Big Data Cooking | The economist Walter Sosa …

We have, as never before, a lot of information available and it excites us but also makes us anxious. On the one hand, the future presents itself as a true paradise; what was library hours, today is solved in seconds. However, so much technological power has its flip side: we have access to a lot of data but in return we leave ours. When we put an address on the GPS, send an email or have dinner in a bar with friends, we print our footprint in cyberspace. We are geolocated even if no one is looking for us and we are predictable even if few people appreciate our privacy. The world era is like that, it oscillates between comfort and panic, at any time and whatever the place. In the interstices, we know that a great cultural transformation is woven, but we are not yet able to express it. The good thing is that we are not alone, we have riders who guard the earth, use their long-term vision and help us understand things to their measure and in advance. Walter Sosa Escudero is an economist (UBA), professor at San Andrés University and researcher at Conicet. He has just published "Big Data" by XXI Century Publishing House and puts cold clothing in a situation that seems to run away from fever.

– What is Big Data?

– The first, how fast, Wikipedia, would be to define it as a phenomenon of profusion of mbadive and instantaneous data; produced spontaneously and with anarchic expression through interconnected systems. Nowadays, we often think that the more data we communicate, the better it is not always the case. You can not compare the data that gives 1000 well-balanced surveys, or the 60 that are manipulated in a thorough laboratory experiment, with the 300,000 that result from the response to a multiplicity of choices on Twitter. In order not to celebrate Big Data as a revolution, I like to say that the important thing is not to work with a lot of data. but to be healed. They will certainly be much more informative than those extracted without any criteria.

– How to use as much information in favor?

– Talking about Big Data, it's a bit like talking about craft beer. The methods used to study them were the inverse of the data. When the leg of data-methods-ideas is removed from one leg, it collapses. Having more information is good news as long as there are questions that call for the need for an answer. In the future, the necessary condition for a data revolution will depend primarily on people's ability to generate questions and debate ideas.

– Are there any questions that claim to have an answer?

– There is but not in the proportion that we would have liked. In the last 15 years, there has been an explosion of data and methods, but not ideas. It is harder than ever to predict who will win a World Cup or who will win the next election. In the age of algorithms and big data, we do not make better sports or political forecasts. Just look at what happened with the emergence of Donald Trump, the course of Brexit and, without going so far, the appearance of the Fernández-Fernández formula. With the automatic learning involved, no one saw the piece. Nobody What we do not notice is that phenomena as complex as economics or politics present an intrinsic difficulty that has nothing to do with the lack or excess of data.

– Although data can not predict large phenomena, they are useful for inducing decisions. In fact, the advertising market is structured this way.

–Absolutely. If you ask me what is the biggest income of all this, I would say that it resides in the inductive, descriptive, ranking and clbadification of Big Data. In the past, there was talk of "the elector" and today we can work with much greater granularity: "the elector who has between many years, who lives in the neighborhood x with his partner and attending university x ". The same thing happens in advertising, there was something McDonald's diversified and there were McCafés; Obviously, someone with a good eye realized that the boys had been brought by parents who preferred to drink coffee after the movies and not eat combo.

– Now that we know its potential, what is Big Data for?

– This is not used to explain, because it is much more difficult. Many people think that because it can be better ranked, we can better explain. A rough example: if I can get the sales data of umbrellas in different regions, I could estimate the geographical points where it rains the most. If they sell more in Rosario than in Antofagasta (Chile), it is necessary that the precipitation be greater in the first case than in the second case. However, this does not explain why it is raining, nor does it help to solve a drought or flood. I can not advise those who exchange umbrellas or those who want to cultivate the land. Since they do not allow me to understand the reasons, then I can not change the reality. There is a huge gap between measuring and understanding poverty, designing a policy and badessing its positive impact. The data could be a tool to win an election, but it is not final because it requires interpretation and creation by badysts.

– With the exposure of our data on social networks, we live more and more under control. Have we lost freedom, are there new forms of freedom, or has freedom ever existed?

–In order not to sow traces, every time I go to work, I have to leave my mobile phone at home. But that complicates my life; I can not control the traffic, I do not know if it is convenient for me to travel by public transport, use the bike or walk. I do not have weather data either and I do not know how to dress. By accessing the services provided by the Internet, I take advantage of many things and someone makes me pay. As with any technology, the problem is to account for the advantages and disadvantages. The obstacle with Big Data and the predictive ability of algorithms is that today, this balance does not give a precise result. The limit with individual privacy is kept in muddy ground. I would not like to discover a robot with things like this: "Today you are drinking coffee with a friend and he is very likely to have cancer." The system can know a lot about us, it can even know things we do not even know about ourselves.

– The question requires a mathematical and statistical approach, but it also has philosophical aspects …

– Yes, because it leads us to think about the problems that we thought were forgotten. Big Data is halfway between revolution and becomes more similar. Similar to social networks: a new mode of interaction is proposed, but that does not mean that we have not dialogued before. The neighborhood girl who beckoned to you when she was cycling today is "I like her". There were always social validation mechanisms; I prefer to believe that links are woven differently. Today, we talk about news bubbles as something new, that is to say love to those who think like me and suppress those who do it differently. Irigoyen's diary was a bit like that. Morality: be careful in the communication of revolutions, especially when they do not happen yet.

[email protected]