Neighbor @Neighboritto

Seventeen years before federal agents finally arrested the man behind the largest financial fraud in American history, one person had already solved the entire case from his own desk, using nothing but a phone, a calculator, and a stack of paperwork nobody else had bothered to check. His name was Edward Thorp, the same mathematician who decades earlier had become famous for working out how to beat blackjack using pure probability, and who later built one of the first successful quantitative hedge funds on Wall Street. In 1991 the consulting firm McKinsey and Company asked him to review their pension and profit sharing investments, one of which was quietly managed by a highly respected money manager delivering remarkably steady returns, the kind of consistency almost nobody else in the industry could match. Thorp asked to visit the firm's offices directly and was refused at the door, so he requested the actual trading records instead. The math simply did not add up. On one specific day in April 1991, the fund claimed to have purchased 123 call option contracts on a single stock, but when Thorp checked, the entire market had only traded about twenty of those contracts that day, total, for everyone. He checked other trades the same way through contacts at a major Wall Street firm and kept finding the exact same problem, transactions claimed on paper that the real market could never have actually handled. That fund manager's name was Bernie Madoff, the same man who would go on to run the largest Ponzi scheme in American history, finally collapsing in December 2008 and wiping out tens of billions of dollars in client money. Thorp convinced McKinsey to pull every dollar out immediately. He never went to federal regulators. He simply told people in his own circle exactly what he had found, laid out clearly, and at least one person who heard the full story chose to keep investing with Madoff anyway, right up until the scheme finally collapsed seventeen years later. One of the sharpest mathematical minds in the country had already cracked the biggest financial fraud in American history nearly two decades early, laid the evidence out in plain language for people he trusted, and some of them looked directly at the math and still chose to believe the story instead. If someone can show you hard proof that something is mathematically impossible and you still choose to believe it anyway, what does that tell you about how much of what people invest in actually gets checked before the money goes in?

@0xlexlab both methods worked by measuring physical wear the house treated as pure chance

A doctor won so much money from a single casino that the casino banned him from an entire country, and when he fought back, he actually won that fight too. His name was Richard Jarecki, a university professor in Germany who got curious about something almost nobody else seemed to notice. Casinos replaced their cards, dice and balls every single night, but the roulette wheels themselves, expensive precision machines, often stayed in service for decades without ever being swapped out. Jarecki figured that decades of constant use had to wear every wheel down just slightly differently, creating tiny mechanical biases nobody was bothering to measure. So he and his wife Carol recruited a small team of helpers and had them sit at specific wheels for hours at a time, sometimes tracking as many as ten thousand spins on a single wheel before ever placing a real bet, later feeding the numbers into a statistical model he built on an early computer at the University of London. Once they found a wheel with a genuine bias, Jarecki would quietly start betting heavily on its favored numbers. Between 1964 and 1969 he pulled in somewhere around one and a quarter million dollars, worth roughly eight or nine million dollars today, enough to buy a Rolls Royce and a luxury apartment near the casino at San Remo where he won the bulk of his fortune. Dealers there started calling him the Roulette Tornado. The casino eventually got so sick of losing to him that they tried to ban him from the entire country of Italy. Jarecki fought back through the courts, and within weeks the ban was overturned and he was right back at the tables. Casinos eventually had no choice but to completely change how roulette wheels were manufactured worldwide, tightening the tolerances on every wheel built after him, just to make sure nobody could ever do again exactly what one patient doctor and his wife did with nothing but notebooks, a university computer, and thousands of hours of watching a wheel spin. If one doctor and his wife quietly forced an entire global industry to redesign how its equipment gets manufactured, how many other industries out there do you think still have a flaw nobody has bothered to patiently sit down and actually measure yet?

@TheOnyxis Patience like that is what turns hidden flaws into real advantages

@verumidest The computer story was the cover. The real edge was notebooks and tens of thousands of spins logged by hand

@Kv0r1n overestimating the edge and betting double Kelly can drive expected log-growth to zero or negative

In 1956 a scientist working for one of the most powerful companies in America wrote a paper that quietly contained one of the most valuable math formulas in history, and his own employer made him disguise the title so nobody would realize what it actually was. His name was John Kelly, a physicist at Bell Labs, the research arm of the telephone company AT&T. Kelly had been thinking about an old news story involving a scam on a television quiz show, where contestants secretly received advance answers over a signal, and it made him wonder about a completely different problem. If a gambler received secret tips about horse races through an unreliable telephone wire, exactly how much of his own money should he bet on each tip to grow his fortune as fast as mathematically possible without ever going broke. Working from his colleague Claude Shannon's theories about sending information through noisy channels, Kelly worked out a precise formula that answered that exact question. He originally wanted to title the paper Information Theory and Gambling, but Bell Labs executives got nervous. Bookmakers and casinos were actual paying customers of the phone company at the time, and the company did not want its name publicly tied to a formula for winning bets. So Kelly quietly retitled it to something almost nobody would ever want to read, A New Interpretation of Information Rate, and it was published in 1956 looking exactly like a dry engineering paper about telephone static. Kelly himself reportedly never once used the formula to gamble for his own profit. Years later, though, a young mathematician named Edward Thorp found the paper, realized exactly what it actually was, and used it to work out precisely how much to bet on each hand while counting cards in blackjack. From there it spread even further, quietly becoming one of the mathematical tools some of the most successful investors in modern history are said to lean on, including Warren Buffett himself. One of the most valuable formulas in the history of gambling and investing almost never saw daylight because a phone company was embarrassed about what it implied, and today it still sits buried in plain sight behind a title boring enough that almost nobody who was not specifically looking for it would ever bother to read it. If one of the most valuable formulas in modern finance was deliberately disguised behind a boring title just to avoid bad publicity, how many other genuinely important ideas out there do you think are hiding in plain sight today simply because nobody gave them an interesting enough name?

