One thing that Google does. Amazon made. Walmart made. And according to the news published, the US government did.
What are they doing? They analyzed the vast “big data” of virtual data collected and stored every day about almost every aspect of life, thereby identifying patterns of behavior, giving clues. correlation and predictive assessments.
Amazon uses customer data to make suggestions based on the history of old purchases. Google uses search data and other information it obtains to sell ads and serve other services and products.
The US National Security Agency is also collecting millions of call records from Verizon customers according to court-ordered confidentiality requirements – “no-one and in large numbers” and “regardless of whether they are doubt about bad behavior ”. According to The Guardian and The Washington Post, the agency also conducted another intelligence program called Prism, in which they collected data in email, audio chat, video chat, photos, documents, and usernames from Leading network companies such as Microsoft, Google, Yahoo, Facebook and Apple, aim to track targets outside the United States.
Why do you need to throw such a wide net just to find a small number of suspects terrorists? After that need to collect data on such a large scale? Jeremy Bash, a former Pentagon official, worked under Leon E. Panetta – former Director of the US Central Intelligence Agency (CIA) said: “If you find a needle in a haystack then you will need a haystack ”.
In the book “Big Data” (Vietnamese name: Big Data), Viktor Mayer-Schönberger, a professor of Internet administration at the Internet Research Institute of Oxford University, with Kenneth Cukier, data administrator (data editor) for The Economist, pointed out that the nature of intelligence operations has changed.
“In the spirit of Google and Facebook, we now have to recognize that people are a combination of social relationships, network interactions and content links. In order to fully investigate an individual, researchers need to look at the widest range that the dark half of the data can be centered around that individual – not just those they know, but also those who they know, and so on. ”
Cukier and Mayer-Schönberger point out that big data analysis is completely changing the way we see and handle the world – they even compare its consequences to the consequences of the printing process. Gutenberg created it. In this publication, they give readers a statistic that is both interesting and alarming about the growing influence of data growing almost all areas: business, state, science, medicine, privacy and even the way we think. They say that, when we understand the mechanisms, we can be aware of causality, which in turn leads to a correlation.
The data is growing rapidly – according to some sources, it has more than doubled each year – and the authors of this book claim that when storage costs drop to the arts The math is growing, data-crunching technologies that were previously used only in intelligence agencies, research labs and giant corporations, are now increasingly become more popular.
Big data has led to the emergence of a range of new companies, helping existing companies promote customer care and find new ways to grow. Thanks to big data, Walmart discovered that, before the storm hit, flashlight sales increased with Pop-Tarts sales (a sugary breakfast in the US). Since then, Walmart stores have started to sell Pop-Tarts boxes next to storm supplies, both to “make customers’ lives easier” and to help increase sales. The authors also refer to UPS installing the sensor system with GPS sensors so that employees can control, optimize the route and know when to maintain the vehicle.
Baseball teams such as Oakland A’s of Billy Beane have also successfully adopted a “data-grinding” approach to tracking athletes. Obama’s 2012 campaign also used complex data analysis to build a sophisticated political machine, to detect supporters and mobilize them to vote. New York City also uses data analysis to find new effective models in everything, from how to respond to natural disasters to identifying smuggled tobacco stores, to informing home inspectors. stay at the houses they need to notice most. In the coming years, as Mayer-Schönberger and Cukier argue, big data will gradually become “part of solutions to global issues such as climate change, epidemic eradication and consolidation.” main and economic development.
Of course, big data also has its dark side, and the authors have deftly analyzed the hazards that they anticipate will happen. Privacy protection will become increasingly difficult. In particular, old protection strategies such as “personal notice, voluntary agreement, cancellation and anonymity” will become less effective or even invalid.
The two authors write: “Every tool that we use every day, from websites to smartphone applications, has the ability to collect personal data.” Once there are millions of ways to reuse, re-target, and resell data to other companies, it’s hard for users to get permission before their data is “used for secondary purposes.” “, What they will not imagine when the data was originally collected.
The second danger caused Cukier and Mayer-Schönberger to worry that there are many similarities with the context from the fictional film “Minority Report”, in which we can make key predictions. so accurate that we can order people to arrest before they commit a crime. The authors suggest that in the near future, large data analysis (which in the film is a modern Pre-Cogs device) may give rise to the situation that “judging that someone who is guilty can be based on predicting their future behavior ”.
Currently insurance companies and release rulers have used data analysis to predict risks, and according to the authors, more and more places in the US use “forecasting policies.” “With” crushed data “to” choose which streets, groups and individuals should be more strictly controlled, because the algorithm has shown that the area is capable of high crime. ”
In an NBC report on an air raid, “the CIA does not necessarily know who it is killing”: in air strikes, “intelligence officers and unmanned aircraft operators can kill suspects just based on their behavior patterns – without positive identification. ”
Cukier and Mayer-Schönberger also point out, another problem related to depending on predictions of potential behavior is that it can deny “the idea of innocent speculation.”
“If we can force people to be responsible for future behaviors that are predicted, things that they may never do, then we deny the chance of moral choice. of human.”
At the same time, they also observed that large data exacerbated “a problem that has existed for a long time: depends on data, while they have a much higher likelihood of deviation than with us can imagine “. They lead to escalation in the Vietnam War under the direction of Robert S. McNamara (former Secretary of Defense under President John F. Kennedy and Lyndon B. Johnson) as a case study case pictured in “data analysis errors”: As a pioneer in data analysis, McNamara relied too much on matrices as the number of troops to measure the progress of the war; while it is clear that the Vietnam War is a war of will rather than numbers.
More recently, we can see many other failures of data analysis, including the collapse of Wall Street in 2009, when it fell into serious crisis due to extremely complicated trading systems. depends on the algorithm. In the 2012 best-seller book “The Signal and the Noise”, statistician Nate Silver pointed out the mistakes of the data in areas such as science. earthquake, finance and biophysical research. He also argued that “forecasting in the age of Big Data” was “not going too smoothly” (although he himself was very successful in making forecasts in areas such as politics and baseball).
At the same time, as a computer scientist and musician Jaron Lanier pointed out in his excellent new book: “Who Owns the Future”, there is one thing. The big difference between “big data of science, such as data on galaxy formation, nature, flu outbreak” – which requires a lot of research, with “big data about people “- which are things that are uncertain, contradictory and often unreliable as anything related to humans ever.
It is worth noting that Cukier and Mayer-Schönberger also recognized the limitations of the data. This is a book that can help us see the importance of big data as a tool to “quantify and understand the world”, but it also warns us not to become victims of ” data dictatorship ”.
“We should be careful not to rely too much on data, instead of repeating the mistake of Icarus, who was too confident in his flying ability to use it the wrong way and fell into the deep sea.”
ITZone via tramdoc