William Poundstone - Stone Breaks Scissors. How to outwit anyone: a practical guide

William Poundstone

The stone breaks the scissors. How to outwit anyone: a practical guide

Predicting fateful accidents seems easy. In fact, this is not so, and one of the reasons is that in matters of almost random sequences, intuition often deceives us. Incorrect prediction can lead to tragedy.

In recent decades, psychologists have studied an object that at first glance seems abstract - a person's perception of chance. They studied how we make random, or arbitrary, choices, and how we predict unpredictable events (such as the situation on the stock market, the outcome of a basketball game, and the “future”). It turned out that this subject is of great practical importance. One way or another, we all make predictions. They can be simple, like in the game “stone, scissors, paper”, when the loser pays the bill in the bar. The stone breaks the scissors. The best predictor wins. Anticipating the thoughts and actions of others is very important for winning an argument or in a game, in order to get a date or a promotion, to get rich. The success of a personal life or business often depends on whether you are more accurate than others in your forecasts.

In this book you will learn how to use psychology to improve the quality of predictions. In particular, it shows how to predict the behavior of people trying to be unpredictable. This will require a practical approach, in particular, a description of how a few simple principles can be applied to a large number of everyday situations. You will learn that outwitting a person is easy, fun and often profitable. Here are some examples.

Tests with a choice of several possible answers. Compilers of the test try to arrange the answers in random order. In most cases, this fails, which makes it possible to gain an advantage in guessing.

Tote. Your colleagues are betting on the result of football matches, on the position in the standings of the National Student Sports Association, on the award of an Oscar. Their choice is more or less predictable. You can win a bet by predicting what others will bet and building your strategy on it.

Games and strategy. In almost every game - from tennis to poker and “rock, scissors, paper” - there is a prediction element. A player anticipating an opponent’s strategy can earn points and win.

Recognition of financial fraud. “White-collar crime” is always associated with fictitious figures: inflated funds for hospitality expenses, fake profit and loss statements, distorted tax returns. When people come up with numbers, these numbers obey a predictable pattern. Anyone who recognizes this pattern can quickly check the accuracy of financial documents.

Investments. An investor or real estate buyer who realizes that market fluctuations over very long periods are quite predictable, will be able to outperform the majority and exceed the average market performance.

All of the practical applications in this book are based on one simple idea. When people make arbitrary, random, or strategic choices, they subconsciously adhere to certain patterns that can be predicted.

Part one

Experimental modeling of randomness

Zenith Broadcasts

"Commander" Eugene Francis MacDonald Jr. loved cage suits and a cocktail of gin and pistachio ice cream. He lived on his 185-foot Mizpah yacht moored in a Lincoln Park Chicago yacht marina. As the CEO of Zenith Radio Company, he lived so interestingly that any major businessman could only envy. The range of his interests was unusually wide - from Arctic research to the search for pirated gold.

MacDonald's main contribution to American business is a publicity stunt. In 1934, he sent a telegram to all tire manufacturers and oil companies: “PAY ATTENTION TO THE ABSENCE OF PEOPLE IN THE STREETS BETWEEN THE ELEVEN AND HALF OF THE TWELFTH AT THE PRESIDENT'S SPEECH”. Indeed, during Franklin Delano Roosevelt’s intimate conversation with the radio listeners, the streets were empty. The telegram was followed by letters telling about the influence of the radio. B. F. Goodrich Company has agreed to sell Zenith radios through its dealer network of 1,200 tire stores. After the stock market crash, many radio stores went bankrupt, making room for Zenith.

So MacDonald invented hidden advertising. He supplied Zenith radios for Hollywood movie studios. In 1929, after the beginning of the television era, Zenith radios began appearing in television films, from musicals directed by Busby Berkeley to The Night of the Living Dead? (Night of the Living Dead). Could they be seen in films about the war, in "comedians of cranks", in "black cinema" and in the television series "Three Dunce"? (Three Stooges). In one of the episodes of The Three Dunce, Curly is hit on the head with a Zenith radio - it must be expressing the feelings of regular movie theaters.

