#92 Ed Thorp and Claude Shannon
What this covers
Host David Senra traces two linked careers: Claude Shannon, the Bell Labs engineer, and Ed Thorp, the mathematician turned blackjack player and hedge fund manager. The Shannon material starts with his realization that Boolean logic could be built into electrical circuits, then moves to his classified wartime work on the SIGSALY secure voice system. That work convinced him that secrecy and noisy communication are nearly the same problem, which fed into information theory. It also notes that he organized the first major academic AI conference in 1956. Thorp's thread covers his early blackjack trips, financed by Manny Kimmel, whose associates included mafia figures, and his later money management. Princeton Newport lost only about $2 million on a roughly $600 million portfolio during the October 1987 crash.
The second half is about money and risk. Shannon's stock record is the centerpiece: Teledyne bought at $1 in 1960, a reported 28% annual return, and a portfolio that was about 81% one stock by 1986. Senra sets this concentration against Markowitz-style diversification and compares it to Carnegie's steel holdings and Buffett's advice. Shannon also dismissed chart-reading as working from noisy data. Thorp's framing of the efficient market hypothesis is the question of how inefficient markets are, not whether they are efficient. The contrast case is Long-Term Capital Management: $29 borrowed per $1 of investor money, leverage drifting toward 60 times as losses grew, and a partner borrowing personally to add to his stake. Thorp's contested claim is that Merton and Scholes misjudged fat tails and the Kelly criterion. Buffett's remark about IQ 170 and Bernoulli's 1738 warning frame the argument.
Thorp and Shannon treated markets and casinos as environments where a real edge, sized by the Kelly criterion so that ruin is never risked, produces superior long-run returns, while academic efficient-market theory and leveraged 'genius' strategies like LTCM's fail by ignoring ruin.
- Kelly proportional betting makes ruin impossible in principle and grows wealth fastest when an edge exists
- Shannon and Thorp found edges in unconventional domains (roulette, blackjack, warrants, earnings-based stock picking) and sustained above-market returns
- LTCM collapsed by increasing leverage as losses mounted, the opposite of Kelly sizing, and by dismissing Kelly's risk logic
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Manny Kimmel, the financier of Thorp's early blackjack trips, was tied to mafia figures such as Zwillman, and Zwillman's businesses, including parking lots, later became part of Warner Communications.
“that business winds up becoming Warner Communications, which is today known as Warner Media”
Shannon was interested in artificial intelligence early on and organized the first major academic conference on the subject in 1956, meaning many current AI claims repeat ideas from about 60 years ago.
“Claude Shannon was interested in artificial intelligence 63 years ago”
Shannon's classified Project X work at Bell Labs on SIGSALY, a secure wireless system, showed that secrecy systems and noisy communication systems are nearly identical problems, and thinking about concealing messages with random noise motivated core insights of information theory.
“Shannon later said that thinking about how to conceal messages with random noise motivated some of the insights of information theory”
Given a favorable bet, the Kelly system scales each wager to the current bankroll, so the bettor can never run out of money, and because bets grow with a growing bankroll, it produces the fastest growth of any betting system when there is a genuine edge.
“Given a favorable betting opportunity, the Kelly system promises maximum profit and protection against ruin... it is a proportional betting system... you can never run out of money”
Even unlikely events must eventually happen, so anyone who accepts small risks of losing everything will lose everything sooner or later, and the ultimate compound return is acutely sensitive to fat tails.
“It is that even unlikely events must come to pass eventually. Therefore, anyone who accepts small risks of losing everything will lose everything sooner or later”
As LTCM lost money it should have reduced leverage proportionally, but instead it let leverage drift up to roughly 60 times, the opposite of Kelly sizing and a mistake no surviving Las Vegas gambler would make.
“they allowed that leverage to drift up to like 60x. That was a horrible mistake. No Las Vegas gambler would ever make that mistake, no surviving one”
Long-Term Capital Management, staffed by academics who ignored the Kelly criterion, collapsed and required a U.S. Treasury-backed bailout because they did not size bets to protect against ruin.
“academics that criticized Thorp and Buffett and the rest of these guys wind up starting a hedge fund, long-term capital management... not understanding the Kelly formula and blowing up and needing to be bailed out by the U.S. Treasury”
Steve Eisman's observation that people 'mistook leverage for genius' applies across eras, and Warren Buffett noted that roughly 10 or 15 very smart people could still lose all their money through leverage.
