If data is the new oil… who are the new oilmen?
“Data is the new oil” is a phrase that is often repeated these days. It was coined in 2006 by Clive Robert Humby, a British mathematician and entrepreneur, expert in what we now call data science and which only a few years ago we would have called statistics. In fact, for many decades, statistics have been the science of data, or the mathematical discipline dedicated to data analysis, whatever their nature. From census data, for the analysis of which IBM was founded in 1911, to economic data, which is why statistics is considered the mathematics of economists.
For some years now, however, a new trend has emerged that tends to associate data analysis with Artificial Intelligence (AI).
AI has made significant progress only in recent years by introducing computational techniques based on a relatively new concept: the artificial neural network. The idea behind this new concept is relatively simple to explain. An artificial neural network (made on the inspiration of how the brain is structured) is a set of connected nodes. Each node can be in an on or off state in relation to what the nodes it is connected to are doing. In a network there are input nodes and output nodes and, depending on how the switching on or off rules are established and how much the individual connections weigh, calculation algorithms can be coded that provide the result in the output nodes. Using this type of schemes, it is possible to implement the computing technologies known as deep learning, at the basis of current AI. They have been efficiently applied as classifier schemes, and successfully used for image recognition and speech analysis.
These powerful techniques, in order to work, need data a lot of data. For example, to “train” an image recognition algorithm containing a cat, we need many (tens or hundreds of thousands) of images picturing a cat. These images are used to create the artificial neural network that will then have to implement the deep learning algorithm. Similarly, in other sectors, it is necessary to have a lot of data to create the algorithms that will be used in practical applications (hence the expression “Big Data”).
A journalistic case that has brought international relevance to these analysis techniques is the one linked to the scandal of the English company “Cambridge Analytica” which in 2018 was the focus of a judicial investigation for using personal data acquired by Facebook in order to build a profiling of voters at political events such as Brexit and the US presidential elections. Since then, many have realized that data has the potential to be of great value even though, as Michael Palmer of the Association of National Advertisers noted, “Data is just like crude. It’s valuable, but if unrefined it cannot really be used”.
The new oilmen
Just as the old oilmen are those who extract crude oil from the ground and then refine it, the new oilmen are those who are capable of capturing data and analyzing them to obtain useful predictions.
After the experience of Cambridge Analytica, the owners of the oil fields (Facebook, Amazon, Google, …) followed the example of some Arab countries: they sent away foreign oilmen and began to extract and refine their data themselves. One of the most interesting applications of this new policy is the so-called “predictive marketing”. We experience this technology whenever we receive advice via email or on the web about the purchase of a boat (after having visited sites and conversations related to boats in the previous days) or a special sunburn cream (after having visited sites on a dermatological and beach themes).
In all this we must not forget a fundamental aspect: who produces the drills? In fact, oil cannot be extracted without drilling and those who produce the extraction technologies have a considerable advantage over all aspiring oilmen. Apple, Samsung and the like have so far been the producers of modern drilling rigs for data acquisition and in fact they too have begun to grow oil ambitions (think of the subtle invasiveness of Siri or Alexa).
Smart sensors are the new data drills
In the future, scenarios of great interest are opening up for all those who will be able to produce new type of data drillsthat go beyond the smartphone. It is the vast prairie called IoT (Internet of Things) where it will no longer be just humans who will produce data, but mostly things. Smart sensors are the new data drills. Think of the vibrational sensors inserted in the infrastructures of a motorway viaduct capable of measuring the structural health of the bridge and measuring the traffic load, just to give an example that is already a reality.
The black gold rush has already begun, those who want to take part are warned: we are in the new far west, saddle your horses and design your sensors.
