Data and AI: the age of digital enlightenment?

Collection, storage, valorisation … On the occasion of its event at the EPFL on February 6, 2020, Softcom convened five experts to discuss the issues related to data and artificial intelligence.

Each decade has its own successful keyword.

Since the advent of Big Data at the turn of the decade, the digital economy has been dominated by mega data from social networks, connected objects, mobile phones and sensors. For her, this treasure was the promise of innovation, but above all of new markets to conquer. The lines shifted a little in 2020 with the advent of AI. Artificial intelligence is on everyone’s lips today. This technology, fuelled by the ever-increasing growth of digital data, promises to revolutionize our daily lives as connected men and women, economically, ethically, politically and societally.

Artificial intelligence and the data that is used as fuel at the heart of the digital issues, because they raise crucial questions even in their own definition: what is AI, big data? What are the purposes of these technologies? How do we make sense of raw data? What about the integrity of this information? What will our future be and what place will humans be able to claim? We could continue the list on several pages. It’s to answer these and many other questions, that Softcom invited five leading experts on 6 February 2020 to its event “Hey Siri, what will be the Switzerland of tomorrow”  at EPFL.

Mastering what we’re talking about

At the EPFL, Rémi Sabonnadière is an expert on data in all its forms. First for studying it after a degree in data science. But above all because he is now one of the members of the management team of the EPFL Extension School. This online continuous training accompanies companies in their digital transition. “The Extension School was created with the stated objective of enabling all job profiles to acquire the skills needed to succeed in a changing world of work. Companies come to us because they want to make a successful digital transition.”

To achieve this, they need to train their employees in these technologies, which are far more accessible than many people think and bring value at all levels. But before we get there, we still need to understand how AI and data work: “For many people, artificial intelligence is used for prediction,” emphasizes Rémi Sabonnadière. That is, this technology can explain the present and the future through past data.” But things are changing in the face of their exponential growth. Because 90% of the data produced by humanity has been produced in the last two years. An ever-increasing volume and variety of data raises new questions for data scientists: how to deal with this volume, how to navigate through the variety of information, how to cope with the speed at which new data is produced, and finally, are there technologies that can judge the veracity of this information?

The issues of storage

According to Liviana Zürcher, data scientist at Microsoft, artificial intelligence is a continent to be apprehended. She invites companies to take the first step in this unknown territory but  synonymous with promising. Only in this way will we be able to “unleash the power of data”. We still need to store and protect it. Microsoft is following the trend by offering home storage solutions. With its Azure service, the technology giant from Redmond has just inaugurated two data centres in Switzerland. One in Geneva. The other in Zurich. It offers its cloud services to businesses, public administrations and NGOs. The objective of these openings is to offer Swiss companies and administrations the possibility of keeping their critical data in Switzerland while taking advantage of the replication systems between the two datacenters in order to ensure the continuity of services in the event of an incident.

If data is a raw material, how can it be meaningful ? The young Fribourg company Swiss-SDI does not confuse data science with big data. Founded recently by three former EPFL mathematics students, Swiss-SDI prefers to draw on artificial intelligence for useful results rather than amassing huge amounts of data. In other words, the company has its feet  on the ground and offers concrete solutions adapted to the reality and strengths of Swiss companies. According to Xavier Bays, one of the founders, flexibility and tailor-made service, in close cooperation with the customer, are more important than the complexity of the analysis processes. Xavier Bays stresses to his customers on the need to remain pragmatic, as much on the goals as on the problems and the tools.

A profound cultural change

The co-founder of Swiss-SDI rightly points out that data science and data engineering are two different but complementary professions. The person who collects, structures and analyzes does not exist. Xavier Bays adds the importance of flexibility. A range of tools and skills exist. The essential thing is to know what you are looking for, for which purpose, with what means and for what purpose. If the company is not clear on these points, it will not take advantage of the potential offered by AI. In other words, artificial intelligence is not just a technological evolution. This is a real cultural change. A point that Blaise Lugeon, Enterprise Architect at Softcom Technologies, insists on.

According to the expert, just because the technology exists does not mean that the challenges are solved. He quotes Gartner’s latest study that shows that 82% of Big Data projects fail and that 62% of these failures are not due to technology, but to inadequate business skills. So what are the best practice ? Blaise Lugeon invites companies to think about data from end to end, from production to sharing, storage, anonymization and control. This culture change does not happen by itself. It requires a connection with all the players who are active in some way or another in the field of data science. It is the sharing of data, the sharing of knowledge and concrete cases that will allow it to move forward and enhance the value of his work in this new territory.

Trust, the mother of all battles

Finally, there is no growth without trust, said François Hollande in 2012. In the face of the growing robolution of society, can we – as individuals – trust the companies and services to which we delegate a growing part of our daily lives as connected men and women? Digital confidence is at the heart of the digital transition. Philippe Gillet  understood this. The head of security at Sicpa reminds the importance of not forgetting that behind dematerialized data, there are real impacts. Whether in the agri-food, health, environmental, financial or political sectors, the use of data and artificial intelligence has real effects. Hence the importance of opting for good practices upstream in order to avoid bias. A requirement to ensure that the digital transition does not shoot itself in the foot.