Seth Stephen Davidowitz is a philosopher and a doctor in Economa. In addition to having worked as a data scientist at Google, I wrote the best seller Everyone Lies, about what the Internet and mass data (
big data, in English) can you tell us about ourselves.
During his visit to the country to participate as a guest of the Federation of University Teachers (Fedun) in the XIII International Congress of Economics and Management ECON 2019, of the Faculty of Economic Sciences, this 37-year-old professional spoke with the nation about How we use social networks.
-You predicted the triumph of Donald Trump after analyzing Google searches for racist terms and other xenfobas words. The key to his study was to understand that in private people reveal much of their true interests and motivations, while they hardly do so in public.
-It is a universal theme. Platforms like PornHub show that we all watch pornography, although nobody admits it in public.
– That means that the traditional surveys are becoming obsolete?
-The surveys show erroneous results for several reasons. Among them, that very few people are answering them, and on top of that, they lie. For example, in the survey they express that they are going to vote for one candidate, but then cast their vote in favor of another. Another contradiction: if you are asked what they care about, the usual thing is to listen to the economy, global warming and work, but in Google one of the most popular searches is "Alberto Fernndez's son is drag queen". So that is what people are interested in.
-Or maybe Fernndez's son is curious. Curiosity and genuine interest are not synonymous.
-It is true. It may be that the search about Alberto Fernndez's son is related to a momentary curiosity, but it is also possible that it is something deeper than that.
– Did you have the opportunity to analyze the behavior of Argentines in social networks?
-Yes, in this country many people talk openly about issues that are private, but they may also be lying when they express themselves.
-If your main hypothesis, that we all lie, is true, the companies that work based on that data to do personalized marketing will not achieve successful results.
-Yes, that is a big problem for companies, because as social networks are aspirational, consumers pretend they believe in their products, but not really. My recommendation for brands is to analyze what happens in social networks, but also do it in search engines, which are a kind of virtual confessional. For example, we all know that sex moves the world, but when browsing pornography sites everyone has their preferences. That is why brands are wrong to always choose the same physical stereotypes, because this does not appeal to all people.
-Following your thinking I performed a very impressive exercise. In my social networks how my friends talk about their husbands. In all cases they are "My husband is the best in the world", "he is the love of my life", and so on. Then I put in Google "My husband is." and the five main ways to complete that sentence are: "an idiot", "the best in the world", "a lazy", "boring" and "a useless". Based on this, what advice can you give readers who read social media?
-The danger is when users read social networks and this leads them to feel bad about their own lives. Because, clearly, all couples argue. In short, my advice is to remember that everyone lies on the networks, and that nobody has such a wonderful life. In fact, we all have similar problems.
-You affirm that the algorithms know us better than we know ourselves. How should we act taking this into account?
-We should regulate companies like Facebook and Google because they can deceive us. In addition, as artificial intelligence (AI) becomes more powerful, companies can gain more advantages from us. And these technologies, such as AI, should also be regulated. The same thing that was done with cigarette manufacturers must be done with them. In this sense, the amount of nicotine that each cigarette can contain is stipulated and where they can spread their advertising. Today, as there is no regulation, all of us are laboratory rats, and that is evident in the A / B tests, which are tests that consist of spreading two versions of the same content to measure what works best.
– Should we stop using social networks?
-You can use them, but with caution, knowing that everyone takes advantage of them to show off.
. "Everyone lies (t) and you have to take with tweezers what you read on social networks" – LA NACION