In 1937, the radio network was at the peak of influence. A few words from NBC news host Gerba Morrison was enough to destroy an entire industry. “It flashed,” breathed Morrison, watching the disaster of the Hindenburg airship. - Oh my God!" After that, no one wanted to fly on airships. In 1937, Arturo Toscanini? led the NBC Radio Orchestra, and young Orson Welles voiced a popular comic book and novel hero named Shadow. McDonald’s most successful publicity stunt used the power of the radio, but not one of the 1937 radio programs could compare with the show invented by the “Commander” MacDonald.

Across the country, Zenith dealers began distributing decks of cards. During the years of the Great Depression, it was difficult to pass by a free deck, but these were not intended for a normal game. Their shirt shimmered with a bright pattern with the Zenith logo and the following words: DEVELOPED IN THE PARAPSYCHOLOGICAL LABORATORY OF DUKE UNIVERSITY. MacDonald expected to make money on a craze for extrasensory perception, and the cards advertised a new Sunday evening radio series.

In the mid 1930s Joseph Banks Rhine attracted the attention of the whole country with psychological experiments at Duke University. He performed sessions of telepathy, clairvoyance and telekinesis. Piercing eyes and gray hair gave Rain - a botanist by profession - a special persuasiveness. The press reviews, from New Yorker to Scientific American, were mostly friendly. As one of the reporters noted, “Rain made an extrasensory perception a craze for women's clubs across America.”

... On a hot June evening, Rain and his wife dined on board the MacDonald's yacht. The owner of Zenith wrote down his ideas for a psychic perception test, which he was going to conduct using radio. Listeners got the opportunity to test the possibilities of their own psyche. This would be the largest experiment in history and the best possible evidence of the existence of telepathy.

Rain was not sure the new science was ready for prime time. Skeptics suspected that Rain is reporting success and hiding failures (in other words, some of the "telepaths" are cheating).

But skeptics did not bother MacDonald. As one of his business partners put it, "nothing will stop the street crowd that loves to fight." MacDonald portrayed Mephistopheles, seducing Rain with the prospect of earning on telepathy. He said he had instructed one of his lawyers to take up copyright protection and trademark registration of cards that Rain used to identify extrasensory perceptions. This is the so-called Zener deck, named after Rain's colleague and consisting of cards with five repeating patterns (circle, cross, three wavy lines, square, five-pointed star). Rain will receive a percentage on each deck sold, McDonald promised, and cards (five and ten cents) will be sold in small goods stores.

Raine hesitated. In the end, he allowed himself to be referred to as a “consultant,” realizing that other psychologists would be observing the experiments. MacDonald agreed.

A half-hour show called The Zenith Foundation took place on NBC’s Blue Network on September 5, 1937 at 10:00 p.m. By the name of the show, it was impossible to guess the content. Advertising enticed: "The program is so UNUSUAL - so AMAZING - so interesting - that it will become familiar to the whole country." The word “foundation” was associated with philanthropy, such as the Rockefeller Foundation, but MacDonald saw no obstacles to the peaceful coexistence of public service and profit. In an information letter sent to dealers, it was reported: “Zenit Foundation broadcasts are intended to increase sales of Zenith radios ... You should take the opportunity to the maximum. Double your efforts. ”

MacDonald was afraid that the word "telepathy" would alienate the most practical listeners, and therefore almost was not mentioned in the first programs. Initially, the plots were dedicated to prominent thinkers whose ideas were subjected to undeserved ridicule. Within a few weeks, the program developed a pattern familiar in the modern world of cable television - a dramatic story about physical phenomena that are supposedly real, and comments by a motley group of “experts”.

A new element - the MacDonald telepathic experiment - was introduced in the fourth program. A group of ten “transmitters” from a locked studio in Chicago tried to send thoughts to listeners all over the country on the radio. Listeners were invited to record their feelings and send the texts by mail.

In the first test, September 26, the "transmitters" broadcast a random sequence of two colors, black and white. To eliminate fraud, the choice was made directly during the broadcast using the roulette wheel.

Host: It’s best to record the sensations right away. Do not meditate or try to find logic. Record your feelings in sequence - at the same speed as they arise. The machine is ready to select the first number.

ROTATING ... STOP ... SIGNAL ... PAUSE ... SIGNAL

Host: It was the first number. Now the car will choose number two ...

When the letters from the radio listeners began to arrive, almost immediately it became clear that something unusual was happening. It was necessary to guess the sequence of five elements, alternating black and white. Most listeners made a mistake in only one case. Probably pleased Rain was relieved - the result testified in his favor.