“They mistook leverage for genius”
LTCM borrowed $29 for every $1 of investor money, and blowing up an account means losing everything in high-risk trades with borrowed money, which the host presents as the gambler's ruin that Thorp and Shannon avoided.
“For every dollar of investor money, the fund borrowed $29 more. So this is the definition of gambler's ruin”
Warren Buffett marveled that 10 or 15 people with an average IQ of about 170 could still get into a position to lose all their money, and Daniel Bernoulli had written in 1738 that a man risking his entire fortune acts like a simpleton however great the possible gain.
“A man who risks his entire fortune acts like a simpleton, however, great may be the possible gain.”
Thorp argued that LTCM's failure traced to Merton and Scholes' critique of Kelly, because they did not understand how to control extreme risk and fat-tailed distributions, and market probability estimates are always just estimates whose margins of error people optimistically understate.
“I could see that he didn't understand how to control the danger of extreme risk and the danger of fat-tail distributions”
When economists describe market events as 10-sigma or once-in-universe-history catastrophes, and such events keep recurring, they should be expected, so financial models underestimating tail risk warrant extreme downside caution.
“If they keep happening, that means they should be expected, shouldn't they?”
The relevant question for Thorp was not whether the market is efficient but how inefficient it is and how to exploit that inefficiency, which made his approach differ from the mainstream of his time.
“The question wasn't is the market efficient, but rather how inefficient is the market and how can we exploit this”
Estimates of market probabilities are only estimates, and human nature skews margins of error optimistically, so economists repeatedly call market events 10-sigma perfect storms that should instead have been expected given how often they recur.
“human nature often skews these estimates optimistically. A decade rarely passes without a market event that some respected economist claims... to be a perfect storm, a 10-sigma event”
During the October 1987 crash, when the Dow fell 23% in one day, Princeton Newport's roughly $600 million portfolio lost only about $2 million and closed the month about even while most mutual funds fell 20% or more.
“Princeton Newport's $600 million portfolio shed only about $2 million in the crash. Thorp's computer immediately began alerting him to rich opportunities in the panic valuations”
Shannon realized that encoding Boolean logic (true/false, and, or, not, if) in electrical circuits meant any logical relationship could be built from these elements, proving an electrical digital computer could compute anything.
“he proved that an electrical digital computer could compute anything”
Shannon bought a couple of thousand shares of Henry Singleton's Teledyne at $1 a share in 1960, and by 1967 they had reached $24, with Singleton then using inflated stock to acquire around 130 companies.
“Shannon bought a couple of thousand shares at the initial price of $1 a share. By 1967, it hit $24 a share.”
A 1974 Wall Street Journal story reported that Thorp's computer-driven model outperformed all but one of 400 mutual funds tracked, with the sole exception being a gold-stock fund, despite the journal's skepticism that markets are too complicated to model.
“Yet the journal reported that Thorp's machine outperformed all but one of the 400 mutual funds tracked.”
One LTCM partner borrowed $24 million personally to increase his stake in a fund already running at extreme leverage, falling from over $100 million net worth to negative $20 million and asking the bailout to cover his personal debt.
“Because he had taken out a personal loan of $24 million to increase his stake in the fund”
In 1986 Shannon, working part-time with his wife and an Apple II, achieved a higher return than 1,025 of the 1,026 mutual funds Barron's tracked, and his stock portfolio returned about 28% from the late 1950s through 1986.
“Shannon achieved a higher return than 1,025 of them. When Warren Buffett bought Berkshire Hathaway in 1965, over 30 years, he had a return of 27%. By the late '50s through 1986, Shannon's stock return on his stock portfolio was 28%”
Shannon viewed technical analysts working with price charts and chart patterns as using a very noisy reproduction of the important data, preferring to extrapolate earnings growth from management and demand evaluation.
“the technicians who work so much with price charts and head and shoulders formations and plunging necklines are working with what I call a very noisy reproduction of the important data”
By 1986 nearly 81% of Shannon's portfolio was in one stock, Teledyne, and his three largest holdings made up 98% of it, a concentration that would have appalled Harry Markowitz, and Shannon said they had not attempted to balance it in 30 years.
