0:15

Pushkin. It

0:26

would be hard to think of a better example

0:28

of a game of chance than roulette.

0:32

Beneath the romantic French terminology

0:34

and the myriad rules of etiquette, each

0:37

spin of the roulette wheel is utterly

0:39

random. The casino's advantage

0:42

is small that it cannot

0:44

be overcome. The game

0:47

is remorseless over the

0:49

long haul. The only way to win

0:52

is not to play, or

0:54

is it? One day in

0:56

August nineteen sixty one, Claude

0:59

and Betty Shannon stroll up to

1:01

a roulette table in Las Vegas, pretending

1:03

not to know their companions, Ed and

1:06

Vivian Thorpe. Claude and the ladies

1:08

a nerve us, but they don't show it.

1:11

Ed Thorpe isn't nervous, he's excited.

1:14

He's still in his twenties, but he's

1:16

an old hand in the casinos. Claude

1:19

Shannon stands right by the wheel. He's

1:22

forty five years old, slim and good

1:24

looking, with fine cheekbones and dark

1:26

eyebrows. He's misdirecting

1:29

the attention of the floor manager by scribbling

1:31

down numbers. He looks like he's got

1:34

some crazy system that will inevitably

1:36

bankrupt him. Thorpe is at

1:38

the other end of the table, far from

1:40

the wheel and far from Shannon. He

1:42

has dark hair, a round face, and

1:45

a smile. He's having fun placing

1:48

his bets with the confidence of a man who

1:50

knows the unbeatable game is

1:52

about to be beaten. This

1:55

is a defining moment in a project

1:57

that has been quietly ticking over for

1:59

a year. When it began, Thorpe

2:02

and Shannon didn't know each other. Edward

2:05

O'thorpe was a junior mathematics

2:07

instructor at MIT. Claude

2:10

Shannon was the greatest computer

2:12

scientist in the world. Ed

2:15

Thorpe had a plan to beat Roulette,

2:17

and he needed Shannon to help

2:20

him. Systems to beat

2:22

Roulette are like blueprints for perpetual

2:24

motion machines or formulas to turn

2:26

lead into gold. They're absurd,

2:29

the pseudo scientific obsessions of Cranks

2:32

and Claude. Shannon's secretary had already

2:34

warned Thorpe that Professor

2:36

Shannon doesn't spend time on topics

2:39

or people that don't interest him.

2:41

Shannon was a legendary figure. People

2:44

in his field talked about Shannon the

2:46

way physicists talk about Albert

2:48

Einstein. What ed Thorpe

2:50

was doing was much like buttonholing

2:52

Einstein and saying, Hey, Albert,

2:55

I've got a sure fire scheme for beating the bookies

2:57

at the racetrack. An unknown

2:59

young mathematician, a patently

3:02

futile goal, Claude Shannon,

3:04

the computing legend, didn't

3:07

hesitate. Take

3:09

a seat. He said to Ed Thorpe.

3:12

We have a lot to talk about. I'm

3:16

Tim Harford. You're listening

3:19

to cautionary tales. Repeat.

3:44

Please please some floora for the present.

3:46

How do you receive? Send floa? Please

3:49

see if you can read this? Can you read

3:51

this? Yes? How are signals?

3:54

Do you receive? Please send something?

3:56

Please send bes and bees? How our

3:58

signals? Those messages

4:01

from eighteen fifty eight represent

4:04

a full day of attempted

4:06

conversation via Morse code

4:09

who were cable lying three miles

4:11

under the surface of the Atlantic Ocean.

4:13

The cable had been enormously

4:16

expensive, and, as he might

4:18

guess, it wasn't really working. In

4:21

an attempt to boost the signal, the

4:24

project's engineer, a man called

4:26

Wildman white House, cranked up

4:28

the voltage. The cable

4:30

melted. It had

4:33

survived only twenty eight days.

4:36

Over the years, telegraph engineers

4:38

figured out how to work around the problem

4:40

of noise on the line. They built stronger

4:43

cables with better insulation and more

4:45

sensitive detectures at the far end. But

4:48

nobody fully solved the problem of noise.

