are unglamorous: find the best relationship manager in the building and shadow him. put ego aside. be available - first impressions die hard. take a position, never equivocate.
“there’s no more powerful way to establish credibility with a client,” he says, than handing them advice that costs you the fee.
then the point most bankers and lawyers miss entirely: they treat signing day as the finish line. donovan’s read is the opposite - the deal is the audition, not the payoff. stay close after the ink dries and the next mandate finds you. walk away and someone else gets the call.
the fee closes on one transaction. the trust compounds across the next ten.
tony robbins is 24 and living in a 400-square-foot apartment in venice beach
he checks his bank account. $21. he’s washing dishes in his own bathtub because the water heater doesn’t work
this is 1983. no book deal. no seminars. no “unleash the power within.” just a guy who flunked out of a normal career path, cleaning up other people’s messes to make rent
he decides one thing is still his to control — not the bank balance, not the apartment, but the next sixty seconds of his own state of mind. he starts practicing it daily. calls it “personal power”
three years later guthy-renker puts him on late-night tv. 24 cassettes. 30 days. one promise: change your physiology, change your results
“there is a powerful driving force inside every human being that once unleashed, can make any vision, dream, or desire a reality”
the infomercial runs for over a decade. millions of copies sold. robbins goes from janitor to the guy fortune 500 CEOs and sitting presidents call for advice. today his seminars run tens of thousands of dollars a seat and his consulting fee is reported near seven figures a day
my read: the $21 was never the constraint. broke people stay broke by waiting for the number to change. robbins just changed what he did in the sixty seconds after finding out he had none.
a seat in tony robbins’s living room used to cost $125. today, a single day with him runs $1 million.
the difference is thirty years and a videotape nobody was supposed to see again.
it’s filmed inside his own house, decades before billionaires started paying seven figures just to sit in the room with him. 21 minutes, flip chart, no stage, no lighting rig. just robbins explaining why people say no.
two excuses, he tells the room. not enough time. not enough money. neither is true. the real objection is state, not money - they don’t believe it’s worth it yet.
his fix: attack and confess. don’t argue the objection, confess your own first. “I had a chance to go to this thing six months ago and I didn’t go until two months ago,” he tells them. “I can’t even imagine the time I lost.” nobody argues back. the room goes quiet.
then the yes train - stack small agreements until the final ask is harder to refuse than to accept. five yeses deep before anyone signs anything.
$125 bought that whole framework in a living room outside chicago. $1 million buys it today, unchanged, delivered by a man who has since sold the same eleven-minute idea to a generation of ceos.
my read: the price went up 8,000x. the material didn’t change at all.
that’s the business - not the secret, the scarcity of access to it.
a seat in tony robbins’s living room used to cost $125. today, a single day with him runs $1 million.
the difference is thirty years and a videotape nobody was supposed to see again.
it’s filmed inside his own house, decades before billionaires started paying seven figures just to sit in the room with him. 21 minutes, flip chart, no stage, no lighting rig. just robbins explaining why people say no.
two excuses, he tells the room. not enough time. not enough money. neither is true. the real objection is state, not money - they don’t believe it’s worth it yet.
his fix: attack and confess. don’t argue the objection, confess your own first. “I had a chance to go to this thing six months ago and I didn’t go until two months ago,” he tells them. “I can’t even imagine the time I lost.” nobody argues back. the room goes quiet.
then the yes train - stack small agreements until the final ask is harder to refuse than to accept. five yeses deep before anyone signs anything.
$125 bought that whole framework in a living room outside chicago. $1 million buys it today, unchanged, delivered by a man who has since sold the same eleven-minute idea to a generation of ceos.
my read: the price went up 8,000x. the material didn’t change at all.
that’s the business - not the secret, the scarcity of access to it.
a seat in tony robbins’s living room used to cost $125. today, a single day with him runs $1 million.
the difference is thirty years and a videotape nobody was supposed to see again.
it’s filmed inside his own house, decades before billionaires started paying seven figures just to sit in the room with him. 21 minutes, flip chart, no stage, no lighting rig. just robbins explaining why people say no.
two excuses, he tells the room. not enough time. not enough money. neither is true. the real objection is state, not money - they don’t believe it’s worth it yet.
his fix: attack and confess. don’t argue the objection, confess your own first. “I had a chance to go to this thing six months ago and I didn’t go until two months ago,” he tells them. “I can’t even imagine the time I lost.” nobody argues back. the room goes quiet.
then the yes train - stack small agreements until the final ask is harder to refuse than to accept. five yeses deep before anyone signs anything.