After the first test, Woolworth’s department store sold all the decks of cards available in the warehouse and placed an additional order. The symbols depicted on the cards were used in several of the following tests. It is said that during the existence of the program 150 thousand decks of cards were printed. They can still be bought on eBay today.

Next week, a choice of five vegetables was to be made: carrots, beans, peas, corn and beets. This complicated the task, because for each position in the sequence there were five options. Most of the listeners guessed in two cases out of five - a satisfactory result, twice the probability with a random choice.

In the next two shows, two colors were again used, black and white. On October 10, the majority guessed four out of five options, and on October 17, five out of seven.

On October 24, radio listeners were offered a choice between a circle and a cross. The transmitted sequence looked like OXXOX, and most of the answers were absolutely correct.

This does not mean that each individual listener guessed the sequence with such accuracy. But for some reason, most of the answers turned out to be accurate - was it really mass telepathy? In many ways, the aggregate results looked more impressive than the results of individuals. Given that the parapsychological game is statistically significant, the Zenith experiment can be compared with a powerful microscope or supercollider, which can accurately detect weak effects. For 15 weeks, while the transmission cycle continued, the studio collected more than a million responses, which was the most ambitious test of extrasensory perception ever conducted. In many programs, the statistical significance of the correct answers from the listeners was incredibly high. Subsequently, the Zenith Foundation issued a report stating that the likelihood of such a coincidence of results is 10,000,000,000,000,000,000 to one. But the listeners did not need this suspiciously round figure to feel that they had become participants in something supernatural.

Zenith invited several well-known psychologists to design and conduct the experiment. Behind the scenes there was a hot fight.

Rain himself preferred to stay away from the radio show - it was not difficult because he was in his laboratory in Durham, North Carolina - and therefore the role of experts fell to the role of two psychologists from Northwestern University, Harvey Gault and Louis D. Goodfellow. Gault, who had only a few years to retire, had long been interested in experimenting with telepathy. Goodfellow was a young psychologist from the Gault department. He wore professorial glasses, and combed his hair in the middle. Both shared MacDonald's belief that radio provides a unique opportunity to test the reality of telepathy.

It was not difficult to repeat Rain's experiments, since they only needed a deck of cards and a senior student who agreed to donate an hour of time. Psychologists were disappointed. In science, the inability to confirm the discovery is fatal. In real life, things are not so simple. Rain claimed that telepathy is a very subtle thing. It is not 100 percent accurate and cannot manifest itself in any person at any time. Failure to try to replicate Rain's results may simply mean that the subjects are devoid of the gift.

Goodfellow and Rain had been arguing for a long time about details that were substantial and not very. Gault was annoyed by both. After several first transmissions, Goodfellow discovered a curious thing that infuriated Raina. He could predict the answers of the listeners!

Very interesting, but not what McDonald and Rain wanted to hear. Goodfell’s opinion threatened the industry of extrasensory perception, which brought more and more profit (oh, these people!). Goodfellow was declared the enemy of the paranormal and dismissed as an expert. Meanwhile, a program dedicated to extrasensory perception began to annoy radio listeners, and its rating declined. In early 1938, MacDonald closed the program.

Goodfellow, independently of the rest, published Zenith results in the journal Journal of Experimental Psychology. He offered a convincing explanation of the results, which had nothing to do with extrasensory perception. Time magazine wrote that Goodfellow "pierced the rainbow bubble of MacDonald's telepathy." In addition, Goodfellow dispelled some of the myths planted in the program. It told how a psychic led the police to the body of a dead woman buried in a woodshed. Goodfellow found the records of the court hearings and found out: the body was found at the prompt of a boy peeping through a hole in the board.

After this, the internecine strife between psychologists began to resemble a children's quarrel. Goodfellow, whose character did not fully correspond to the surname?, Launched an attack on Raina, hiding behind a pseudonym. Chicago-based board game maker Cadaco-Ellis has unveiled a new game called Telepathy. It was invented by a certain "Dr. Ogden Reed", and in the instructions to it, Rain's science was characterized as "replete with evasions." Without any telepathy, Rhine realized that "Dr. Ogden Reed" was actually Dr. Louis Goodfellow. “Is it permissible,” wrote Raine Goodfellow, “to a scientist to use fraudulent methods (in this case, a fictitious name) in order to avoid responsibility for his words?”