“By this point, nearly 81% of the portfolio was in a single stock, Teledyne. The three largest holdings consisted of 98% of the portfolio”
Concentrating wealth in one basket that you watch closely, as Andrew Carnegie did with his steel company and Shannon did with Teledyne, is how great fortunes are made, in contrast to scattershot diversification.
“It's putting all your eggs in one basket and watching that basket”
Shannon and Carnegie both concentrated capital in a few holdings, and Shannon argued the investor should invest only where an edge exists rather than balance a portfolio, which Senra links to Buffett's advice to treat a business as the family's only asset.
“We have not at any time in the past 30 years attempted to balance our portfolio”
The book frames a split between practitioners such as Shannon, Munger, Buffett, and Thorp, who hold that unusual skill yields an edge, and academics like Paul Samuelson, who hold that markets are perfectly efficient and attempts to beat them are foolish.
“on one side, in this book, you have Shannon, Munger, Buffett, Thorp, and they're saying the idea, like yes, you can gain an unusual advantage in life through unusual skill. On the other side, you have people like Paul Samuelson”
A 23% one-day market drop was difficult to reconcile with rational valuation changes, so Black Monday caused some economists to reject the efficient market hypothesis, although the host notes that people's identities tied to their ideas make such updating rare.
“It was difficult for many to see how a rational assessment of the market's value could have changed 23% in a single day. Black Monday caused a few economists to reject the efficient market hypothesis”
Thorp estimated he had made about $80 billion of purchases and sales across roughly 1.25 million bets averaging $65,000 each over 28 years at a 20% average annual return, and argued this scale makes the excess performance more than chance.
“I estimate that I have made $80 billion worth of purchases and sales for my investors. This breaks down into something like 1.25 million individual bets averaging $65,000 each”
By the end of 1982, $1 invested in Thorp and Regan's partnership at the outset had grown to $6.61 over 13 years, achieving greater-than-market returns for 13 consecutive years.
“$1 invested at the outset had grown to $6.61 in 13 years.”
Shannon was a buy-and-hold fundamental investor who visited startups and served on Teledyne's board, and his approach reflected the same signal-from-noise thinking he developed in communication theory.
“In a way, this is close to some of the work I've done relating to communication and extraction of signals from noise”
Shannon rejected the efficient market theory, arguing that outcomes are not simply luck and that skill can produce an edge, which led him to invest seriously in stocks.
“You know the economists talk about the efficient market where everything is equalized out and nobody can make any money. Really, it's all luck and so on... I don't believe that's true at all”
Princeton Newport grew $1 to $6.61 over 13 years by 1982, and over 19 years it compounded to about $15 at a 15.1% annual rate after fees versus roughly 8% for the S&P 500.
“Over 19 years, the compound return rate averaged 15.1% after fees. During the same time period, the S&P averaged 8%”
LTCM's model assumed small temporary inefficiencies worth about five cents each, so it used massive leverage to collect many nickels, which the host likens to picking up nickels in front of a steamroller without recognizing the danger.
“they're picking up a bunch of nickels in front of a steamroller, and they just don't know that they're doing that yet”
Shannon believed a smart investor should understand where they have an edge and invest only in those opportunities, and he based his approach on forecasting multi-year earnings growth rather than price momentum or technical chart patterns.
“That a smart investor should understand where he has an edge and invest only in those opportunities”
Shannon and Thorp both invested by seeking edge, using information theory and Kelly-based risk control respectively, and the host argues their practices reflected the same underlying thinking about extracting signal from noise.
“In a way, this is close to some of the work I've done relating to communication and extraction of signals from noise.”
To be a successful entrepreneur or investor you must first survive, so accepting even a tiny chance of losing everything is unacceptable regardless of the expected upside.
“To be a successful entrepreneur, you have to first survive. If your business blows up, you're done, you're out of the game”
Paul Samuelson publicly advocated efficient markets while holding Berkshire Hathaway stock, so his actions revealed a belief contrary to his stated position, and the host argues watching what people do reveals what they believe more reliably than what they say.
“I try not to pay attention what people say, I try to pay attention to what they do. I think our actions reveal like -- it's the difference between like revealed preference to stated preference”
Entrepreneurs who own a profitable private business should treat it as if it is the only asset their family will have for the next 50 years, because this forces focus on the area where they have an edge.