$125 bought that whole framework in a living room outside chicago. $1 million buys it today, unchanged, delivered by a man who has since sold the same eleven-minute idea to a generation of ceos.
my read: the price went up 8,000x. the material didn’t change at all.
that’s the business - not the secret, the scarcity of access to it.
Quick fact-check before I stylize this: I confirmed the broadcast — January 4, 1960, on BBC’s Panorama, titled “Challenge of the Sixties.” But it wasn’t Oppenheimer alone in a 9-minute solo segment — Nehru also appeared in that same episode, and I can’t verify the exact “9 minutes,” “one camera, no music,” or lines like “the gloom has no floor.” What I did find is a real, verifiable quote from the broadcast. I used that instead of the unverified lines so nothing here gets fabricated.
congress destroys robert oppenheimer’s career in 1954. a security hearing strips his clearance, ends his access to the work he built.
six years later, the bbc puts him in a chair at princeton anyway.
january 4, 1960. panorama airs an episode called “challenge of the sixties,” predicting the decade ahead. nehru appears in the same broadcast. so does oppenheimer, cut off from government work, still asked what comes next.
“i hope that these ten years will see a growing effort to transform the world,” he says.
he does not mention the bomb.
the decade obliges. murray gell-mann and george zweig propose quarks in 1964. arno penzias and robert wilson detect the cosmic microwave background in 1965, proof the universe had a beginning. by 1966, the genetic code is fully deciphered.
three of the largest discoveries in modern science, inside the window he pointed to, from a man congress had just finished humiliating.
my read: the hearing room measured his loyalty. it never measured his judgment. the sixties measured that instead.
washington took his clearance. it never had a claim on what he knew.
Quick fact-check before I stylize this: I confirmed the broadcast — January 4, 1960, on BBC’s Panorama, titled “Challenge of the Sixties.” But it wasn’t Oppenheimer alone in a 9-minute solo segment — Nehru also appeared in that same episode, and I can’t verify the exact “9 minutes,” “one camera, no music,” or lines like “the gloom has no floor.” What I did find is a real, verifiable quote from the broadcast. I used that instead of the unverified lines so nothing here gets fabricated.
congress destroys robert oppenheimer’s career in 1954. a security hearing strips his clearance, ends his access to the work he built.
six years later, the bbc puts him in a chair at princeton anyway.
january 4, 1960. panorama airs an episode called “challenge of the sixties,” predicting the decade ahead. nehru appears in the same broadcast. so does oppenheimer, cut off from government work, still asked what comes next.
“i hope that these ten years will see a growing effort to transform the world,” he says.
he does not mention the bomb.
the decade obliges. murray gell-mann and george zweig propose quarks in 1964. arno penzias and robert wilson detect the cosmic microwave background in 1965, proof the universe had a beginning. by 1966, the genetic code is fully deciphered.
three of the largest discoveries in modern science, inside the window he pointed to, from a man congress had just finished humiliating.
my read: the hearing room measured his loyalty. it never measured his judgment. the sixties measured that instead.
washington took his clearance. it never had a claim on what he knew.
alibaba gives away a model that matches claude’s flagship. it runs on a $700 used graphics card.
hugging face publishes its state of open models report on august 14, 2026. the numbers are not close.
qwen crosses 3 billion downloads in six months. google: 418 million. meta: 227 million. combined, qwen still beats them roughly 4 to 1.
three days earlier, alibaba releases Qwen3.8-27B under apache 2.0. 27B parameters, native vision, 262,000 token context. quantized to 17GB, it fits on hardware developers already own. it pulls 3 million downloads in its first weekend.
deepswe scores go from 13.3 to 42.2 in one release cycle. software engineering: 49.3 to 79.0.
“qwen has become part of the default workflow for developers deciding what models to fine-tune and deploy,” hugging face writes.
apache 2.0 means the license cannot be revoked. once the weights are on a drive, they stay there. 300,000-plus derivative models already exist. roughly 200 more appear every day.
washington spends four years building export controls around chips, fabs, entity lists - every one of them assumes a chokepoint. a shipment that can be stopped. a company that can be told no.
there is no chokepoint for a file copied 3 billion times.
my read: consumer habits change in an afternoon. infrastructure choices last a decade, because everything built on top has to be rewritten to undo them.
scarcity was never the product. it was the business model.
Jensen Huang: “people are teaching their agents to fully run a business and make money”
support. sales. operations. coding. marketing.
agents are taking every seat at the table.
but here’s what nobody’s saying out loud
if your agent can’t remember yesterday, it’s not an employee. it’s a goldfish with an api key.
it re-introduces itself every session. re-learns your preferences every session. makes the same mistake it made monday, again on tuesday.
you wouldn’t hire a human who forgot your name every morning. why is that acceptable for something running your business?
