Vexo

[Vexo @Vexo200]

Benoit Mandelbrot walked onto a stage at 85, apologized for having to sit down, and then told the room that the length of a coastline is not a large number or a small number. It is not a number at all.

The talk is TED, February 2010. He died that October.

It has been sitting free on TED and YouTube ever since. Seventeen minutes. He first published the argument in 1967 and had already made it on that same stage in 1984.

The idea is one line. Some quantities are not properties of the object being measured. They are properties of the ruler.

Shorten the ruler and a coastline gets longer, with no ceiling. There is no true length waiting underneath. The number was never there.

3:20. A friend hands him a picture of a curve to bug him and asks how rough it is. He glances at it and says just short of 1.5. The real value is 1.48. He mentions how long the answer took him: no time at all.

4:40. He arrives at coastlines and lakes and picks his word carefully. Not approximation. Not simplification. Complete fallacy, there is no such thing.

5:30. He brings up the human lung. Anatomists were arguing whether its inner surface is the size of one basketball court or five. He treats the size of the disagreement as the finding rather than the error.

12:50. Two curves on screen. Standard and Poor's in blue. The same index in red with its five largest discontinuities removed. Five days, out of decades, and the two lines describe different worlds.

The thing to watch is where he draws the border between what he saw and what he proved. He measured Brownian motion, got 1.33, and could not prove it. Three mathematicians spent twenty years closing that gap and one of them took a Fields medal for it. He says this out loud, on a stage, about his own work, with no visible discomfort.

It lands harder in 2026 than it did in 2010. Any tool now returns a number for any question put to it, instantly and with total confidence. Separating the numbers that belong to the thing from the numbers that belong to the ruler is the part nobody automated.

He spent fifty years putting numbers on things everyone had called unmeasurable, which is exactly why he was the one qualified to say a coastline has no length.

[Sammy @sammy_capital]

https://t.co/cOYkpn7mej

[Vexo @Vexo200]

@isssa0x bro bought an entire block just for the office ngl

[Vexo @Vexo200]

An event can have probability zero and still happen. John Tsitsiklis shows why in the first hour of MIT's probability course, using nothing but a square and a dart.

He has taught 6.041 at MIT for decades. This is lecture one, filmed in 2010 and sitting free on MIT OpenCourseWare since 2012. Close to one and a half million people have opened it.

The setup is a perfect dart thrower and a square target. The dart always lands inside, but where it lands is random, and every point in the square counts as a possible outcome.

That breaks the obvious plan. If probability were handed out point by point, each point would get zero, and zero for every outcome tells you nothing. So probability gets attached to regions instead, and a region's probability is simply its area.

20:09. He introduces the square and says out loud that this sample space is infinite, not the tidy sixteen outcomes of the dice roll he just finished.

21:36. He starts to say probability is assigned to outcomes, stops himself mid sentence with well, not exactly, and rebuilds the whole idea around sets.

21:49. He asks the chance of hitting one exact point to infinite precision and answers zero, even though the dart is certain to land on some such point.

44:40. He finds the probability that the two coordinates sum to less than one half by drawing the triangle under a line and reading off its area, one eighth.

Watch the seam between his two examples. In the dice world every outcome carries real weight, one sixteenth each, and he adds them up like coins. In the dart world he quietly drops that habit, because the thing that actually happens is a point the model swears almost never happens.

He never resolves the tension in this lecture, he only names it and moves on, which is why a dart landing on a zero probability point is the cleanest picture of what the rest of the course exists to handle.

[Vexo @Vexo200]

An event can have probability zero and still happen. John Tsitsiklis shows why in the first hour of MIT's probability course, using nothing but a square and a dart.

He has taught 6.041 at MIT for decades. This is lecture one, filmed in 2010 and sitting free on MIT OpenCourseWare since 2012. Close to one and a half million people have opened it.

The setup is a perfect dart thrower and a square target. The dart always lands inside, but where it lands is random, and every point in the square counts as a possible outcome.

That breaks the obvious plan. If probability were handed out point by point, each point would get zero, and zero for every outcome tells you nothing. So probability gets attached to regions instead, and a region's probability is simply its area.

20:09. He introduces the square and says out loud that this sample space is infinite, not the tidy sixteen outcomes of the dice roll he just finished.

21:36. He starts to say probability is assigned to outcomes, stops himself mid sentence with well, not exactly, and rebuilds the whole idea around sets.

21:49. He asks the chance of hitting one exact point to infinite precision and answers zero, even though the dart is certain to land on some such point.