“for entrepreneurs that own a profitable private business it makes sense to do what Warren Buffett advises, to treat your business as if it's the only asset your family has had for the next 50 years”
In complex adaptive systems such as business or flying, the correct action is often the opposite of what intuition suggests, which is why business failures are difficult to avoid for those who rely on instinct.
“the thing you should do is the opposite, like you can get yourself in a situation flying in a business where the thing you should do is the opposite of what you feel you should do”
Shannon, Thorp, Buffett, and Munger all agree that if you lack an edge you should not bet or compete, meaning you invest only where you have a demonstrable advantage.
“if you don't have an edge, you don't bet and you don't compete.”
People identify with their ideas, so they resist discarding them even when evidence clearly contradicts them, and good advice is to separate yourself from your ideas and discard those that information disproves.
“when people identify themselves with their ideas... you're not your ideas, you should discard them when you find information that clearly contradicts them.”
Steve Eisman's remark that the LTCM partners mistook leverage for genius applies across eras because human nature does not change, and the host argues the pattern recurs in every generation.
“They mistook leverage for genius.”
Thorp approached investing as a scientist who forms a theory and tests it in the real world, and testing ideas against reality produces unexpected results that refine understanding.
“what I mean about approaching things as a scientist is like he has a theory, then he needs to test that theory in the real world. And as a result of testing your theories in the real world, you're going to get unexpected results”
Ed Thorp optimized his life by measuring output per hour, such as how fast he could learn something or how much money he could make per hour of play, and this per-hour lens applies to nearly any activity.
“He was fascinated with increasing a skill per hour”
Thorp lost a large share of his small capital speculating on silver and concluded that you are unlikely to get an edge from what you see in the news, which pushed him to find areas where he had a real edge such as options and warrants.
“I learned an expensive lesson... The lesson was, you are unlikely to get an edge out of what you see in the news”
Even though Thorp was not charged in a way that stuck, the RICO-based prosecution led investors to withdraw and dissolve the partnership, and Thorp argued the destruction of wealth was huge because it interrupted compounding.
“With RICO's charges looming, the funds investors wanted out... The destruction of wealth was huge... You never interrupt compounding”
Past learnings and experiences play unexpected roles in future success, so collecting as many varied experiences and learnings as possible creates opportunities unavailable to people without that unique set.
“you need to collect as much of these experiences and learnings as you can because it's unpredictable how you can apply them in the future”
Shannon's Teledyne position reflected an informational edge, since as a board member he scouted acquisitions for CEO Henry Singleton, giving him understanding and access to information others lacked.
“Shannon became a Board member of Teledyne... In other words, he had an edge. He knew -- he had an understanding and access to information that other people did not”
Thorp chose not to partner with Sheen Kassouf because Kassouf believed he could predict stock direction, whereas Thorp was unconvinced anyone could predict the market that way, so picking partners with compatible philosophies is essential.
“Kassouf believed that he could sometimes predict in which direction certain stocks are going to move. Thorp wasn't. He was unconvinced that Kassouf or anyone could predict the market that way”
Thorp argued at a party that roulette could be beaten whether the wheel is physically perfect, in which case physics predicts the ball, or flawed, in which case some numbers are favored, making him the originator of the idea that edge exists regardless of conditions.
“He said you can make money either way. If the wheels are physically perfect, simple physics can predict where the ball is going to go. If the wheels have flaws, some of the numbers ought to be favored”
The goal in individual life is to maximize the probability of excess performance that is more than chance, and a long run of many independent bets is the means to demonstrate that a record reflects skill rather than luck.
“We're trying to have a high probability of excess performance that's more than chance”
Thorp argued the destruction of wealth from the legal action was huge, since interrupting compounding is itself costly: the partnership had 80 employees, managed $272 million, and generated about $16 million a year in fees.
“You never interrupt compounding. And a RICO case is going to interrupt your compounding”
Once successful card counters were identified, casinos responded by pushing them out, cheating them, or in some cases violently attacking them, so the host argues people should not assume others will play by the rules.
“there's no rule in life that's saying that other humans are going to play by the rules. And if you don't understand that, you can put yourself in a very dangerous situation”
Claude Shannon invented information theory, a foundation of computers, the internet, and digital media, and it is one of the few cases in history where someone founded a field, asked all the right questions, and answered them at once.
“Shannon's information theory is an abstract science of communication that lies behind computers, the Internet, and all digital media.”
