Patterns of Heat

Description

The scientific consensus that climate change is man-made is constantly being challenged, especially in social media. Is it possible to use artificial intelligence to make this stream of misinformation visible?

For the work »Patterns of Heat«, artist Gaëtan Robillard drew on recent research in technocognition: using the latest methods, it is possible to examine large amounts of online data in the context of climate skepticism and identify misleading claims. The network implemented in the installation detects and classifies misinformation posted on the social media platform Twitter. Every four minutes, 896 tweets are analyzed.

»Patterns of Heat« shows the flow of information on three levels: The sculpture on the floor heats up in proportion to the amount of computation involved in analyzing the tweets. The changing numbers lighting up on the sculpture reveal what kind of misinformation it is. (You can find the meaning of the number codes on the screen on the wall.) Finally, the results of the analysis are displayed on the screen in a real-time animation: The warmer the installation and the higher the number of misinformation, the warmer the color tones.

The photo of the flooded city of Erftstadt, titled »07/17/2021«, evokes the memory of a recent weather event and points to the destructive power of global warming.

Team

Gaëtan Robillard, Özlem Sulak, Vincent Nozick, Laurine Capdeville, Jolan Goulin

Machine learning algorithm and training set

John Cook, Travis Coan, Constantine Boussalis

Partner

ZKM | Center for Art and Media, Deutsches Museum, Ldi Esad TALM

»Patterns of Heat« was created within the framework of the »intelligent.musuem« project at the Hertz-Lab of ZKM | Center for Art and Media and at the Deutsche Museum. It is part of the research project »Critical Climate Machine« by Gaëtan Robillard, which is supported by the »MediaFutures« programm and funded by the European Union's »Horizon 2020« framework. »Patterns of Heat« was also supported by LDI Esad TALM.