4:51

Nobody even fully understood it. Not

4:54

until nearly a century later,

4:56

along came Clawed Shannon. Shannon's

5:00

career was defined by two thunderbolts

5:03

of insight. When he was twenty

5:05

one in nineteen thirty eight, his

5:07

master's thesis showed that any

5:09

logical statement could be evaluated

5:11

by a machine, with true or

5:13

false being represented by switches

5:16

being open or closed. Those

5:18

dots and dashes of Morse code were

5:21

just a hint at the possibilities. Armed

5:23

only with open or closed on or

5:25

off dot or dash zero

5:28

or one, machines could

5:30

perform any operation in mathematics

5:33

or logic, And rather than

5:36

merely proving the point in abstract,

5:38

Shannon, who was barely old enough to

5:40

buy a beer, showed electrical

5:42

engineers how to efficiently

5:44

build a logic machine. Claude

5:47

Shannon had bridged the vast gap

5:50

between electrical wiring diagrams

5:52

and thought itself unlocking

5:55

the age of the digital computer.

6:00

Shannon's second thunderbolt was published

6:02

in nineteen forty eight, when he was working at

6:05

Bell Lamps alongside several

6:07

future Nobel prizewinners,

6:09

including the team that invented the transistor,

6:12

Shannon returned to the deep problem underlying

6:15

the Transatlantic cable fiasco.

6:18

He created a unified mathematical

6:21

theory of transmitting information.

6:24

Some of that theory seems obvious from

6:26

the viewpoint of the twenty first century.

6:29

We now take it for granted that information

6:31

bits and bytes and gigabytes might

6:34

represent anything a computer

6:36

game or a spreadsheet, or music

6:39

or pornography. But that idea

6:41

started with Shannon. Before

6:44

him, researchers only dimly grasped

6:47

the distinction between the meaning of a message

6:50

and the quantity of information it contained.

6:53

The idea of compressing a file

6:56

so that it took up less space were

6:58

Shannon's, and so too was

7:00

the utterly radical idea that

7:02

any amount of noise on a line could

7:04

be overcome. We didn't do

7:07

that by cranking up the voltage and melting

7:09

the undersea cable, nor did you need

7:11

to build a better listening device or a thicker

7:13

cable. No matter how much

7:16

distortion there was, you

7:18

could convey any message if

7:20

you had enough patience. All

7:22

you had to do was add redundancy

7:25

to the data. It's the inverse

7:27

of compressing a file. You add

7:29

extra data to make the message more

7:31

likely to be recoverable. Even in

7:33

the presence of interference, that

7:36

idea was unthinkable,

7:38

right up to the point that Claude Shannon

7:41

improved how to do it. This

7:46

new theory of information was revolutionary

7:49

and so elegant and general that it

7:51

could be applied to anything from the Internet

7:54

to genetic information in DNA,

7:56

even though the Internet did not then exist

7:58

and the double helix structure of DNA

8:01

had not yet been discovered. Shannon

8:04

wasn't merely ahead of his time. He

8:06

was the one who had wound the clock and

8:09

set it running. All this

8:12

and he'd barely turned thirty. So

8:15

what did Shannon do for an encore? Is

8:18

a description from his biographers, Jimmy

8:21

Sony and Rob Goodman of Shannon's

8:23

work ethic. Shannon

8:25

arrived late, if at all, and

8:28

often spend the day absorbed in games of chess

8:30

and hecks in the common areas. When

8:33

not besting his colleagues at board games,

8:35

he would be found piloting a unicycle

8:37

through Bell Labs's narrow passageways

8:40

Occasionally while juggling. Sometimes

8:42

he would po go stick his way around the Bell Labs

8:44

campus, much to the consternation reimagine

8:48

of the people who signed his paychecks.

8:51

Shannon wasn't goofing off completely.

8:53

He often worked hard, but the

8:55

projects he worked tom seemed whimsical.

8:58

For example, he spent many hours

9:00

at home playing with a colossal

9:03

erector set. He built a robot

9:05

mouse that could explore a maze and

9:08

by trial and error on the first attempt, learn

9:10

how to reach its target flawlessly on

9:13

the second run. The robot

9:15

mouse was clever and thought provoking,

9:17

and it might have represented real progress

9:19

towards artificial intelligence if

9:22

Shannon had persisted with it. But

9:24

he didn't. Shannon

9:27

built perhaps the first chess

9:29

playing computer, albeit

9:31

one that could play only a radically

9:33

simplified setup the end game

9:35

with six pieces. He published

9:37

a theoretical paper on computer chess.