There is a betting system that feels so mathematically bulletproof almost everyone who hears it for the first time cannot find a single flaw in it, and it has personally bankrupted a world famous lover, a legendary artist, and one of the oldest banks on earth. The idea is almost insultingly simple. Bet one unit on anything with roughly even odds, like red on a roulette wheel. If you lose, double your next bet. Keep doubling after every single loss, and the moment you finally win, that one win recovers every previous loss plus your original profit, guaranteed, as long as you never run out of money or hit a betting limit. The strategy actually got its name from an eighteenth century British gambling house owner, though the first clear description of someone using it comes from the memoirs of Giacomo Casanova, who described doubling his bets at a Venice casino in 1754 and winning for days straight during a lucky streak. Casanova got lucky that particular carnival season, but he was not lucky forever, and neither was the famous artist Marcel Duchamp, who read Casanova's memoirs, became personally obsessed with the system, and actually designed and sold physical bonds in 1924 meant to fund a gambling operation built around it, a strange art project that also lost money. The mathematical flaw hiding underneath the whole idea has a formal name, gambler's ruin, a proven result showing the strategy only actually works with infinite money and no betting limits, two things that do not exist anywhere in the real world. Casino betting limits exist specifically to stop this exact system, because losing just ten times in a row, which happens far more often than people expect, means your next bet already has to be over a thousand times your original wager. The most expensive example of this mistake did not even happen inside a casino. In 1995 a British trader named Nick Leeson, managing money for one of the oldest banks in the world, started quietly doubling down on failed trades using the exact same logic, convinced the market simply had to turn back in his favor eventually. It never did in time. His losses eventually passed eight hundred million pounds and single handedly collapsed Barings Bank, a 233 year old institution that had once helped finance the Napoleonic Wars and personally banked for the British royal family. People keep reinventing the exact same doubling strategy every single generation, convinced they finally found the one flaw nobody else noticed. The math has had the same answer since before the United States existed, and it has never once changed. If a mathematically proven flaw has personally ruined a legendary lover, a famous artist and one of the oldest banks on the planet, why do you think the exact same strategy keeps getting reinvented by somebody new every single generation anyway?

Long before any computer ever set foot inside a casino, four complete unknowns quietly solved a problem that mathematicians and gamblers had been arguing about for over a hundred years, using nothing but hand cranked adding machines, and almost nobody today has ever heard their names. Their names were Roger Baldwin, Wilbert Cantey, Herbert Maisel and James McDermott, four Army mathematicians stationed together at the Aberdeen Proving Ground in Maryland in the early 1950s. It started almost by accident, when Private Baldwin was dealing a casual blackjack game with friends and had a sudden thought, that the dealer's moves in a real casino were not random at all, they followed a fixed, predictable set of rules. If that was true, Baldwin realized the entire game could theoretically be solved with math. The problem was that computers barely existed yet, so the four of them spent their nights and weekends over the next year and a half using nothing but military desk calculators, the kind people at the time still called adding machines, grinding through what later came out to roughly twelve combined man years of raw calculation. What came out the other end was the first mathematically accurate basic strategy for blackjack ever produced, along with the earliest legitimate version of a card counting concept, all published quietly in 1956 inside an academic journal called the Journal of the American Statistical Association, followed a year later by a short ninety two page book that barely sold any copies at all. Here is the part almost nobody realizes. Years later a young mathematician named Edward Thorp came across their paper, used their own hand calculations to verify his ideas, and built directly on their work to publish Beat the Dealer in 1962, the book that eventually made blackjack strategy famous worldwide and helped inspire everything from professional card counting teams to an entire Hollywood movie industry built around the game. The four men who actually solved the math first, using nothing but mechanical calculators and their own free time, mostly faded quietly back into ordinary careers. People assume the math behind beating blackjack belongs to whichever gambler eventually got famous for it. It actually belongs to four army guys with adding machines who solved the hardest part first and never got the credit for it. If four unknown soldiers with nothing but mechanical calculators could out think an entire industry built on keeping its odds secret, how many other supposedly impossible problems out there are just waiting for someone patient enough to actually sit down and do the math?

