Small data for big solutions

Analysts predict that 2022 will bring advances in artificial intelligence that will make it possible to analyse information in less time, on the move and with less energy.

Josep Lluís Micó

La Vanguardia

October 11, 2021

https://www.lavanguardia.com/tecnologia/20211011/7778713/datos-pequenos-grandes-soluciones.html

The development of data science has driven fourth industrial revolution technologies such as natural language processing, computer vision and deep learning. Without this discipline, artificial intelligence, a field that is transforming the way we work, play and interact, would be less intelligent and more artificial. Well, experts such as best-selling author Bernard Marr predict that by 2022 there will be remarkable advances in this area.

Analysts recall that as the importance of understanding data and knowing how to operate with this information grows, the science on which it rests becomes more accessible. Just ten years ago, this field was seen as the intersection between mathematics, statistics and computer science, and only a handful of universities had dared to offer specific studies. Today, its relevance in business, commerce, etc. is undisputed and courses and training paths for further learning abound.

Reasons such as these justify talk of the "democratisation" of data science, a trend that will be accentuated in the coming years. In addition to big data, the concept of "small data" is already being used as a paradigm to facilitate the rapid cognitive study of the most salient material in situations where time, available bandwidth or energy consumption are vital. These algorithms are designed to run on low-power hardware, close to where the action is.

Thanks to this solution, the performance of household appliances, cars, agricultural machinery or industrial equipment, among others, is improving, because "small data" makes them more efficient and useful. Likewise, when companies and institutions used the information of their consumers or users to provide them with more and more pleasant content and services, they began to talk about the "data-driven customer experience".

The interactions of citizens around the world are increasingly digital, an extreme that has been amply demonstrated during the coronavirus pandemic and the various confinements or physical distances. Organisations already measure various aspects of the connections and engagements that these people manifest. From chatbots to Amazon's cashierless convenience stores, they do this on a daily basis.

On a larger scale, so-called "generative artificial intelligence", naturally integrated into sectors such as entertainment, is successfully achieving its goal: to generate a reality with data... that does not exist as such. This is how American director Martin Scorsese managed to get actor Robert DeNiro to appear in the film The Irishman at a younger age than he is today. Engineers will try to train image recognition systems to detect signs of rare and rarely photographed cancers in medical records.