9:40

It could have been the start of something, but

9:43

again he lost interest. It

9:46

seemed a shame. If anyone could

9:48

make progress with computer chess, surely

9:50

it was Shannon. He was good. Shannon

9:53

once traveled to Moscow and played chess

9:56

with three time world champion Mikhael

9:58

Botvinick, and he made Botfinick

10:00

sweat. When it wasn't

10:02

chess, it was juggling. Shannon tried

10:04

to figure out how to juggle upside down

10:07

by hanging from the ceiling and bouncing the balls

10:09

off the floor. He built juggling

10:11

robots too, and a variety

10:13

of machines designed to play abstract

10:16

games, such as hex and a

10:18

Rubic's Cube solving robot, and

10:20

the Jugglometer and a flame

10:23

throwing trumpet and the

10:25

Ultimate Machine. The

10:27

Ultimate Machine is a box with a switch and

10:29

a trapdoor, and you flick the switch

10:32

to turn it on. A robot finger

10:34

pops out of the trapdoor and flips

10:37

the switch back again to turn itself

10:39

off. Shannon made

10:41

giant styrofoam shoes so he

10:43

could walk on water at a nearby lake.

10:46

After Shannon learned to juggle, ride

10:48

a unicycle and walk a tight rope,

10:51

he formulated the aim of juggling on

10:53

a unicycle on a tight rope

10:56

alas he never got further than two

10:58

out of three.

11:02

Claude Shannon's boss, Henry Pollock,

11:05

said, Shannon has earned

11:07

the right to be nonproductive, and

11:11

of course he had, but

11:13

come on, you're a genius.

11:16

Claude, You're thirty three

11:18

years old. You're the Einstein of computer

11:20

science, and you're unicycling,

11:23

poe going and playing board games.

11:26

Shannon never again published

11:28

anything like his theory of information.

11:31

He never even came close. Once

11:34

he promised the editor of Scientific American

11:37

an article on the physics of juggling.

11:40

If that didn't seem trivial enough, he

11:42

followed it up with an unapologetic

11:45

letter, You probably think I've

11:47

been frittering. I say, frittering away my time

11:49

while my juggling paper is languishing on the

11:51

shelf. This is only half true.

11:53

I have come to two conclusions recently.

11:56

One, I'm a better poet than scientist.

11:59

Two. Scientific American should

12:01

have a poetry column instead

12:04

of his juggling research. Shannon

12:06

enclosed a seventy line poem about

12:09

Rubik's cubes, to be

12:11

sung to the tune of Tarara

12:14

Bundier. He

12:16

added, I'm still working on the

12:18

juggling paper. Shannon

12:21

never finished it. Not

12:24

only was he not producing thunderbolts,

12:27

he wasn't even producing a study of juggling.

12:30

Perhaps we should not be surprised

12:32

that Claude Shannon was happy to put

12:34

aside serious re search when

12:36

the young mathematician Ed Thorpe approached

12:39

him for help in hacking the roulette

12:41

table in Vegas. Cautionary

12:47

tales will be back in a moment.

12:57

If we know anything, we know we're supposed

12:59

to stick to a task. Psychologists

13:02

have developed some attractive ideas about

13:04

how success depends on practice

13:07

and determination. Is

13:09

Angela Duckworth, who's popularized

13:11

the idea of grit, Carol

13:14

Dwex research on the growth mindset,

13:16

and the late and as Ericsson, the

13:19

source of the ten thousand hour rule

13:21

made famous by Malcolm Gladwell. There

13:24

are subtleties to each of these research programs,

13:27

but the versions that have broken into popular

13:29

culture are simple enough. Like some

13:32

motivational poster, nothing

13:34

in this world can take the place of

13:37

persistence. Talent will

13:39

not. Nothing is more common than

13:41

unsuccessful people with talent. Genius

13:44

will not unrewarded genius

13:47

is almost proverb. The slogan

13:50

press I has solved

13:52

and always will solve the problems

13:54

of the human race. Isn't

13:56

that great? It's often attributed

13:59

to President Calvin Coolidge, but

14:01

it's older than that. Claude

14:03

Shannon, however, seems not to

14:05

have gotten the message he

14:08

achieved so much, But if it's

14:10

stuck to a task, couldn't he have achieved

14:12

so much more? Instead,

14:15

he was playing with flamethrowing trumpets,

14:17

juggling robots, and silly poems, oh

14:20

and the impossible task of beating the casino

14:23

at roulette. For

14:27

a junior academic, ed Thorpe

14:29

spent a surprising amount of time in

14:31

casinos, using some ferocious

14:34

mathematics and the best computers

14:36

he could access at MIT. Thorpe

14:39

had figured out that it was possible to beat

14:41

the dealer at the casino staple

14:43

blackjack by keeping track

14:45

of the cards that had been played in

14:48

placing bets when the deck was offering

14:50

favorable odds. Card

14:52

counting is a familiar idea these days.