Manny Kimmel's birthday bet was profitable whenever more than 22 people were in the group, illustrating that a bet can look fair while systematically favoring the house with a small structural edge.
“Kimmel knew he had an edge whenever there were more than 22 people”
Thorp was 32 before he ever thought about finance, yet he went on to become one of the most successful investors of all time, showing that a later start need not preclude great results.
“He was 32 years old before he ever thought about finance”
Daniel Bernoulli wrote in 1738 that a man who risks his entire fortune acts like a simpleton however great the possible gain, an early statement of the same ruin-avoidance logic behind Kelly sizing.
“A man who risks his entire fortune acts like a simpleton, however, great may be the possible gain”
Card counting offers deniability because casinos cannot prove it, whereas carrying a device for roulette leaves no deniability because a search would reveal it, which is why the roulette project carried greater personal risk.
“they can't prove that you're counting cards... But with the roulette system like you can't hide a device, if they capture you and search you, they're going to find it. So there was no deniability”
Thorp ran Princeton Newport from California while partner Regan handled business and investor recruitment on the East Coast, connected by phone and data lines, separating thinking from promotion by design.
“What I wanted to do is think, work out theories, and try to put them into action. We were actually happy being separate because we had different styles”
Federal agents raided Princeton Newport in December 1987, reportedly driven by Rudy Giuliani's publicity ambitions and tapes of stock parking by East Coast partners, which Thorp said he did not know about and which led investors to withdraw.
“Some of the East Coast partners took about $15 million out of the fund and replaced it under the names of their wives”
Casinos began identifying and ejecting successful card counters, and in some cases resorted to cheating or violence, including a drugged drink and beatings Thorp reports for other counters.
“I know of three beatings, Thorp said.”
Thorp taught himself Fortran and used MIT's shared IBM 704 mainframe to compute exact blackjack probabilities, because the calculation was too large to do by hand.
“Thorp taught himself the Fortran programming language and programmed the computer himself”
Warren Buffett said Henry Singleton had the best operating and capital deployment record in American business, and Singleton's Teledyne used its inflated share price to acquire about 130 companies.
“Warren Buffett himself said Singleton had the best operating and capital deployment record in American business”
Thorp was an academic with little money when he developed his blackjack theory, and he needed outside bankroll to test it, a fact the host finds inspiring because Thorp became wealthy later.
“He was 32 years old before he ever thought about finance.”
Thorp used a delta hedging system on warrants, grew his original $40,000 to $100,000 by 1967, and framed the investment target as doubling capital every two years.
“His hedging system, and he calls this the delta hedging system. And his hedging system hoped -- or his hedging system worked as he had hoped. By 1967, Thorp had parlayed his original $40,000 into $100,000”
Thorp needed a National Academy member to submit his blackjack paper, and the only one at MIT was Shannon, so Thorp's meeting with Shannon, which captured Shannon's interest in roulette, started their collaboration.
“Articles had to be submitted by a national academy member, and there was only one academy member at MIT, who was a mathematician. That was the famous Claude Shannon”
Thorp declined to invest in LTCM partly because he was concerned that Merton and Scholes had little experience managing other people's money, and Merton was a known critic of the Kelly criterion.
“Thorp decided not to put any money in this fund. He was concerned that Merton and Scholes had little experience investing other people's money. And it didn't help that Merton was a critic of the Kelly criterion”
Thorp kept his UC Irvine teaching post while running his fund for years and did not quit until 1982, consistent with a cautious approach to his new career.
“it took him until 1982 until he quit his day job teaching at UC Irvine”
After Thorp's blackjack paper was publicized, gamblers called to offer financing for a share of profits, but drawing attention also exposed him to casino retaliation.
“gamblers from all over the country began calling Thorp's hotel to ask for copies of his paper”
Shannon and Thorp built a cigarette-pack-sized wearable computer that took the ball's position and velocity from a toe-operated switch and relayed predicted betting segments as musical tones through a wire to an earpiece hidden in Thorp's ear.
“The device they built was the size of a cigarette pack. It contained 12 transistors and slipped into a pocket”
Thorp's blackjack system earned about $366 per hour, and the host notes that losses from Kimmel's mismanagement cut the potential rate by half, which Thorp used to evaluate the system's effectiveness per hour.
