Bits of Books - Books by Title


Range: How Generalists Triumph in a Specialised World



David Epstein



More books on Success

More books on Education and Learning

It takes time - and often forgoing a headstart - go develop personal and professional range, but it's worth it.

Highly credentialed experts can become so narrow minded that they actually get worse with experience, even while becoming more confident.

Learning is best done slowly to accumulate lasting knowledge, even when that means performing poorly on tests of immediate progress.

Three areas where early specialisation and lot of practice give best results. 1) Tiger Woods trained from a very early age (before he could talk) 2) Laslo Polgar trained his daughters to become chess stars 3) the fire chief who pulled his crew out of a burning room just before floor collapsed.

But chess, golf and firefightig are 'kind' learning environments. You learn (a large number of) patterns which repeat over and over, and you get immediate and accurate feedback. The learning environment is 'kind' bc a learner improves simply by engaging in the environment and trying to do better.

The flip side is the 'wicked' domains, where the rules are often incomplete, patterns may not be obvious, and feedback is delayed and/or inaccurate.

AI powered comp programs eventually beat best humans at chess, and then Go. But here it'soperating in a constrained, rule bound world. But video games a different story. Turns out that games such as StarCraft require many layers of thinking. And, this applies to most real-world problems.

AI systems are like savants: they need narrow worlds with clear rules and answers that don't change over time.

Danger of treating the wicked world as if it is kind. Consultants from high powered business schools did really well on business school problems that were well defined and quickly assessed. But in the real world they applied 'single loop' learning - they used the first familiar solution that came to mind. Whenever that solution proved wrong, the consultant became defensive (blamed everyone else).

IRL no-one has told you the rules, and those rules are liable to change without notice.

If the amount of narrow, focused learning were the key to innovative performance, savants wd dominate every domain they touched, and child prodigies would always go on to adult eminence.

There are other narrow domains - accountants, bridge and poker players, surgeons. But they are all based on long expwerience of learning possible patterns. When the rules are changed, even slightly, the 'expert' finds it harder to adapt and learn than newbies approaching task for first time.

Electrical engineer Claude Shannon launched the Information Age thanks to a philosophy course he was forced to go on for engineering degree. There he learned of a British logician George Boole, who assigned a value of 1 to true statements, and 0 to false ones, and showed that logic problems cd be solved like maths equations. This resulted in absolutely nothing of practical value until 70 years after he died, when Shannon did a summer internship atBll Labs research fcility. There he realized that he cd combine call-routing tech with Boolean logic to encode, store and transmit any type of information electronically. This fundamental insight is what computers rely on. "It just so happened that no-one else was familiar with both those fields at the same time" Shannon said.

More books on Computers

Modern work requires knowledge transfer: the ability to apply knowledge to new situations.

Venice C17 and C18 had all female musicians called figlie del coro, literally 'daughters of the choir'. They sang and played instruments usually reserved for men. The best became Euro-wide celebrities.They performed behind screens, bc although sang like angels, it turnned out they were mostly very ugly/deformed/old. They were all foundlings dropped off at a ospedalia, charitable institiution founded to save babies, usually prostitute's, who wd otherwise be dropped in a canal.

The ospedali governors realized lots more people came to church when women sang, so started promoting concerts. Couldn't charge people to enter a church, but you cd rent a chair if you wanted to sit at the front. You weren't allowed to applaud in church, so had to cough and blow nose and stamp feet to show appreciation.

More books on Music

The women typically learned every instrument the institution owned, and swapped roles frequently within concerts. They were so talented that craftsmen made special instruments just to extend their prowess, and composers such as Bach, Vivaldi and Mozart wrote works that cd only be performed by these women.

Top jazz guitarist Jack Cecchini self taught. "I cd show someone in two minutes what wd take them years of screwing around on a fingerboard like I did to find it ... It's slow, but at the same time , there's something to learning that way."

In contrast to the Tiger Mom parenting style, if you want creative kids you have to be a lot less prescriptive. Creativity is difficult to nurture, but easy to thwart. Study found average home had six household rules for kids, compared to one in households with extremely creative kids. Parents with creative kids made their opinions known after their kids did something they didn't like, they just didn't try to proscribe it beforehand.

Scenario: you are a surgeon with a patient with malignant stomach cancer that cannot be surgically removed. There is a kind of ray that can be used to destroy the tumour if high enough intensity can reach it. But at this intensity all the healthy tissue that the ray passes through on way to tumour will aso be destroyed. At lower intensities the ray won't harm healthy tissue, but won't be enough to kill tumour. What to do?

Here's a little story while you think: once was a general trying to capture a castle. He has=d a big enough army to take the castle if they were all there at once, but all the roads leading to the castle were narrow, so only a few men at a time cd pass. So he split his forces and sent small groups down each of the roads,co-ordinated to all get there at same time.

Another story: A fire chief arrived at a fire in a small shed but that cd easily spread to adjoining houses. A group of neighbours were filling buckets from the nearby pond and tossing water onto flames, but weren't winning. The fire chief told them all to sstop, each get a bucket of water,then on count of three, everybody dumped water at once. Te fire quickly extinguished.

So, have you saved the patient yet? The solution is to send low intensity rays frommultiple sources, focused so all arrive at tumour, so tumour gets full power rays, but surrounding flesh only gets low dose.