14:55

It all started with Ed

14:57

Thorpe. Thorpe's

14:59

ideas were sophisticated enough to be worth

15:01

publishing as an academic paper, which

15:04

he did, but he wasn't

15:06

content with that. He wanted

15:08

to beat casino too. To

15:11

do that, Thorpe had to learn to

15:13

spot crooked dealing, where

15:15

a disguise count cards unobtrusively

15:18

late into the night, and above all,

15:21

make sure he didn't get killed. That

15:24

was no idle worry. One

15:27

day Thorpe made a little too much

15:29

money, and the casino spiked

15:31

his coffee with something mysterious

15:33

that blurred his vision for hours.

15:36

He came back the next day

15:38

and the casino tried it again, but

15:41

Thorpe wasn't scared. His

15:43

idea to beat Roulette was the boldest

15:46

of all. He didn't have in mind

15:48

a clever mathematical system. There

15:50

are loads of them, and he knew that none of

15:52

them work. Instead,

15:55

he planned to build a computer that could

15:57

predict where the ball would land. That

16:00

would be hard even today, but

16:02

at a time when computers were the size of

16:04

pianos, this computer

16:07

needed to be one that you could seal

16:09

inside your clothes, the

16:11

world's first wearable

16:13

computer, decades before the

16:16

fitbit, Google Glass or

16:18

the Apple Watch. Thorpe

16:21

had done some experiments on the timing of

16:23

a Roulette wheel with his wife, Vivian,

16:25

a woman who was both intelligent and

16:28

indulgent, as you'd need to be

16:30

if you were married to Ed Thorpe. But

16:33

to crack the problem he needed

16:35

to team up with perhaps the best

16:37

gadgeteer in the world, Claude

16:40

Shannon. Thorpe spent

16:42

twenty hours a week at Shannon's house

16:45

he was in heaven. The basement

16:48

was a gadgeteers paradise.

16:50

Moders, transistors, switches,

16:53

pulleys, gears, condensers, transformers.

16:56

I was now happily working with the

16:59

ultimate gadgeteer. Shannon

17:01

and Thorpe were able to time the spinning of

17:03

the ball around an upper loop and the contrary

17:06

motion of the wheel itself. With practice,

17:08

they were to start a clock within one

17:10

hundredth of a second and then stop the

17:13

clock after ten revolutions. That

17:15

gave them both the speed and the position of

17:17

the ball relative to the wheel, and Newtonian

17:19

physics could do the rest. The

17:22

result of months of experimentation

17:25

taught them that, using their computer to

17:27

compute the path of the ball, they could

17:29

predict that it would fall into one of five

17:31

numbers just over one eighth

17:33

of the wheel and expect to be right

17:36

twenty percent of the time. It seems

17:38

a modest advantage, but the potential

17:40

profits were enormous. All

17:43

they had to do was to figure out

17:45

how to miniaturize that computer, making

17:48

it small enough to slip into a pocket

17:50

and carry into the casino undetected.

17:53

It was an astonishingly audacious

17:56

project and a huge effort

17:58

For the final three weeks, Thorpe was

18:01

effectively living at Shannon's house, but

18:03

by August nineteen sixty one, the

18:06

device was ready with their

18:08

accomplices, Vivian Thorpe and Claude's

18:10

wife, the mathematician Betty Shannon.

18:13

The two gadgeteers then took it to the casinos.

18:16

The Einstein of computer science was

18:19

going to Las Vegas. Looking

18:25

at Claude Shannon's career from age

18:27

thirty three onwards, it's hard to

18:30

escape the conclusion that he might have

18:32

achieved more, much more, if

18:34

not for his habit of flitting between

18:36

whimsical projects and typically setting

18:39

them aside before they were finished. But

18:41

some very smart people would disagree.