MacDonald was furious. He told Rain that he should sue the game maker, and promised to pay legal fees.

Cards that brought Rain a percentage of profits became a constant source of trouble for him. One of the complaints was this: paper was wrinkled from the paint used in the production of cards. In order to reduce costs, they were printed on such thin paper that newly appeared psychics could see through it. Psychologist B. F. Skinner “guessed” 23 out of 25 cards - to the delight of students. This made Rain an object of ridicule, although he had nothing to do with cheap cards and used others in his laboratory.

“Rain and Goodfellow sent me copies of their“ love letters, ”wrote Gault, chief psychologist for the show, to MacDonaldo. - I am not surprised that R. climbed into the bottle. Strictly between us, I do not care about his attacks against G. The latter is an excellent specialist, and as such he is very useful to me. But in some ways he’s a complete fool. ”

Judging by these words, Goodfellow did not enjoy much love at Northwestern University. During the war, he left there and became the director of the center for training technical personnel for aviation. Then he got a place at the Department of Psychology at the University of Pennsylvania in the hometown of Altoona. The rest of his professional career passed in a quiet and calm backwater - he taught, published good work, but did not achieve anything that would attract general attention. Today, Goodfellow is remembered almost exclusively by Zenith's experiment. It is also popular with those who classify themselves as scientific skeptics - along with Harry Houdini or James Randy. Demonstrating that reading thoughts on the radio is a hoax, Goodfellow opened up real reading of thoughts.

Goodfellow didn’t even try to do what the listeners were doing - to catch the broadcast thoughts of the “transmitter”. They were determined by the position of the roulette wheel and really were random in nature, as Goodfellow himself was convinced. Instead, he predicted the responses of radio listeners trying to guess a random sequence.

During the first broadcast, psychologists tricked the listeners into believing that seven options were broadcast. In fact, there were only five. For the third and seventh “broadcasts”, members of the group were simply asked to quickly consider to themselves and not think about two “broadcast” versions, black and white.

Books "/\u003e

Here is a practical guide: how to benefit from the predictable behavior of other people and use this skill in everyday life. After reading this book, you will learn how to predict the actions of friends, bosses, teachers, competitors, and just others. Most people are programmed to make certain choices and follow simple trends, and therefore their behavior is predictable even when they try to be original. William Poundstone has the gift of extracting practical advice from psychology and behavioral economics and proves that forecasting is easy, pleasant, and sometimes profitable! ISBN: 978-5-389-08659-3

Publisher: "ABC-Atticus" (2014)

ISBN: 978-5-389-08659-3

Other related books:

Author	Book	Description	Year	Price	Book type
Poundstone william	Stone breaks scissors	Here is a practical guide: how to benefit from the predictable behavior of other people and use this skill in everyday life. After reading this book, you will learn how to predict ... - ABC Atticus,		299	paper book
William Poundstone	The stone breaks the scissors. How to outwit anyone: a practical guide	Here is a practical guide: how to benefit from the predictable behavior of other people and use this skill in everyday life. After reading this book, you will learn how to predict ... - ABC Atticus, e-book	2014	249	eBook

Prologue
Prediction Machine

Upset, Hagelbarger took a few days off to read and think. He decided that he would now take up the mind-reading machine.

Announcing a decision out loud was part of the play. In the 1950s no machine could recognize a human voice. She made a choice before the player opened his mouth.

(a) The person’s choice is not accidental. The sequence of moves in the game is influenced by experience and emotions. For example, some people, having won twice in a row, are afraid to “frighten off luck” and repeat the actions. Others, on the contrary, do not want to “tempt fate” and change their mind. But in both cases the car will catch them.

(b) In order to confuse the opponent, the machine will try to predict its actions only if it is won, and if it loses, it will make a random choice.

Shannon's interest in languages \u200b\u200binspired Pierce, under the pseudonym J.J. Kopling, to write an article about the researcher for the journal Astounding Science Fiction. There, he also presented Shannon's ideas for computer music. Subsequently, music theorists suggested that the listener constantly foresees the next few notes based on the previous few. The perception of music largely depends on how it meets or does not meet our expectations.

