Big data that doesn't pay attention to human nature, just Big Bluff

&http://www.aliyun.com/zixun/aggregation/37954.html ">nbsp;

Stalin once said: "A person's death is a tragedy, 1 million people's death is the data." If you take the terminology of the medical profession, this is a kind of empathy fatigue, and if you switch to the most popular term, we still can't deal with big data.





last week to participate in the Tencent idea, the theme is "How big data will affect social change." field for large data, proposed a different voice, there is "Data Island theory": the existing large data is broken and closed, such as Tencent said that they have a certain aspect of the full data, but whether there are Baidu, Ali's? There is a "data conspiracy theory": big business and government agencies are now doing things with big data, and if we are not mastered by ordinary people, they are controlled by an invisible net. The role of the "black box" problem of large data processing from the angle of economic security. There are also practical ways to talk about the use of big data in business. But what touches me most is the following two points. Here is a simple excerpt to chiselled readers.




The big data that
don't pay attention to human nature is big flicker--Liu De





's current mainstream understanding of big data is based on Victor's two-time improvement in the big data age. But there are two very questionable views, one is the extreme contempt for sampling, and the other is the absence of principled esteem. Large data is a sort of abandonment of random analysis (sampling) to deal with all the data, then there is a question posed by Stanford Trevor Hastie, how to find a needle in the straw, if many straw is as long as the needle. This is the biggest risk to all of our big data research, and the fact that the data is too big is actually a pattern of loss and distortion, and never ignore the sampling.





To put aside these two ideas, it is even more frightening that the big data today is rarely about human nature. First of all in life, we have encountered problems, a person went online to buy 5 liters of laundry fluid, the entire process cost less than 1 minutes. The next day browsing the web, he found that there was a wide variety of laundry detergent. What is it? Precision marketing based on large data? I'm afraid it's based on the precision of the big bluff. A bit of common sense people know that 5 liters of laundry fluid even if the family more than one months, and that person's process is so short, it must be brand loyalty, the promotion should be when the brand of laundry liquid will be discounted and so on, this is the big data. The method of early human research is to rely on experience, perception, experience, introspection, and so on, something that seems to be irrelevant to large data is just the core of large data, because it is thought.





Google 2008 made a very powerful thing called the flu trend forecast, it predicted the results than the U.S. Centers for Disease Control, which was a sensation in the world. As a result, things became more and more messy, seriously overestimating the state of influenza. Why? This is what the victor genre is talking about when it comes to big data. Influenza and the onset of time, and the United States, such as middle school basketball match that time point is exactly the same, these two concepts can be related? The problem is that as long as the middle school basketball game is searched, it constitutes one of the main words of influenza prediction. There are so many things like that. Because at Google's prediction, there is no expert on disease control and public health, these things are a very important prerequisite for large data prediction.





Genetic engineering is the real big data




How many cells do
people have? The magnitude is 10 to 14. One of the cells that is cancerous can lead to the end of your life. Isn't that big data? The real big data is life-wide data, genetic screening can eliminate congenital diseases and prevent cancer, and it is not difficult for humans to revive after millennia. However, the development of genetic technology has encountered countless practical bottlenecks and ethical challenges.





question 1: From a civilized and religious point of view, the "perfect Man" created by genetic engineering is another species, is this "perfect man" still human?





Question 2: Genetic problems and large data problems the biggest harm is the violent destruction of personality independence and privacy.


   


    In response to these two issues, Wang Jian, director of the Hua Da Genomics Institute, gave his interpretation:     you don't do it, the Europeans are doing it, the Americans are doing it, in one of the most pessimistic terms, rather than letting the white man kill us and kill ourselves. When I was in Seattle in 1993, the old Bush era launched the Human Genome Project, and that was a very intense discussion. Genetic science will make a great difference to the current human morality, law, and the way we live and produce medical care, which may be difficult to change with human will, which is a little too fast in this time dimension and we feel too fast ourselves. When I was in the academy, it was because of these things, so we left. After leaving, the result is quicker, from dozens of years time to thousands of people, next year may be tens of thousands of people, the next year, pure basic research institutions of tens of thousands, may be the country's current research institutions of the destructive and challenging is very large, industry development will be very large. But it is in some ways responsive to the needs of the people.