18:44

Vanavar Bush arguably knew

18:46

more than anyone about the way scientific

18:49

progress occurred. He guided

18:51

science policy for the United States

18:53

during the Second World War, coordinating

18:56

the efforts of six thousand researchers.

18:59

Bush said that great scientists should

19:01

range widely and keep changing things

19:03

up. In a speech to professors

19:06

at MIT, Bush advocated

19:08

breath rather than depth. It

19:11

is unfortunate when a brilliant and creative

19:13

mind insists upon living

19:16

in a modern monastic cell. Bush's

19:19

idea was later backed up by scientific

19:21

investigation of scientists themselves.

19:26

In nineteen fifty eight, a remarkable

19:28

study was launched by a young psychologist

19:31

named Bernice Agison. The

19:33

study followed a group of promising

19:35

researchers as their careers unfolded,

19:38

periodically interviewing them and continuing

19:40

even after Agison herself died in

19:42

nineteen eighty five. Four of

19:45

the scientists eventually won Nobel

19:47

prizes. The findings

19:49

of the Agison study support Shannon's

19:51

habit of flipping from one project to

19:53

another. The scientists who'd

19:56

most flourished over the decades had

19:58

switched back and forth dozens

20:00

of times. Once you start

20:02

looking for this pattern, you see it everywhere.

20:05

Isaac Newton is most famous for formulating

20:07

the law of gravity, but made huge

20:10

advances in mathematics and optics.

20:13

He was the master of the Royal Mint and

20:15

was fascinated by economics, and

20:17

devoted as much attention to alchemy as

20:19

to anything else. Einstein

20:22

published four astonishing scientific

20:24

papers on four different topics,

20:26

all in the same year nineteen o five.

20:29

Charles Darwin worked simultaneously

20:32

on the theory of evolution, the definitive

20:34

two volume work on Barnacles

20:37

and a book about the human infant, begun

20:40

while his son William was a baby,

20:42

and published just in time for William

20:44

Darwin's thirty eighth birthday. Multiple

20:48

projects aren't unusual at the highest

20:50

level of science, they're the norm.

20:53

Not only that high achieving scientists

20:55

often have time consuming side interests,

20:58

pursuing photography, fine art, or

21:00

music at or near a

21:03

professional level. Nobel

21:05

Prize winning scientists are substantially

21:07

more likely to have serious hobbies

21:10

than other leading scientists, who

21:12

in turn are more likely to have them than the

21:14

rest of us. The

21:16

later part of Shannon's career fits

21:19

right into this highly diverse pattern,

21:22

but then so does the early part.

21:25

Back in nineteen thirty nine, shortly

21:27

after his first thunderbolt, he

21:29

wrote a note to an academic

21:31

mentor, Dear doctor

21:33

Bush, Yes, Van of our Bush,

21:36

the man who knew everyone who mattered

21:38

in mid century American science. Of

21:41

course, he was there to support the young

21:43

Claude Shannon. Dear doctor

21:45

Bush, I've been working on three

21:48

different ideas simultaneously, and strangely

21:50

enough, it seems a more productive method

21:52

than sticking to one problem.

21:55

When Shannon wrote to Van of oar Bush,

21:57

he wasn't working on engineering or logic.

22:00

He was working on genetics. He knew

22:02

nothing about the subject, but swiftly

22:04

produced a completely new kind

22:06

of algebra to describe and analyzed

22:09

genetic inheritance. The work

22:11

was intriguing and wholly original, but

22:14

needed developing. Did Shannon

22:16

develop it? He did not. In

22:19

fact, he never even bothered publishing

22:21

it. Neither did he ever return

22:23

to genetics. Later scholars

22:25

lament the loss. His new algebra

22:28

might really have advanced the field, but

22:30

sticking with genetics might also have meant

22:33

he never had his second thunderbolt on

22:35

information theory. Between

22:38

those two thunderbolts, Shannon didn't

22:40

just switch fields. He lived

22:42

a rich and complicated life. He

22:45

married and then divorced within a year.

22:47

He moved to Manhattan to spice things up.

22:50

It played chess in Washington Square Park.