A normal genius would spend the rest of his creative life cultivating the fertile field that he plowed the very first. But Shannon, having created a fundamental work, abandoned the theory of information. His interests have shifted to the field of computers and, to some extent, the human mind. “We hope,” he once wrote, “that research in the field of gaming machines will help us understand how the human brain works.”

Shannon spent a lot of time designing the most incredible devices. In 1950, he created one of the first machines capable of playing chess, and later a pair of mechanical hands that could assemble a Rubik's cube. Shannon's desktop calculator called THROBAC worked with Roman numerals. But his most famous invention was Theseus, a mechanical mouse that found its way into an aluminum maze. A short film was made about Theseus, after which Shannon gained stardom.

Then, approximately in 1952, the “absolute machine” appeared. A curious observer was asked to press the toggle switch to turn it on. After that, a mechanical arm extended, returned the toggle switch to its original position, and again hid. This surreal action was doomed to become a meme. Try typing in Google "Absolute machine". On the market and in scientific museums you can find a huge number of copies and imitations of it. There’s even a YouTube video with an “absolute machine” made from Lego blocks.

Like many shy people, Shannon was prone to extravagant antics. It is known that he rode along the corridors of the Bell Labs research center on a unicycle, sometimes while juggling. For a while, he used stilts as an alternative means of transportation from cabinet to cabinet. Looking from the side, one would think that he was transformed from a scientific genius into an illusionist clown. However, he explored, in his own way, deep problems. One of them was formulated as follows: how complicated should a machine be to outwit a person? As his colleague David Slepyan put it: “In intellectual terms, he was the best fraudster in the world.”

Among the few Bell Labs employees that Shannon welcomed was Dave Hagelbarger. They often dined together in the Hagelbarger laboratory, having fun with all sorts of ingenious devices. Once, the Hagelbarger connected Shannon to an electroencephalograph to check if his mind’s brilliance reflected on the configuration of the curves. Nothing unusual. Then they connected the encephalograph to a prediction instrument. The resulting curves were similar to signals coming from Shannon! It turned out that the engine inside the machine rotates at a speed that coincides with the period of the alpha rhythm of the human brain.

Of course, Shannon liked the prediction machine, and he decided to design his own. His device was not a copy - it surpassed the prototype. In a long series of games with the same person, Shannon's machine won 65 percent of the time. Her superiority over man was obvious, and this news became the main one in the smoking rooms of the laboratory. In the 1950s a series of brilliant and ambitious researchers, engineers, and mathematicians went through Bell Labs. Shannon’s predictive machine became their “sword in stone.” Each claiming outstanding intelligence considered it his duty to challenge Shannon. Few were able to overcome the temptation, and even fewer were those who escaped humiliation. The tension was intensified by the fact that only Shannon himself could outplay the predictor machine.

He described his device on March 18, 1953 in a memo entitled "A machine for reading thoughts (?)." It was noted there that the game of coincidence has a glorious history and even found reflection in literature. It was “analyzed from the point of view of game theory by [John] von Neumann and [Oscar] Morgenstern, and from the point of view of psychology by Edgar Allan Poe in the story“ The Stolen Letter ”. Oddly enough, the machine uses the Po method rather than Neumann. ” The hero of the psychological detective, composed by Poe, reveals crimes, initially assuming that people, trying to imitate randomness, behave predictably.

Nowadays, supercomputers play chess, and there is nothing surprising in the victory of a machine over a person. But in the 1950s. the success of the machine seemed almost magical, and the term "mind-reading machine" used by Shannon was based on the reaction of most people. The longer a person played with Shannon's machine, the better she guessed his thoughts.

Bell Labs' Manfred Schroeder wanted to show off and showed the car to a math lab guest, Fritz Hirzebruch. Hirzebruch won the first 13 games in a row. Has Shannon’s car really found a worthy rival?

Nothing like this. The car won the 14th game. And 16 of the following 17, ahead of the famous mathematician. Hirzebruch continued to play, but in vain. He never managed to keep in the account.

For many years now and then I came across references to Shannon's predictive machine, and finally wondered if it had survived. I knew that Shannon was a flea marketer, littering his homes with toys, various devices and memorabilia. After his death, the family donated all these treasures to the Museum of the Massachusetts Institute of Technology. I looked at the museum’s website and found what I was looking for - “a machine playing coin comparison.”