'Outsider thinking' came from Eli Lilly posting online 21 problems that had stymied their best scientists. They got suggestions and solutions from all walks of life. So spun off a separate company, InnoCentive where people cd pay to post challenges, and solvers were rewarded.

The most interesting finding was that the further the probblem was from the solver's expertise, th emore likely they were to solve it. Expertise is increasingly specialised, but when that specialised knowledge can't generate an answer, they're stuck. Then someone who's far away from the surface of the problem reframes it in a way that unlocks the solution. Knowledge is a double-edged sword. It allows you to do some things, but it also blinds you to other possibilities.

The more (specialised) info that specialists accumulate, the more opportunity for curious dilettants to contribute by finding unsuspected links.

Nintendo was a struggling playing card company until their machine serviceman started tinkering with old technology. The attraction of 'withered tech'was that it was cheap, well understood, and readily available. He developed the Game Boy - a handheld console that played any game a developer cd put on a cartridge. The graphics were crap but it played for days on a couple of AA batteries. Old tech was very familiar, and they pumped out Tetris, Super Mario Land and a slew of sports games. They sold 118 million units ofGame Boy, far and away the best selling console of C20.

Freeman Dyson talked of frogs and birds. Birds fly high in the air and survey vast vistas out to the horizon. Frogs live in the mud and see only the flowers that grow nearby. They delight in the details of particularobjects, and they solve problems one step at a time. Dyson's concern was taht science is incrasingly overflowing with frogs, trained only in a narrow speciality and unable to change as csience does.

Modern comic books require one or more creators to integrate narrative, dialogue, art and layout design. Expectation that the most successful comics would come from most experienced artists. Instead found that the spectacular successes came from creators who had dabbled widely in many of the 22 different genres of comic books - comedy, crime. NF, SF, fantasy, adult etc.

Hayao Miyazaki, the Japanese artist responsible for Spirited Away, the highest grossing film in Jpan ever, but he also drew comics in just about every genre.

In kind environments, where the goal is to reproduce performance as perfectly as possible , teams of specialists work superbly. Surgical teams work faster and make fewer mistakes as they repeat specific procedures. If you need surgery, you need a doctor who specialises in the procedure and who has done it many times.

Same applies to air crews. Analyse major flight accidents found 73% occurred on flight crew's first day working together.

Utah U study of serial innovators found these characteristics - high tolerance for ambiguity, additional knowledge from peripheral domains, abilityto connect disparate info in new ways, synthesising info from different sources, broad range of interests, widely read, having friends in different knowledge domains.

Hedgehogs and Foxes: Philip Tetlock. Hedgehog experts were deep but narrow. They spent years studying a single proble. They fashioned tidy theories about how the world worked, and then bent every event to fit them. Foxes drew from a wide range of data, explored shallowly, and tolerated ambiguity. Hedgehogs performed especially badly in areas of their own expertise, and actually got worse as they accumulated experience.

Narrow experts are an invaluable resource, but you have to realize that they may have blinkers on. So what you have to do is take facts fromthem, not opinions.

Hedgehogs tend to see simple cause and effect with everything framed by their area of expertise, like repeating patterns of chess games. Foxes see complecity, and understand that most situations are probalistic, not deterministic. Even when history seems to repeat, it does not do so precisely.

In wicked domains lacking automatic feedback, experience alone does not improve performance. Effective hbits of mind are more impt, and they can be learned. Few events are truly novel, so try to generate a list of separate events with structural similarities. For example, forecasters were asked if Greece wd drop out of the Euro currency zone in 2015. No country had done that, so superficially it was a unique event, but there of examples of exits from international agreements.

Smokejumpers are trained firefighters who are helicoptered or parachuted into wildfires. Every so often a fire behaves unexpectedly and the men have to run for their lives. The obvious rule is to drop your chainsaw and your heavy pack, but in many cases the men are so wedded to their tools that they are found clinging to them, dead.

Chain of command,but also circle of communication: don't listen just to voices at the top; if you check with the guys at the coalface, they'll tell you where sand is getting into the gears.

Trenching: many cardiologists specialise in treating chest pain by placing stents into blood vessels to force them apart. Seems logical response to a blockage, but randomized clinical trials have repeatedly shown that compared with more conservative treatment of patients with stable chest pain, stents prevent zero heart attacks, and extend lives of patients not ata all.

A 2015study showed that patients with heart failure or cardiac arrest were less likely to die if they were admitted during a national cardiology conference, when all the 'experts' were away.

Advice to grad students: don't end up a clone of your thesis advisor. Take your skills to a place that's not doing the same sort of thing, Take your skills and apply them to a new problem, or take your problem and try completely new skills.

"Two black holes collided in space a billion years ago, and for a billion years those gravitational waves have been travelling through space. When the original signal began, life on Earth was unicellular, and in that time humanity has managed to build interferometers to measure those waves. How cool is that?"

Doctors and scientists need training in basic logic skills. If a test to detect a disease whose prevalence is 1/1000 has a false positive rate of5%, what is the chance tht a person with a positive result actuallyhas the disease (assuming you know nothing aboutthe person's symptoms.) When this was posed to a group of Harvard doctors and med students, the most common answer was 95%.

In fact he correct answer is "less than 2%". If you have a sample of 10,000 people, 10 of them will have the disease and get a true positive result; 5%, or 500 people will get a false positive. Out of 510 people who get a positive result, only 10, or 1.96%, are actually sick.





Books by Title

Books by Author

Books by Topic

Bits of Books To Impress