Gaëtan Robillard 1/4 : Data-portrait of climate scepticism
Data, sets of data: it's practically all we talk about these days. They are the prism through which we view reality, they inform and motivate decisions. We mine them like coal, we extract them, we exploit them, we feed them to our machines...
It was only a matter of time before an artist used it as raw material for a creation. Here we are - following a group from Stuttgart which developed an algorithmic art (an aesthetic which Gaëtan Robillard studies as part of his university research), the Frenchman Gaëtan Robillard is a "data sculptor". His work as a visual artist consists of making "perceptible" (i.e. materially tangible to our senses: sight, hearing, touch, etc.) these data that connect our societies today.
It is often said that art is a way for the artist to share his vision of reality: here, it is the data itself that is diverted to show this vision, this world from a different angle.Gaëtan Robillard, originally from Brittany and is fascinated by the ocean, has, for example, used a meteorological simulator developed by a team of researchers in Brest to “materialise” wave heights during a storm. A long slab is laid on the ground and parallel beams, set up at regular intervals, display the evolution of these heights on an equivalent area of ocean. This makes the intensification of storms due to climate change immediately understandable.
Gaëtan Robillard. Laëtitia Ngha, Lycée International de l'Est Parisien et Anastasiya Balan, Université Gustave Eiffel. Critical Climate Machine, 2021.
When the European network Mediafutures (of which IRCAM is a partner) sought, at the end of 2020, to launch and support artistic projects that highlight the major role taken by data in our experience of reality, the terms of the call seemed tailor-made for Gaëtan Robillard. This is how a new piece, entitled Critical Climate Machine, was created. 
For the occasion, he used the system imagined for the materialisation of storms, and of course, modified it. He began by feeding it with other data, to address, once again, the crucial subject of climate change. However, rather than focusing on its effects, and in order to better understand the sometimes-harmful influence of data in our societies, he turned this time to the climatosceptic discussions spread through social networks.However, there is no longer any question of modelling here: the installation gives a sense of the degree of online disinformation in real-time. To do this, Gaëtan Robillard and his team *, assisted by cognitive science researchers **, trained an artificial intelligence to classify the various climate sceptic speeches. From pure lies to the misappropriation of quotes, including quotes out of context or the questioning of scientific consensus, this AI - which resembles those developed by the social networks themselves to hunt down fake news - assigns a code (like an "error code") to all the tweets from certain Twitter accounts identified as belonging to the climatosceptic movement, in real time.
The digital translation of the tweet (in an AI-readable form), as well as its error code, are then displayed on the device, which is also capable of heating up - the heating temperature depending of course on the quantity of tweets analysed and detected by each of the units in the sculpture.
To complete the installation sonically, Gaëtan Robillard draws on the research and creation work of IRCAM, starting with that of Jérôme Nika. With the help of computer music designer Dionysios Papanikolaou, he uses the OM-DYCI2 machine (a generative engine allowing to generate many variations from the same compositional idea) to generate new phrases from a database of specially designed and played dialogues containing various climate sceptics and their refutations.
© Gaëtan Robillard
“Playing on the opposition between order and disorder, the new phrases deconstruct the original meaning more than totally losing it," says Gaëtan Robillard.
On the one hand, they are the result of a “disordering” of recorded speech and, on the other, they reconstruct units of meaning. In other words, OM-DYCI2 makes it possible to generate purely musical phrases, but also to provoke accidents or shifts in meaning that make the nature of language on the climate appear in another way.
The contrast between recorded and generated voices creates a tension that echoes the ethics and dynamics of the relationship between society and technology.
Their accumulation also illustrates the saturation of the information space - which unfortunately allows the climate sceptics to cloud the issue.
© Gaëtan Robillard
* Gaëtan Robillard's team for the development of this project:
Jérôme Nika, researcher
Tony Houziaux, sound designer
Dionysios Papanikolaou, computer music designer
Özlem Sulak, artist advisor
Vincent Nozick, deep learning expert
Laurine Capdeville, engineer designer
Jolan Goulin, engineer programmer
** John Cook, Monash Climate Change Communication Research Hub, Monash University, Australia.
Constantine Boussalis, Department of Political Science, Trinity College Dublin, Ireland.
Travis G. Coan, Department of Politics and the Exeter Q-Step Centre, University of Exeter, United Kingdom.
Part of the installation was developed in July at the ZKM in Karlsruhe.
A first version is currently on show at the Deutsches Museum in Nuremberg (October-November 2021, das Zukunftsmuseum, 'The Museum of the Future').
In the following episodes, we will look in more detail at each aspect of this work, and the team members who are specifically involved.
Critical Climate Machine is part of the MediaFutures project, of which IRCAM is a partner.
Critical Climate Machine has received funding from the European Union's Horizon 2020 program for research and innovation, under grant agreement n° 951962
Photos 1 et 3 : © Gaëtan Robillard