44:40. He finds the probability that the two coordinates sum to less than one half by drawing the triangle under a line and reading off its area, one eighth.

Watch the seam between his two examples. In the dice world every outcome carries real weight, one sixteenth each, and he adds them up like coins. In the dart world he quietly drops that habit, because the thing that actually happens is a point the model swears almost never happens.

He never resolves the tension in this lecture, he only names it and moves on, which is why a dart landing on a zero probability point is the cleanest picture of what the rest of the course exists to handle.

[Vexo @Vexo200]

@Zyron5m Robert Shiller certainly deserves respect

[Vexo @Vexo200]

John Bogle wrote in his Princeton thesis that fund managers should serve investors in the most economical and honest way possible. Fifteen years later he broke every rule in it.

He was hired in 1951, straight out of college, by the man who had founded the Wellington Fund in 1928.

When the performance slipped in the sixties, Bogle was handed the job of saving it. It was the go-go era. He decided the fund needed a hot manager of its own.

Assets fell from two billion dollars to four hundred million.

The management company fired him.

Sixty years later he sits down in front of a camera and walks through the whole thing himself. He does not soften it. He says he was overly self confident, that he ignored the lessons of history, and that what he did was a really dumb thing.

Then he explains the part almost nobody knows, which is how he came back.

They could fire him from the management company. They could not remove him from the board of the fund. A fund and its manager are two separate companies, and the fund's directors do not take orders from the manager.

He sat on that board and used it.

Vanguard came in 1974, structured so the company is owned by the funds themselves. Still the only one built that way. The first index fund came in 1976. Wall Street called it Bogle's Folly. He planned to raise a hundred and fifty million and raised eleven.

He was eighty three when he recorded this and he still calls the whole industry a casino to the interviewer's face.

The man who built the index fund destroyed a fund first, and he is the only person who will tell you how.

[Vexo @Vexo200]

John Bogle wrote in his Princeton thesis that fund managers should serve investors in the most economical and honest way possible. Fifteen years later he broke every rule in it.

He was hired in 1951, straight out of college, by the man who had founded the Wellington Fund in 1928.

When the performance slipped in the sixties, Bogle was handed the job of saving it. It was the go-go era. He decided the fund needed a hot manager of its own.

Assets fell from two billion dollars to four hundred million.

The management company fired him.

Sixty years later he sits down in front of a camera and walks through the whole thing himself. He does not soften it. He says he was overly self confident, that he ignored the lessons of history, and that what he did was a really dumb thing.

Then he explains the part almost nobody knows, which is how he came back.

They could fire him from the management company. They could not remove him from the board of the fund. A fund and its manager are two separate companies, and the fund's directors do not take orders from the manager.

He sat on that board and used it.

Vanguard came in 1974, structured so the company is owned by the funds themselves. Still the only one built that way. The first index fund came in 1976. Wall Street called it Bogle's Folly. He planned to raise a hundred and fifty million and raised eleven.

He was eighty three when he recorded this and he still calls the whole industry a casino to the interviewer's face.

The man who built the index fund destroyed a fund first, and he is the only person who will tell you how.

[koba @kobaHUB]

https://t.co/LiV1RA4cm9

[Vexo @Vexo200]

@VoltexGar free for 14 years and nobody clicked, classic

[Vexo @Vexo200]

@velesxbt bro they spent 300m to be first in line lol

[Vexo @Vexo200]

@0xLoran a stock that doubles or halves goes nowhere over time. Rebalanced 50/50 it compounds about 6 percent a period

[Vexo @Vexo200]

@PenguinWeb3 bro is the entire marketing team

[Vexo @Vexo200]

@PenguinWeb3 thank you and enjoy watching

[Vexo @Vexo200]

The equation that tells you how much to bet was published in 1956 by a physicist at a telephone company. He wrote it as a paper about information rates, never used it himself, and died at forty one.

John Kelly worked at Bell Labs, down the hall from Claude Shannon.

Shannon had written A Mathematical Theory of Communication in 1948 and created information theory. Everything digital you touch runs on it.

Kelly took that mathematics and pointed it at a different question. Not how to send a message. How much to bet.

His answer was a formula for what fraction of your money to put at risk, given your edge. Bet more than it says and you go broke while being right most of the time. Bet less and you leave money behind.

He did not publish it as a betting paper. He framed it as a study of information rates, built around a gambler receiving horse race results over a private wire.

The framing was deliberate. A telephone company did not want its name on a paper about gambling.