@0xlexlab secrecy wins one table; publishing builds the fund that outlasts every table

There is a man who beat blackjack with pure math years before the person history actually gives credit for it, and almost nobody has ever heard his name because he never wanted anyone to know. His name was Jess Marcum, a physicist doing classified nuclear research for the RAND Corporation in the years after World War Two. In 1949, a coworker who liked to gamble on weekends talked him into joining a trip to Las Vegas, and once Marcum actually sat down and watched a blackjack game, his instincts as a scientist kicked in immediately. He suspected the odds had to shift depending on which cards were already gone from the deck, so he worked out the math by hand, reportedly later refining it using an early IBM computer, and eventually built a system that gave him something close to a three percent edge over the house, an enormous advantage in a game everyone else assumed ran on pure luck. He started quietly beating casinos across Las Vegas so consistently that pit bosses eventually banned him outright, and the ban followed him internationally, all the way to Cuba. Here is the part most people never hear. This all happened more than a decade before Edward Thorp, the mathematician usually credited today as the father of card counting, ever published a single word on the subject in 1962. Marcum simply never told anyone how he did it. He avoided publicity, never wrote a book, and reportedly took the exact details of his system and the true size of his winnings to his grave in 1992. People spend their whole lives chasing credit for things they barely accomplished. This man quietly solved one of the most famous problems in modern gambling years before anyone else, and chose to let history hand the credit to somebody else entirely. If the man actually responsible for beating one of the most famous casino games in history never wanted anyone to know his name, how many other breakthroughs out there do you think are sitting with the wrong person's name attached to them?

@batagonx Thorp did mention the early Southern California counter in Beat the Dealer, so the nobody heard of him claim overstates it

@xxlazy57 yep

@verumidest once the edge was public, casinos rewrote the rules to erase it

@verumidest Convincing AI demos still need independent verification before any real portfolio decisions

Tavus gave me early access to Griffin I had just tried another agent, then went straight to Tavus and decided to run an experiment. I didn't ask anything abstract. I asked Griffin to advise me on my investment portfolio. It was unforgettable. Normally you type a request and wait for text. This was different: you just talk, and on the other side someone listens, looks at you and reacts to every word. I kept catching myself forgetting who I was talking to. And the topic wasn't a light one, because it was my money and my decisions. Now I think differently about how we'll interact with AI. Fewer text boxes, more real conversation. It's only early access, but the impression is already strong. If you want to see where this is heading: tavus.io/griffin @tavus @hassaanraza

Tavus @tavus

a week ago

Introducing Griffin, the first model to pass the video Turing test. 48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video. It’s the first Human Interaction Model (HIM).

3K 4K 39K 21.0M 27K

@batagonx the empire ran quietly for years; the record fine only marked its public ending

There was a man who once controlled how fast an entire country found out who won a horse race, and for years he quietly used that tiny sliver of time to build one of the largest illegal empires in American history. His name was Moses Annenberg. Back before television, before the internet, the only way a bookie sitting hundreds of miles from a racetrack could learn who had actually won a race was through a telegraph wire service relaying results city to city. Annenberg realized that whoever controlled that wire controlled every single bookie who depended on it, so he bought up competitor after competitor until his company was piping results from nearly thirty racetracks into bookmaking operations spread across thirty nine states, reaching all the way into Canada, Mexico and Cuba. Bookies who refused to pay his subscription fee got cut off entirely or fed the results with just enough of a delay for sharper operators to bet against them before the real numbers ever arrived. To keep the whole operation running smoothly, Annenberg reportedly paid Al Capone's organization about one million dollars a year in protection money just to be left alone. Federal investigators eventually caught up with him, not for the wire service itself, but for the mountain of cash it quietly generated. In 1939 he was indicted for evading more than three million dollars in income taxes, and the following year he pleaded guilty and stood before the exact same judge who had once sent Al Capone to prison. He walked away with an eight million dollar fine, the largest single tax fraud penalty in American history at that time. His son Walter inherited what remained of the family business and spent the rest of his life trying to bury his father's name under decades of relentless philanthropy, eventually building one of the largest charitable foundations in the country, one still quietly funding schools, hospitals and universities today. If one of the most respected philanthropic names in the country today was built on top of an illegal information monopoly and a fortune paid out in protection money, how many other respectable fortunes out there do you think are hiding a very similar story underneath them?

@useKled If this really works, public datasets will soon just be a warm-up

A guy with degrees from Yale and Harvard quit his job as a senior vice president at a stock exchange to get banned from casinos all over the world. His biggest bet on one single hand was 12,000 dollars. And then he did something no card counter did before: he sued the casinos and won. His name is Ken Uston, and he did not do it alone. He ran a whole team of card counters. One person sits at the table and only counts, and when the deck gets hot he signals a big player who walks in and drops the huge bets. The math behind it is simple: running count divided by the number of decks left gives the true count, and when that number is high the player has the edge, not the house. One team he joined won about half a million dollars before the Sands in Las Vegas spotted them. Later he kept coming back in different costumes, and 60 Minutes even made a story about him in 1981. Then came the court fight. A New Jersey court ruled that casinos could not ban a player just for counting cards unless the state had an official rule for it. The casinos answered the only way they could: more decks in the shoe and worse rules for the player. So the casino paid lawyers for years to lose against a man with a brain and a simple division. Be honest, could you count cards in your head under pressure while a pit boss stares at you? Real 1981 news report about the fight between New Jersey casinos and card counters, with Uston ↓