Despite the importance of this machine in the history of artificial intelligence, it is not in a permanent exhibition. To see her, I had to visit the storerooms of the museum - a brick building with no windows the size of a Costco warehouse store in Somerville. It houses a curious collection of incredible inventions and a variety of props for university draws. For example, a huge can of Jolt Cola energy drink hung from the ceiling beams, like a stuffed crocodile in an old Kunstkamera.

Shannon's predictive machine is a plexiglass box measuring approximately 30 × 30 cm, with an opaque black lid slightly tilted towards the player for better visibility. The upper square part is made with a fair amount of humor and resembles a face. Two bulbs instead of eyes, a button instead of a nose and a red toggle switch in the black slit of the mouth. The device is portable, although heavier than a laptop. On the left is a strong handle for transportation.

Plexiglass housing allows you to see the design from all sides. I was even able to turn the car over and peek through the transparent bottom. The wiring harness is braided with a neat pigtail, as in the old AT & T switchboard.

Shannon replaced the rows of Hagelbarger bulbs with a retro-style counter. It uses the same principle as Newton’s Pendulum, a table toy that consists of several steel balls suspended from a thread. Depending on who wins, the car or the player, a steel ball shoots into one of two glass tubes. The ball transfers energy to several exactly the same balls, the last of them goes to the column for scoring. Each victory is accompanied by a loud click.

“To get an idea of \u200b\u200bthe mental activity of a predictive machine, think about this: a person has 1010 neurons, the most dumb ant warrior has 200 neurons, and this machine has less than 100 relays.” So described his car Hagelbarger. Shannon halved the number of relays. The memory of his device was only 16 bits, that is, 2 bytes or 0.0000000018 gigabytes. And this turned out to be enough to defeat a person, even if the player developed a strategy and thought, cunning and wise.

On Shannon's car, two choices are labeled “right” and “left.” In the first batch, the machine makes a random choice using a rapidly rotating switch - a kind of mechanical roulette. Then gradually reveals unconscious patterns in the actions of the opponent. A convincing victory of the information accounting system over the program! Imagine a chess player who wrote down all of Garry Kasparov’s answers to Blumenfeld’s gambit in his games: this will predict Kasparov’s moves at the next gambit draw.

Both Bell Labs prediction machines split the coin comparison game into eight standard situations. I will give an example of one. Suppose you win twice in a row by making the same choice. What will be your next move? You can not back down from a successful strategy or make a different choice, perhaps on the grounds that three identical moves in a row do not look like an accident.

When faced with this situation, the machine each time remembers the opponent’s decision. This solution is encoded as “1” or “0” and stored in one of 16 bits of memory. For each of the eight standard situations, Shannon's machine only remembers the last two decisions. This information occupies all 16-bit memory.

When the machine needs to make a prediction, it watches how the opponent did the previous two times. If his actions coincide, the machine believes that now he will do exactly the same. Otherwise, the choice of the machine is random - the same continuously rotating roulette wheel.

The main difference between the Shannon machine and the Hagelbarger device is simplicity. The Hagelbarger machine calculated the outcome of eight standard situations as a percentage. The higher the percentage, the more likely it was to predict a repetition of the past. It may seem that this is more logical and elegant than the “all or nothing” principle used by Shannon, but in practice his device predicted better.

When changing the enemy, both machines took some time to build the game. They should have compiled a digital dossier. Fritz Hirzebruch was probably lucky at the first stage, since Shannon's car had to make a random choice every or almost every time.

As if children who were organizing a toy war, Shannon and Hagelbarger decided to drive cars against each other. They constructed an “intermediary” that generated the same random sequences for both devices. “All three cars were connected,” said Shannon, “and left to work for several hours, which was accompanied by a bet on small amounts and loud cries of approval.” To Shannon's delight, his car won with a result of 55:45 (in percent).