So the most important result in position sizing sat inside an engineering journal that nobody in finance read.

Kelly never used it. He died in 1965, at forty one, on a street in Manhattan.

Edward Thorp read the paper. He used it to beat blackjack, then built a wearable computer with Shannon and walked into a Las Vegas casino in 1961 with it hidden in his shoe.

Then he took the same mathematics to the market. His fund ran nineteen years without a losing quarter.

Shannon kept a chart of his own stocks on the wall of his house and compounded at roughly twenty eight percent a year for three decades. He never published a word about it either.

When interviewers came in the 1980s and asked about information theory, he wanted to show them his gadgets instead. Juggling robots. A flame throwing trumpet. The roulette computer.

The equation that decides whether being right makes you money was written by a man who never traded, hidden inside a paper about telephone lines, by a company that did not want to be associated with it.

[Rossst.03 @Rossst_03]

https://t.co/eMlCMF7liL

[Vexo @Vexo200]

@Sancho_Wizard the formula itself is one line. Bet your edge divided by the odds. Most people size ten times that and call it conviction

[Vexo @Vexo200]

The equation that tells you how much to bet was published in 1956 by a physicist at a telephone company. He wrote it as a paper about information rates, never used it himself, and died at forty one.

John Kelly worked at Bell Labs, down the hall from Claude Shannon.

Shannon had written A Mathematical Theory of Communication in 1948 and created information theory. Everything digital you touch runs on it.

Kelly took that mathematics and pointed it at a different question. Not how to send a message. How much to bet.

His answer was a formula for what fraction of your money to put at risk, given your edge. Bet more than it says and you go broke while being right most of the time. Bet less and you leave money behind.

He did not publish it as a betting paper. He framed it as a study of information rates, built around a gambler receiving horse race results over a private wire.

The framing was deliberate. A telephone company did not want its name on a paper about gambling.

So the most important result in position sizing sat inside an engineering journal that nobody in finance read.

Kelly never used it. He died in 1965, at forty one, on a street in Manhattan.

Edward Thorp read the paper. He used it to beat blackjack, then built a wearable computer with Shannon and walked into a Las Vegas casino in 1961 with it hidden in his shoe.

Then he took the same mathematics to the market. His fund ran nineteen years without a losing quarter.

Shannon kept a chart of his own stocks on the wall of his house and compounded at roughly twenty eight percent a year for three decades. He never published a word about it either.

When interviewers came in the 1980s and asked about information theory, he wanted to show them his gadgets instead. Juggling robots. A flame throwing trumpet. The roulette computer.

The equation that decides whether being right makes you money was written by a man who never traded, hidden inside a paper about telephone lines, by a company that did not want to be associated with it.

[Vexo @Vexo200]

@tonyprediction original title was Information Theory and Gambling. AT&T made him change it, they had leased those wires to bookmakers for decades

[Vexo @Vexo200]

@Degen_calls_sol without CLAUDE.md the agent just sees thousands of markdown files, with it every session starts with an actual map

[Vexo @Vexo200]

@SpikeCalls repetition as a filter is the whole trick, weak ideas just naturally stop rising back to the top

[Vexo @Vexo200]

A 19-year-old opened his laptop and for a second I forgot how to speak.

No notes. No folders. No Obsidian.

A brain. An actual brain, running on his computer.

Prefrontal cortex. Hippocampus. Cerebellum. Neurons firing in real time as he typed.

"I stopped saving information," he said. "Now it just grows."

This thing doesn't only store what you write. It watches how you work, learns your patterns, and reads the emotion underneath them.

Every note becomes a neuron. Thousands of them, wiring themselves together while you sleep.

He opened the full map and it was alive. Pulsing. Connecting itself.

I've used note apps for ten years. Turns out they were graveyards and nobody told me.

The scary part is I already know I can't go back.

[kai @0xkkai]

https://t.co/JQY6hCnPiZ

[Vexo @Vexo200]

Someone just turned ordinary home Wi-Fi into an X-ray that sees straight through walls.

No camera. No lens. Just the radio waves already flying around your house.

And it doesn't only clock that you're home.

Watch the moment the person lies down on the bed in the next room.

It maps their skeleton, counts 9 breaths a minute, and locks onto a 67 BPM heartbeat. 98% confident, live.

Your body bends Wi-Fi in a way nothing else does. The AI reads the distortion, never you.

The whole rig runs on a $5 ESP32 chip anyone can buy today.

And every line of the code is open source on GitHub.

Your router was never just a router. It's a motion-capture studio that works through solid brick.