22:52

It played clarinet. He loved the jazz

22:54

scene in New York. He swam,

22:57

played tennis, stayed up too late,

22:59

and played his music too loud. All

23:01

this was happening when Shannon was at the peak

23:04

of his intellectual powers. Shannon

23:06

didn't just hit thirty five then

23:09

and in serious thinking in favor of playing

23:11

around. Shannon was

23:13

playing around all along. Maybe

23:16

Shannon's love of frittering, I say,

23:18

frittering away his time on

23:20

juggling or unicycling, or

23:22

music or chess. Maybe

23:24

that's not the reason he produced only

23:27

two truly brilliant ideas. Maybe

23:30

it's the reason he produced two truly

23:32

brilliant ideas in the first

23:35

place. Cautionary

23:41

tales will be back in a moment. I

23:49

try hard to answer all the people

23:51

who write to me. I get anxious

23:54

knowing that the task is unfinished. Claude

23:57

Shannon didn't feel that same compulsion

23:59

to clear his inbox. He often

24:01

left correspondence unanswered, then

24:04

eventually cleared the decks through the use of

24:06

a trash can marked letters I've

24:08

procrassed, donated on for too long. That

24:11

might seem a trivial thing, but I think it

24:13

points to something deeper. Psychologists

24:16

have identified a tendency called completion

24:19

by us. If you've ever assembled

24:21

a list of things to do, then ticked

24:23

off all the easy ones while ignoring

24:26

the important stuff, you've

24:28

demonstrated completion by us. That

24:30

apparently admirable tendency persistence,

24:33

the determination to finish what we start

24:36

well, it could be twisted and perverted

24:39

if we feel compelled to reach the

24:41

finish line, we also feel

24:43

tempted to choose a short racetrack.

24:46

There's more at stake here than making ourselves

24:48

feel better by cheating with our own to

24:50

do lists. Psychologists

24:53

recently studied completion by us in a high

24:55

stakes setting, a hospital emergency

24:58

department. They found that the busier

25:00

the emergency room becomes, the

25:02

more the doctors look for quick winds

25:05

the patients who aren't really very ill and

25:07

can therefore be treated swiftly and

25:09

ticked off the list, and this behavior

25:12

is counterproductive. The more seriously

25:14

ill patients wait longer, of course, and

25:17

the doctors start to slow down after working

25:20

through a lot of fairly trivial cases.

25:22

I expect we all know the feeling, but

25:24

in their subconscious desire to see

25:27

some work through to completion, doctors

25:30

were harming the patients who were

25:32

in greatest need. Claude

25:36

Shannon's willingness to set aside projects

25:39

starts to look like a strength rather

25:41

than a weakness. Shannons certainly

25:44

could focus, whether building information

25:46

theory from scratch or building

25:48

a wearable computer to be Droulette.

25:51

Yet Shannon also seemed to have an

25:53

inner confidence that allowed him to declare

25:56

victory at any point that suited him.

25:58

If a piece of work was not good enough to publish,

26:01

fine, he was happy to leave it unpublished.

26:04

That juggling paper is an example, but

26:07

so too was his early work on genetic

26:09

algebra. One of Claude

26:11

Channon's colleagues at Bell Labs praised

26:14

him as a man of infinite

26:16

courage. He was talking about

26:19

Shannon's intellectual daring, a

26:22

willingness to march into unknown

26:24

territory to begin the search for solutions

26:26

to problems that seemed as unbeatable

26:29

as Roulette. But perhaps

26:31

courage is not quite the right word to

26:33

describe Shannon's approach. I

26:35

prefer in soussions. Claude

26:38

Channon just wasn't worried. He

26:41

didn't feel completion by us the way

26:43

you and I feel it. He would walk

26:45

away from any project at any time

26:48

without regret. And if

26:50

he was willing to abandon a stalled project,

26:53

where was the risk? And if there

26:55

was little risk, why talk about

26:57

courage. Shannon didn't

26:59

need courage, He just needed

27:02

the ability to move on. In

27:06

August nineteen sixty one, Claude

27:09

and Betty Shannon met ed and Vivian

27:11

Thorpe in a hotel room in Las

27:13

Vegas. Claude and Ed

27:15

prepared the wearable computer system, which

27:18

required both of them to operate. Shannon

27:21

controlled the computer itself, the

27:23

size of a cigarette packet, with twelve

27:25

transistors in it. He used

27:27

his toes to trigger silent mercury

27:30

switches hidden in his shoes. Thorpe,

27:33

whose research into blackjack had given him

27:35

plenty of experience hanging around in casinos,

27:38

was the one who would place the bets. He

27:40

had a radio receiver and an earpiece

27:43

connected to a hair thin steel wire.

