What exactly is a scientific exhibition advisor? A discussion with Stephan Clémençon

Who are the people working behind the scenes at scientific exhibitions? The tasks at hand range from approving content and proposing themes to identifying scientific and societal issues, and much more. To find out more, we interviewed Stephan Clémençon, a researcher specializing in machine learning at Télécom ParisTech and a scientific advisor for the Terra Data exhibition led by Cité des Sciences et de l’Industrie, focusing on digital data.

 

What is the role of a scientific exhibition council?

Stephan Clémençon: Organization for the exhibition began about a year and a half before the event. Our council was complementary in terms of skills, since it was made up of mainly technical specialists in IT data, etc., as well as others who focused on usage and legal issues. Our aim was to identify the topics to be addressed during the exhibition by illustrating with examples. Above all, we wanted to make the link between data and applications. Secondly, the exhibition organizers presented the different workshops to us and what they did was extraordinary.

 

What messages were you wanting to pass on?

SC: We wanted to show that data are not just a way of representing information. For example, we addressed the notion of storage. Often, people don’t realize the network aspect, the fact that there are kilometers of fiber optic cables at the bottom of the oceans. It’s important to show people pictures of that. In practice, people switch on their computer, search for information, etc., but they actually have no idea about the physical and concrete aspect behind this, such as what a data center looks like. The important thing was to demystify data.

 

Which part of the exhibition did you work on the most?

SC: I mentioned to the organizers that biometrics could have an impact on the public. The idea was to follow the digital trace of visitors, who had their photo taken at the exhibition entrance. I also worked on the Algorithmic aspect with Françoise Soulié-Fogelman [a professor in computing at Tianjin University, China]. We illustrated how recommendation engines work and what their principles are. The objective was to demystify the algorithmic aspect. An algorithm is simply a sequence of tasks leading to a result. We explained that it was nothing new and that they are already in use in daily life.

 

What motivated you to participate in this project?

SC: I think that addressing the subject of data is important. People are scared to talk about artificial intelligence, automatic processing by machines, etc. and rightly so, because we are becoming dependent on these technologies. But this is not something specific to machine learning, it applies to technology as a whole. It is therefore very important to explain how these technologies work. Working with Cité des Sciences also allowed us to reach young people who use technology but don’t necessarily ask themselves how it works. I also took part just out of curiosity. I had no idea about how these exhibitions were put together and it allowed me to discover this new world.

 

What can you draw from the exhibition and all that it entailed?

SC: I feel that it was a small but well thought-out exhibition. It was educational on the topic of data and what they are used for. There was a good balance between mathematical aspects and usage. It would be interesting to generalize this kind of exhibition because there is a real need to provide society with information about digital technology. In terms of artificial intelligence which has lately become fashionable and developing robotization, many of these issues are suffering because of received ideas. They deserve to be presented to the general public in the form of an exhibition. We are currently sitting on the threshold of some really significant transformations and it would be good if people started to think about these rather than just letting them happen.

 

Terra Data: the exhibition demystifying data

 

Leave a Reply

Your email address will not be published. Required fields are marked *