When Hagelbarger was about to publish the results of his work, AT & T found that the name “prediction machine” sounds frivolous. Acronyms were very popular in those days, and the Hagelbarger renamed his brainchild into SEER, which stands for "robot - an extrapolator of sequences." Decide for yourself whether it’s more serious or not. Be that as it may, that was the title of the article in Transactions on Electronic Computers magazine. In it, the Hagelbarger asked the obvious question:

“Why design such a machine? Playing with her is not particularly interesting and almost or not at all profitable. By changing the inscriptions on the panel, we can turn it from an opponent into a servant trying to please the operator. ”

A digital servant able to predict the needs and desires of the user can be extremely useful. Hagelbarger gives an example:

“Surely it would be economically viable to create a central telephone exchange to measure traffic and adjust for it. Such a telephone exchange, for example, could track that most calls from the business district of the city fall in the daytime, and from residential quarters in the evening, and switch accordingly, but at the same time, restructure if there is a strong fire in the business district at night.

Perhaps in an unusually difficult situation it will be easier to build a machine that learns to be effective than to design an efficient machine. ”

The Hagelbarger prophecy has come true. This is exactly what happened with the telephone business - and indeed with the business in general. Creating machines that learn to be efficient is the idea of \u200b\u200bthe 21st century.

The prediction machine testifies to the human soul no less than about technology. We all constantly try to predict the actions of others, reserve the right to some unpredictability. The predictive car is a caricature of our tricks: for her, people are ridiculously mechanistic, they have a short memory and a lack of sophistication. Strategic decisions are based on what brought success or failure in the past, as well as the year before last. The success of the machine is proof that this conclusion is not so far from the truth.

The key conjecture of Hagelbarger and Shannon was the idea that people are not able to behave spontaneously. Andy Warhol's “screen test” became an analogue of the predictive machine for the right hemisphere of the brain. Warhol shot an icon of 1960s pop culture on a black and white camera without sound. - Bob Dylan, Susan Sontag, Allen Ginsberg, Yoko Ono and Dennis Hopper. He told them to do nothing at all. For any actor, this is a nightmare, because neither text nor action. Take a look at the results and you will see that almost everyone is resorting to a set of tricks as old as the world. Warhol characters swallow, blink, purs their lips, straighten an almost perfect hairstyle. Most try to look natural. Some choose the opposite line of behavior, grimacing in front of the camera or gesturing exaggeratedly. It takes a few seconds ... and then? Everyone can see the standard signs of awkwardness. Trying to remain calm, they equally showed concern.

Similarly, Bell Labs science stars used a limited set of tricks to mimic random selection. Applying them, they went further away from chance. And they were powerless to stop the car from predicting their behavior.

The exception was Shannon - only one person was able to outwit the car. He revealed his secret in 1953 in a memo. Like a Zen archer, Shannon became a machine. He mentally performed the operations that the machine performs, calculating its predictions. Then he did the opposite. “Running this program in your mind is very difficult,” Shannon admitted with mock humility.

The design of the machine allows those who emulate its work to win in 75 percent of cases (not 100 percent, because sometimes the machine makes random choices). Shannon managed to defeat the car in 60 percent of cases.

They say that some guests of Bell Labs before the start of the game described the principle of the machine, but even this did not help. Odometers appeared on the front of Shannon's car, showing the total score. Under them were attached paper ribbons with pencil inscriptions: “Player” and “Machine”. The final score left for posterity is: Player 3507 - Machine 5010.

Looking into the "face" of the car, I finally understood. The red toggle switch is the language. Shannon's machine shows language to humanity.

Today, predictive devices surround us from all sides. Perhaps one of them is in your smartphone. Talking apps like Apple’s Siri look more humanized than they really are, since human actions are more mechanical than they seem. Siri is able to anticipate many requests thanks to constantly updated statistics that record what questions the owners type on their phones and in which situations. This reinforces the illusion that Siri understands the user (the name Siri comes from SRI International, a former Stanford Research Institute, once known for physics research funded by the CIA).

But the most serious predictive machine known as Big Data is comprehensive algorithms that make it possible for us to track all of our activities in a digital environment to predict which purchases we can drive you to. Apparently, Shannon and Hagelbarger devices used cookies or archives of previous actions for the first time to predict the next actions. Shannon's small car with a human face offered a game that you could play or not play; predictions of faceless “big data” are hard to ignore.