27:46

The earpiece played an ascending musical

27:48

scale. Shannon would use

27:51

the toe switches to time a rotation

27:53

of the wheel and then the counter rotation

27:55

of the ball from the moment it passed

27:57

a reference mark. Thorpe

28:00

would hear the musical scale stop

28:02

on a continuous note at the moment

28:04

that Shannon finished timing the rotation,

28:07

and the pitch of that continuous would

28:09

indicate in which part of the wheel the

28:11

ball was likely to drop. Thorpe

28:14

still had a few seconds to place bets

28:17

and collect the money. Thorpe

28:20

knew from hard experience that they

28:22

had to be careful. Their

28:24

device wasn't illegal, it was

28:27

far too inconceivable for that,

28:30

but it wouldn't go down well if discovered.

28:33

Beating the casino required more

28:35

than just beating the game. That's

28:37

why the Shannons and the Thorpes stroll

28:40

up to the table separately, pretending

28:42

not to know each other. It's why Claude

28:44

Shannon's scribbling numbers down, distracting

28:46

the floor manager from what he's really doing. All

28:49

the while he's gazing intently

28:52

at the wheel from under his dark eyebrows

28:55

and his toe, silently pressing

28:58

and releasing the hidden control

29:00

of the computer. And

29:03

while Thorpe is standing at the other end of

29:05

the table, cheerfully placing his bets

29:07

the earpieces, seeing the signals

29:10

from Shannon's little computer in

29:12

giving Thorpe predictions in the

29:14

form of musical tones, and

29:16

Thorpe is winning. Not

29:22

everything goes smoothly. The fine

29:24

wires to Thorpe's earpiece break several

29:27

times, requiring a trip to the bathroom

29:29

to fix them. At one moment,

29:31

a horrified observer sees the

29:33

earpiece come loose and thinks some strange

29:36

insect is crawling out of Thorpe's

29:38

ear But fundamentally the

29:41

computer works perfectly. The

29:43

chips are stacking up fast. At

29:47

the end of the visit to Vegas, the Shannons

29:50

and the Thorpes pondered their options.

29:53

Ed Thorpe was bullish. He'd beaten

29:55

the casinos before and was happy to do

29:57

it again, but Betty, Claude

29:59

and Vivian weren't so sure. It

30:02

had been an exhilarating day,

30:04

but a nerve racking one, and casino's

30:07

simply banned players who seemed to in too

30:09

much for any reason, so making

30:11

the computer pay on a regular basis would

30:14

require constantly concealing their identities.

30:17

Thorpe was forced to admit they had

30:19

a point. The computer clearly

30:22

worked, and in theory

30:24

they could use it to make millions, but

30:27

was it worth the effort and the risk. Shannon

30:31

and Thorpe had had their fun, and

30:33

they'd proved their point to their own

30:35

satisfaction, and Claude Shannon

30:37

had other projects to play with, so

30:41

after months of hard work, the

30:43

world's first wearable computer was

30:46

retired, undefeated after

30:49

a single trip to Vegas. Decades

30:53

later, Thorpe reflected,

30:56

I have always thought it was a good decision.

31:03

When I first thought about writing this cautionary

31:05

tale, I thought it would be a warning not

31:08

to lose focus like Shannon did. I've

31:11

changed my mind now. I think

31:13

Shannon and Thorpe are inspirational

31:16

figures. The cautionary tale

31:18

isn't a warning to keep your focus. Instead,

31:21

it's a warning not to focus too

31:23

much. Don't commit yourself so

31:26

totally to a project that you lose

31:28

heart, or lose sight of creative

31:30

ideas, or lose your freedom

31:32

to change course. There's

31:35

one last lesson I think we can draw

31:37

from Claude Shannon's ability to move on.