99% of agents ship with zero long-term memory. no continuity. no compounding. just a very fast, very confident amnesiac.
the companies actually winning with agents aren’t the ones with the best prompts. they’re the ones who solved memory first.
here’s how to build an agent that remembers - and actually earns the seat
in 1979, clay christensen graduates from harvard business school. one of his classmates is jeffrey skilling.
by 2000, skilling runs enron, stock trading at $90 a share, wall street’s favorite growth story. christensen is three miles away, teaching strategy at HBS.
a year later enron is worth 26 cents a share. $74B in market value gone in months. skilling is convicted on 19 federal counts, sentenced to 24 years, later reduced to 14. he serves 12 before release in 2019.
christensen keeps a different set of numbers. two of the 32 rhodes scholars in his own class spend time in jail. by his 25th and 30th HBS reunions, classmate after classmate arrives divorced, estranged from children nobody planned to lose.
none of them graduated with that strategy. they got there one rational decision at a time - the meeting that beat the recital, the deal that beat the dinner. work pays out immediately. a child takes 20 years to show you the return.
“the metric by which god will assess my life isn’t dollars but the individual people whose lives I’ve touched.”
business review essay in 2010, then a book with james allworth and karen dillon in 2012. that same year doctors find cancer in his thyroid and pancreas. he keeps teaching anyway. he dies january 23, 2020, at 67.
skilling once earned $100M in a single year running enron. christensen never came close.
only one of them is still worth studying.
charlie munger spent six decades building a reputation for blunt, airtight answers - the kind that ended debates instead of starting them. one of his best had nothing to do with money.
omaha, an early-2020s berkshire hathaway annual meeting. munger - vice chairman of the company since 1978, buffett’s partner for nearly six decades, a man who had already outlived two marriages and the loss of a son.
a woman in the crowd asks something simple. no trap in it. is it true, mr. munger, that you have eight children?
eight - across two marriages, folded into one family over fifty years. a fact he could have just confirmed.
he doesn’t.
“so far.”
buffett has retold that exchange for years. it’s shorter than any breakdown of berkshire’s balance sheet, and it explains munger better than most of them do. he died at 99, november 28, 2023 - still working the week before.
my read: a great answer isn’t measured by completeness. it’s measured by how precisely it lands in the moment it’s given. munger understood that at the negotiating table and at the podium, in the same breath.
numbers age. a good line doesn’t.
@MyWestLord Gates in 2004 at MIT talked about a vacuum robot and spam. The AI and biology he mentioned in passing are now worth more than all of Microsoft. Spam is still alive
charlie munger spent six decades building a reputation for blunt, airtight answers - the kind that ended debates instead of starting them. one of his best had nothing to do with money.
omaha, an early-2020s berkshire hathaway annual meeting. munger - vice chairman of the company since 1978, buffett’s partner for nearly six decades, a man who had already outlived two marriages and the loss of a son.
a woman in the crowd asks something simple. no trap in it. is it true, mr. munger, that you have eight children?
eight - across two marriages, folded into one family over fifty years. a fact he could have just confirmed.
he doesn’t.
“so far.”
buffett has retold that exchange for years. it’s shorter than any breakdown of berkshire’s balance sheet, and it explains munger better than most of them do. he died at 99, november 28, 2023 - still working the week before.
my read: a great answer isn’t measured by completeness. it’s measured by how precisely it lands in the moment it’s given. munger understood that at the negotiating table and at the podium, in the same breath.
numbers age. a good line doesn’t.
charlie munger spent six decades building a reputation for blunt, airtight answers - the kind that ended debates instead of starting them. one of his best had nothing to do with money.
omaha, an early-2020s berkshire hathaway annual meeting. munger - vice chairman of the company since 1978, buffett’s partner for nearly six decades, a man who had already outlived two marriages and the loss of a son.
a woman in the crowd asks something simple. no trap in it. is it true, mr. munger, that you have eight children?
eight - across two marriages, folded into one family over fifty years. a fact he could have just confirmed.
he doesn’t.
“so far.”
buffett has retold that exchange for years. it’s shorter than any breakdown of berkshire’s balance sheet, and it explains munger better than most of them do. he died at 99, november 28, 2023 - still working the week before.
my read: a great answer isn’t measured by completeness. it’s measured by how precisely it lands in the moment it’s given. munger understood that at the negotiating table and at the podium, in the same breath.
numbers age. a good line doesn’t.