A few years ago, a resident of Minnesota was convinced of this from their own experience. Running into Target department store on the outskirts of Minneapolis, he demanded to call the director. “My daughter got it in the mail!” He shouted. The director looked at what the buyer brought. A standard Target mailing list, similar to millions of other mailing lists, addressed to a client’s daughter. The booklet looked quite innocent - photographs of happy babies, children's furniture and clothes for expectant mothers.

“Are you pushing her to get pregnant?” - the client was indignant. His daughter was in high school and, of course, was not married.

The director apologized and promised to sort it out. It turned out that Target uses proactive analysis. The company collects all information about customers: visits to the site on the Internet, visits to real, not virtual stores, calls to customer support, use of coupons or discounts. Then the program analyzes all this “haystack” in order to find the “needles” of pure gold. This allows the seller to make specific, practical forecasts of the behavior of each client.

One of the secret initiatives was to predict a client’s pregnancy. Expectant mothers need a huge amount of goods that they have not bought before. Therefore, pregnant women are especially susceptible to advertising, discounts and everything else, which may motivate them to go shopping at Target. A customer who is used to relying on Target during pregnancy may want to continue to use the services of the store for decades.

Target’s predictions about pregnancy were far more accurate than just guessing, but certainly not 100 percent. Several errors were considered perfectly acceptable. An awkward situation occurred only in those cases when the client was really upset by the wrong prediction.

A few days later, the director called the annoyed client again to apologize again.

“I talked with my daughter,” he answered. - In my house there were events that I did not know about. Give birth to her in August. I must apologize. ”

This is a completely new situation. Department store software is able to determine that a woman is pregnant and her father is not capable. How to react to this? Admire the sophistication of algorithms or reflect on your inability to hear and understand each other?

The advantage of proactive analysis is that, using the program, you can find relationships in large databases that are invisible at first glance. They may not have explicit logic or reason. Target's pregnancy prediction algorithm is based on the purchase of 25 products, including odorless lotions and soaps, dietary supplements with calcium, magnesium and zinc, cotton swabs and hand antiseptics. None of these products alone mean anything. A 50-year-old bachelor can buy a zinc supplement. But if a woman buys several products from the list, this indicates a high probability of pregnancy. Target is not only able to predict the pregnancy of clients, but also to calculate the date of birth with an accuracy of one or two weeks.

Proactive analysis can really be called a kind of reading of thoughts, although its purpose is not to confuse you at all. The organizations using it make sure that you do not find out what your actions predict, they manipulate you. They say that Visa is able to predict which of the married couples, cardholders, are likely to divorce, and takes this factor into account in predicting non-return of debts. Needless to say, it would be tactless to inform potentially unhappy couples about this.

“As for the goods for pregnant women,” one of the Target executives explained, “we were convinced that some women react negatively. Then we began to mix these products with others, obviously unnecessary, so that the advertising of children's products looked random. Lawn mowers were placed next to the diapers; next to the clothes for the newborn, a coupon for wine glasses. As a result, everything looked as if the sample was random. And we found: if a pregnant woman does not suspect that she is being spied on, then she uses our coupons, believing that all the other residents of the quarter receive exactly the same newsletter with the offer of diapers and cribs. If we don’t scare her, then everything works. ”

Consumer behavior - a sequence of impulsive purchases that fit into the framework of economic necessity. We ourselves do not always know in advance what we are going to buy, and we are confused by the thought that someone is able to predict our purchases. But no one complains about the owner of a small store that knows its customers and gives them advice. The difference is that the modern digital recommendation is derived using an algorithm, which, as we know, is quite simple. This serves as an unpleasant reminder of how mechanistic our thinking and decisions can be. New means of prediction challenge notions of privacy, but also of freedom. In a consumer society, buying is the highest expression of free will. Will I become less free if the website on the Internet is able to predict which shoes or movies I will choose? At the end of the day, you can either get what you want, or enjoy complete, existential freedom. Combine will not work.

Related Articles:
An American soldier told why Russians are so scared in the USA The Russians never liked the war. But on the other hand, one cannot “lose ... An American soldier told why Russians are so scared in the USA The Russians never liked the war. But on the other hand, one cannot “lose ... In the territory of the Kazakh Assr

William Poundstone - Stone Breaks Scissors. How to outwit anyone: a practical guide

Other related books:

See also in other dictionaries:

Prologue Prediction Machine

Prologue
Prediction Machine