31:40

In their Vegas hotel room, as

31:43

Shannon equipped Thorpe with his earpiece

31:45

and the fine connecting wires,

31:47

Shannon had cocked his head to one side

31:50

and smiled impishly. What

31:52

makes you tick? It was a joke

31:54

about the fact that Thorpe was plugged into a

31:57

machine, but young Thorpe took

31:59

it as a deep question from an older

32:01

and wiser man. What did

32:03

make him tick? Professional

32:05

gambling, academic mathematics

32:08

or something else? But then

32:11

why choose? Shannons

32:13

seem to do it all, from academia

32:15

to juggling, and so in

32:17

the end, would ed Thorpe. You

32:20

can find interviews with him well into his eighties,

32:23

still as sharp as anything, reminiscing

32:25

about blackjack and academic mathematics,

32:28

and the hundreds of millions of dollars

32:30

he eventually made after analyzing

32:32

the patterns in financial markets

32:34

as one of the first quants. One

32:38

of the intriguing ideas in Claude Channon's

32:40

mathematical theory of communication is

32:43

that a message can be compressed to

32:45

the precise extent that

32:47

it is predictable. A movie

32:49

can be compressed because each frame tends

32:52

to resemble the previous one. A

32:54

compression algorithm doesn't store the new

32:56

frame. Instead, it

32:58

stores a series of dips changes

33:01

from the previous frame. Movies

33:04

with lots of cuts or fast dramatic

33:06

movements are harder to compress.

33:09

The same is true more or less with the way

33:11

we remember our lives. Although

33:13

the brain is not a video recorder and

33:15

doesn't store diffs, it

33:17

does compress memories by recalling

33:20

the gist of an experience. If

33:22

I get up in the morning at the usual time,

33:24

eat my regular breakfast, walk the usual

33:27

route to the station, and catch the same train

33:29

as always to the office, my brain

33:31

doesn't trouble itself to remember much. The

33:34

diffs aren't worth bothering with a

33:37

life that's too predictable, creates

33:39

few memories. That's what prisoners

33:42

sometimes say about their years behind bars.

33:44

Don't remember much because it was all

33:46

the same or the pandemic

33:49

lockdown for me and perhaps

33:51

for you, involved sitting in

33:53

the same seat, doing the same thing every

33:55

day. Life in lockdown

33:57

was thin and forgettable. The

34:01

opposite experience is a vivid vacation

34:04

somewhere, packed with new sights and smells,

34:06

the people, the language, the architecture,

34:09

the food. These complex,

34:11

rich experiences defy compression.

34:14

The diffs are too big, so

34:16

the memories are rich. Has it really

34:18

only been ten hours since I arrived,

34:20

you ask yourself. Feels like a week.

34:23

So if you want a full life rich

34:26

with memories, keep searching

34:28

for new experiences Like

34:30

Shannon, don't be afraid to declare

34:33

victory and start afresh. Shannon

34:36

did everything the jazz and the juggling

34:38

of the chess, the intellectual journey

34:40

from genetics to the Rubik's cube, the

34:43

joky robots, and the flame throwing trumpet,

34:45

and he really did turn upside

34:48

down the way the world thought about

34:50

digital information. Not once,

34:53

but twice isn't

34:55

twice enough. The

35:07

key sources for this episode were Jimmy

35:09

Sony and Rob Goodman's biography

35:12

of Claude Shannon, A Mind at

35:14

Play, and Edward Thorpe's autobiography

35:17

A Man for All Markets. For

35:19

a fullest of references, see Tim

35:22

Harford dot com.

35:26

Cautionary Tales is written by me

35:28

Tim Harford with Andrew Wright.

35:31

It's produced by Ryan Dilley and Marilyn

35:33

Rust. The sound design and original

35:36

music are the work of Pascal Wise.

35:39

Julia Barton edited the scripts.

35:41

Starring in this series of Cautionary

35:43

Tales are Helena Bonon, Carter

35:46

and Jeffrey Wright, alongside

35:48

Nazar Alderazzi, Ed Gochen,

35:51

Melanie Gutteridge, Rachel Hanshaw,

35:54

cobnor Holbrook, Smith, Reg

35:56

Lockett, Missia Munroe, and

35:59

Rufus Wright. The show would

36:01

not have been possible without the work of Mia

36:03

LaBelle, Jacob Weissberg, hell

36:06

Faine, John Schnarz, Carlie

36:08

mc loori, Eric Sandler, Emily

36:11

Rostock, Maggie Taylor, Daniellow

36:14

Lakhan, and Maya Canning. Cautionary

36:17

Tales is a production of Pushkin

36:19

Industries. If you like the show, please

36:22

remember to share, rate, and review.