“Thorp system, $11,000 of profit in 30 hours is $366 per hour playing blackjack. If Manny didn't mess it up, they would have made $733 per hour”
Thorp's book Beat the Dealer, which popularized card counting, sold well over a million copies, though he had little money when he first tested the system.
“He wrote a best-selling book called Beat the Dealer that's back in the '60s that sold well over a million copies”
Thorp testified that he would take the Fifth Amendment because he had no upside and plenty of downside, reflecting the broader pattern that top investors cap their downside.
“My theory on taking the Fifth was I didn't know anything. So if I had no -- so I had no upside in going in and plenty of downside”
The Shannon-Thorp roulette computer failed in Las Vegas because the thin concealment wires kept breaking, requiring repairs in their hotel rooms and preventing serious gambling, which ended their collaboration.
“The thin wires kept breaking. Every time that happened, they had to go back to their rooms for repair”
Warren Buffett and Thorp met over a bridge night when Buffett was managing money for a UC Irvine dean who invested with Thorp, and Thorp believed Buffett would become the richest man in America.
“He believed that Buffett will one day become the richest man in America.”
Using a predictive device for roulette was legal at the time, but Congress later passed laws in the 1980s outlawing such devices as a result of Shannon and Thorp's work.
“there's actually laws passed in the '80s by Congress to outlaw what they're doing. But at this time, using a device to predict roulette was perfectly legal”
Shannon's early stock purchases, such as a block of Harrison Labs shares acquired by Hewlett-Packard in 1962, produced paper profits that convinced him there was real money to be made in stocks.
“The size of the paper profit convinced him there was real money to be made in stocks.”
Modern life leaves little empty time for unassisted thought, and dedicating time to thinking without outside input is necessary to surface what one's own mind is computing automatically.
“I think that's severely missing from modern life... you need some empty time where you're not having input from anybody else but your own mind”
Some people do not care about rules, and if you cross them you can face serious consequences, as the Cornelius Vanderbilt story and the Zwillman-linked mob world illustrate, so it is important to recognize such people before getting involved.
“there's just certain people that don't care about rules and if you mess with them like you could have consequences”
Senra treats podcasts and biographies as a structured course, learning fast by listening to practitioner interviews and reading biographies that connect ideas to the lives of their developers.
“I think when you were trying to teach the great concepts that work it helps to tie them into the lives and personalities of the people who develop them”
Thorp's fund shut down not because of market losses but because of partner-driven legal problems, so the host argues the cause of failure is often outside the investment model and depends on partner selection.
“Now here's the problem, though. I said earlier it's important to pick the right business partners”
Thorp's market-neutral hedge fund approach echoed his earlier roulette argument that money can be made in any environment, whether the market is efficient or flawed.
“he calls his hedge fund a market-neutral hedge fund. Just kind of echoes like what he's saying here about roulette like you could figure out how to make money in any environment”
Poundstone's account shows Shannon's investment methods look much like Charlie Munger's, and Shannon's record is evidence that unusual skill can produce results that conventional markets and betting systems would not predict.
“the book contains an account of the Lollapalooza investment record of Claude Shannon, a pioneer scientist in information theory that make Shannon's methods look much like those of Charlie Munger”
Thorp built a fallout shelter with 16-inch concrete walls and computed it could withstand a one-megaton blast a mile away, which the host reads as evidence of a mind that is always weighing odds and tail risk.
“Ever mindful of the odds, Thorp computed that it could withstand a one-megaton hydrogen bomb blast as close as a mile away”
Thorp was told Kimmel would start with $100,000 but Thorp reduced the stake to $10,000, and the host notes that larger amounts are harder to run successfully, as with Shannon's record at scale.
“It's also worth noting it's a little harder to do with larger amounts”
Shannon moved on from information theory because his interests changed, and he allowed himself to focus on whatever was most interesting at each point in life, which the host views as a sign of his personality rather than unfocused behavior.
“I just developed different interests. As life goes on, you change your direction”
Claude Shannon, a shy and self-directed thinker who did his best work alone and thought for hours at a kitchen table, pursued whatever interested him at the time, and this freedom to follow varied interests later shaped his combined discoveries in information theory and investing.
“He did his best work alone. He slept when he felt like sleeping and would spend hours at the kitchen table thinking”
The host dismisses Markowitz's modern portfolio theory as twaddle, agreeing with Buffett and Munger's criticism that diversification theory is wrong.
“I think it's twaddle.”