In the past ten years, artificial intelligence has encountered such dramatic progress that it is now seen as a tool of choice to solve environmental issues and, in the first place, greenhouse gas emissions (GHG). At the same time, the deep learning community began to realize that training models with more and more parameters require a lot of energy and, as a consequence, GHG emissions. To our knowledge, questioning the complete net environmental impacts of AI solutions for the environment (AI for Green) and not only GHG, has never been addressed directly. In this article, we propose to study the possible negative impacts of AI for Green. First, we review the different types of AI impacts; then, we present the different methodologies used to assess those impacts and show how to apply life cycle assessment to AI services. Finally, we discuss how to assess the environmental usefulness of a general AI service and point out the limitations of existing work in AI for Green.

Résumé La recherche de réponses précises à des questions, aussi appelée « questions-réponses », est une évolution des systèmes de recherche d’information : peut-elle, comme ses prédécesseurs, se satisfaire de méthodes essentiellement numériques, utilisant extrêmement peu de connaissances linguistiques ? Après avoir présenté la tâche de questions-réponses et les enjeux qu’elle soulève, nous examinons jusqu’où on peut la réaliser avec très peu de connaissances linguistiques. Nous passons ensuite en revue les différents types de connaissances linguistiques que les équipes ont été amenées à mobiliser : connaissances syntaxiques et sémantiques pour l’analyse de phrases, rôle de la reconnaissance d’« entités nommées », prise en compte de la dimension textuelle des documents. Une discussion sur les contributions respectives des méthodes linguistiques et non linguistiques clôt l’article.

Recent Machine Learning (ML) approaches have shown increased performance on benchmarks but at the cost of escalating computational demands. Hardware, algorithmic and carbon optimizations have been proposed to curb energy consumption and environmental impacts. Can these strategies lead to sustainable ML model training? Here, we estimate the environmental impacts associated with training notable AI systems over the last decade, including Large Language Models, with a focus on the life cycle of graphics cards. Our analysis reveals two critical trends: First, the impacts of graphics cards production have increased steadily over this period; Second, energy consumption and environmental impacts associated with training ML models have increased exponentially, even when considering reduction strategies such as location shifting to places with less carbon intensive electricity mixes. Optimization strategies do not mitigate the impacts induced by model training, evidencing rebound effect. We show that the impacts of hardware must be considered over the entire life cycle rather than the sole use phase in order to avoid impact shifting. Our study demonstrates that increasing efficiency alone cannot ensure sustainability in ML. Mitigating the environmental impact of AI also requires reducing AI activities and questioning the scale and frequency of resource-intensive training.