Take a photo of a barcode or cover
660 reviews for:
Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are
Steven Pinker, Seth Stephens-Davidowitz
3.82 AVERAGE
660 reviews for:
Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are
Steven Pinker, Seth Stephens-Davidowitz
3.82 AVERAGE
Mostly good. The shock & awe factors seemed a little forced and stretched on a little too long. Really enjoyed the author's discussion of the tools he used and why.
While the book has some interesting tidbits, the author makes assumption upon assumption about the results of his data based on what he expects to see. One example would be that he uses Google searches for the word N****r to demonstrate continuing racism and to estimate how prevalent it is, however he explains that the searchers are obviously not looking for hip hop lyrics because if they were they would use the term N***a. The vast majority of people that I have ever encountered do not make any distinction between the two, it is a matter of pronunciation. If I were looking up the lyrics to the Kanye West song Golddigger and could only remember the lyrics "broke n***a" I would search with it with an "er", not an "a", because that is the word that I hear him saying and I am using the most common spelling. In fact, I can type that into Google as a test and, sure enough, it does not correct my spelling or ask "Did you mean n***a?" It just gives me tons of results.
Another thing that this illustrates is why people may type such things into search engines. I did it to test to see if the N word would give the results that the author assumes I would not be looking for, but now I am in his pool of racists. As another example, the author repeatedly references hate sites on the internet, and on many occasions he talks about the Stormfront.org website, a gathering place for white supremacists. He uses the visitation data and registration data to again demonstrate that racism is more rampant than we think, and that Google can shine a light on how big the problem is. Again, what he fails to take into account are people who are not racist but who visit the site for research, like the author and myself to see what he was talking about (again, I am contributing to his racist numbers), people who have heard about the organization and are just curious to see what these horrible people are saying and witness these train wrecks of human beings, a phenomenon that increases whenever groups such as this are in the news, and people who go there to argue and troll the white supremacists and call them names. I have been to tons and tons of websites that hold views that I would never, ever consider myself, and I know that I am not unique. How many of the search engine queries and hate site visits result from people like me? There is no way to answer that. It could be a few, it could be a huge number, but there is no way to tell who is doing these things or why, which makes his conclusions sketchy at best. The author never acknowledges this, instead he uses words like "obviously" and "has to be" in order to demonstrate the soundness of his conclusions, while the conclusions are anything but obvious to anyone that does not already start with his premise. At one point he even tacks on to the end of a paragraph about how his data could not predict repeatedly how well stocks would do based on search data a conjecture that his failure might be because investment firms already figured out methods like his and are skewing the data. That he even felt the need to add that shows that in his head he is rationalizing why data that "should" work did not work.
One particularly egregious area was where he estimated that Barack Obama lost about 4% of the vote due to racism. He is not saying that 4% of the voters were racist, I would completely agree that a number in the mid single digits for racists sounds about right, he is saying that people would have voted for John Kerry against Bush, but did not vote for Obama against Bush because they were racists. He reaches this conclusion because he found that, other than race, there is no discernible difference between the two candidates. The ridiculousness of this statement is profound in that he seems to assume that he knows all of the factors that go into making a decision about who to vote for. Obama is described as a liberal and Kerry is described as a liberal, so people that would vote for one would vote for the other unless they happen to look different, and specifically if their skin is a different color. I did not vote for either of them, and actually wrote in candidates for the last three elections, but if someone put a gun to my head and asked me to vote for one of the two I would have voted for Obama because he came across as more honest, and I always believed that his heart was in the right place even if I disagreed with his views. I could certainly not same that I got the same impression from John Kerry. Other people liked that Kerry was a war veteran and that he spoke out against Vietnam to Congress, and many people liked that Kerry had a long political track record. Neither of those things are things that Barack Obama could have claimed in his first election, but neither of those are considered as significant differences by the author, and certainly not as significant as skin color. Military service and political experience are considered very important to many people, and while military service is generally considered to be more of a benefit to Republicans than to Democrats, Kerrys record of testifying about Vietnam to Congress endeared him to many Democrats who saw him as an insider who was daring to speak the truth. How much of an effect did this have on people that would have voted for Kerry but did not vote for Obama? We have no idea, but that does not mean that we can draw the definite conclusion, as the author does, that racism cost Obama votes from potential Kerry voters.
The data in this book is interesting, and over time there may be information below the surface of internet searches that can be gleaned, but this author makes so many unsubstantiated leaps in this book that his conclusions cannot be taken seriously.
If someone is looking for a clear cut case of confirmation bias, or if you are interested in seeing what data can be gathered and what some people are using it for, then you may really enjoy this book. Others that will enjoy this book are people that will read his conclusions and think "Aha! I knew it!" because they will be making the same mistake of looking at the data with their minds made up as well. Otherwise I would not bother, since in the end the book does nothing to shed light on anything new.
Another thing that this illustrates is why people may type such things into search engines. I did it to test to see if the N word would give the results that the author assumes I would not be looking for, but now I am in his pool of racists. As another example, the author repeatedly references hate sites on the internet, and on many occasions he talks about the Stormfront.org website, a gathering place for white supremacists. He uses the visitation data and registration data to again demonstrate that racism is more rampant than we think, and that Google can shine a light on how big the problem is. Again, what he fails to take into account are people who are not racist but who visit the site for research, like the author and myself to see what he was talking about (again, I am contributing to his racist numbers), people who have heard about the organization and are just curious to see what these horrible people are saying and witness these train wrecks of human beings, a phenomenon that increases whenever groups such as this are in the news, and people who go there to argue and troll the white supremacists and call them names. I have been to tons and tons of websites that hold views that I would never, ever consider myself, and I know that I am not unique. How many of the search engine queries and hate site visits result from people like me? There is no way to answer that. It could be a few, it could be a huge number, but there is no way to tell who is doing these things or why, which makes his conclusions sketchy at best. The author never acknowledges this, instead he uses words like "obviously" and "has to be" in order to demonstrate the soundness of his conclusions, while the conclusions are anything but obvious to anyone that does not already start with his premise. At one point he even tacks on to the end of a paragraph about how his data could not predict repeatedly how well stocks would do based on search data a conjecture that his failure might be because investment firms already figured out methods like his and are skewing the data. That he even felt the need to add that shows that in his head he is rationalizing why data that "should" work did not work.
One particularly egregious area was where he estimated that Barack Obama lost about 4% of the vote due to racism. He is not saying that 4% of the voters were racist, I would completely agree that a number in the mid single digits for racists sounds about right, he is saying that people would have voted for John Kerry against Bush, but did not vote for Obama against Bush because they were racists. He reaches this conclusion because he found that, other than race, there is no discernible difference between the two candidates. The ridiculousness of this statement is profound in that he seems to assume that he knows all of the factors that go into making a decision about who to vote for. Obama is described as a liberal and Kerry is described as a liberal, so people that would vote for one would vote for the other unless they happen to look different, and specifically if their skin is a different color. I did not vote for either of them, and actually wrote in candidates for the last three elections, but if someone put a gun to my head and asked me to vote for one of the two I would have voted for Obama because he came across as more honest, and I always believed that his heart was in the right place even if I disagreed with his views. I could certainly not same that I got the same impression from John Kerry. Other people liked that Kerry was a war veteran and that he spoke out against Vietnam to Congress, and many people liked that Kerry had a long political track record. Neither of those things are things that Barack Obama could have claimed in his first election, but neither of those are considered as significant differences by the author, and certainly not as significant as skin color. Military service and political experience are considered very important to many people, and while military service is generally considered to be more of a benefit to Republicans than to Democrats, Kerrys record of testifying about Vietnam to Congress endeared him to many Democrats who saw him as an insider who was daring to speak the truth. How much of an effect did this have on people that would have voted for Kerry but did not vote for Obama? We have no idea, but that does not mean that we can draw the definite conclusion, as the author does, that racism cost Obama votes from potential Kerry voters.
The data in this book is interesting, and over time there may be information below the surface of internet searches that can be gleaned, but this author makes so many unsubstantiated leaps in this book that his conclusions cannot be taken seriously.
If someone is looking for a clear cut case of confirmation bias, or if you are interested in seeing what data can be gathered and what some people are using it for, then you may really enjoy this book. Others that will enjoy this book are people that will read his conclusions and think "Aha! I knew it!" because they will be making the same mistake of looking at the data with their minds made up as well. Otherwise I would not bother, since in the end the book does nothing to shed light on anything new.
informative
medium-paced
Remarkable in its ability to evoke freakonomics and inability to be as compelling. The author has interesting ideas but needs better editing or direction. The conclusion stands out for two reasons - that the end of the book is finally here, and for the amount of times the author mentions how important a good conclusion is, how he hopes to create a good one, how he is planning to spend 6 months crafting it ... all without generating much in the way of an actual conclusion.
The best idea in the book is that there are ethical concerns about using big data. Shame it’s not mentioned till 3/4 of the way through, and that it’s not further examined.
The best idea in the book is that there are ethical concerns about using big data. Shame it’s not mentioned till 3/4 of the way through, and that it’s not further examined.
Full with rather fascinating, useful and thought-provoking revelations - both arguments and tidbits - but could use less sports examples and more non-US examples. A short and quite enjoyable read nonetheless.
کتاب واقعاً خوبی بود. همه میدونیم کلی داده در سطح اینترنت هر ثانیه داره ایجاد و جابجا میشه، ولی اینکه بدونیم با چه شیوهای روی این دادهها تحقیق کنیم و بررسیشون کنیم چیزیه که این کتاب یکم بهم نشون داد. تخصصی نیست کتاب و خیلی در سطح عام توضیح داده مسائل رو. ترغیبم کرد به اینکه به بررسی کلاندادهها لاقل فکر کنم.
ترجمهش هم به نظرم روان و خوب بود.
ترجمهش هم به نظرم روان و خوب بود.
So good! I love the push for social science as a real science, and the ways in which big data and our internet usage can be used to improve our lives.
informative
reflective
medium-paced
funny
informative
inspiring
lighthearted
reflective
medium-paced