Course directory

KDDC 2EN23 - Beyond AI hype: its impacts on local ecologies

KDDC 2EN23 · Fall semester 2026-2027

Language of tuition
English
Course type
Cours séminaire
Number of hours
12
Credits
2

Information

Description

For tech CEOs such as Sam Altman (Open AI) and Elon Musk (Space X), the future of artificial intelligence is tied to the future of energy. This means that the race between companies has serious implications for the energy provision of other industries as well as citizens. This also means that any future AI workers must have an understanding of AI’s materiality. Beyond the hype around AI, this class introduces students to the real world impacts this set of technology has on energy markets, land and local ecologies. Unpacking AI relations to local environments requires to see it as an infrastructure composed of developers, chips, data centers and cloud providers. These various entities will be analysed with a specific attention to their relations with issues of electricity, heat, land uses, pollution and biodiversity.

Prerequisites

No prerequisites. Before the start of the course, students must read this article : « Phase G: GPUs and generative AI as a new phase in the environmental history of digitalisation? (Part 1) », Gauthier Roussilhe, March 2025, URL:  https://gauthierroussilhe.com/en/articles/phase-g-gpus-and-generative-ai-as-a-new-phase-in-the-environmental-history-of-digitalisation-part-1

Teaching format

Student will join our class sessions well prepared and ready to contribute to our discussion.Your thorough preparation of all cases is central to the effectiveness of the classes. The more you prepare, the better these will be. Student will engage with this module's content through assigned background readings, interactive presentations/discussions led by the session leader. As always, sessions will be highly interactive, with the session leader/faculty member encouraging you to ask questions, engage in discussions, debate contentious/ethical issues, and share your own experiences.

I won’t use AI to prepare my class except DeepL for translation.

For the individual article review as well as the group class, the use of any Generative AI is forbidden.

Since students are not experts on the topic they won’t be able to assess if a given Gen AI model has made a mistake. Moreover, reading and synthesising content without a Gen AI is part of the process to become a critical citizen/worker.

Assessment

  • Oral participation (individual) 10 %
  • Review of an article to be done remotely (individual) 30%
  • Group case study analysis to do in class 60%

Readings

Main readings 2

  1. Roussilhe, Gauthier, « Phase G: GPUs and generative AI as a new phase in the environmental history of digitalisation? (Part 1) », March 2025, URL:  https://gauthierroussilhe.com/en/articles/phase-g-gpus-and-generative-ai-as-a-new- phase-in-the-environmental-history-of-digitalisation-part-1

  2. Wang, Xiaowei R., «  Toxic Residue and Cold Wars: Building the Chips that Power the Cloud « , Data & Society Blog, April 2025, URL : https://datasociety.net/points/toxic- residue-and-cold-wars-building-the-chips-that-power-the-cloud/

Further readings 5

  1. AlgorithmWatch (2022) sustain: Sustainable AI in Practice. AlgorithmWatch Magazine. https://algorithmwatch.org/en/sustain-magazine-2022 (accessed 30 March 2025).

  2. Brevini, Benedetta, « Black boxes, not green: Mythologizing artificial intelligence and omitting the environment ». Big Data & Society 7(2): 1–5, 2020.

  3. Cellard, Loup, Parker, Christine, Haines, Fiona, « Beyond AI as an environmental pharmakon: Principles for reopening the problem-space of machine learning’s carbon footprint », Environment and Planning E, 2025., URL : https://journals.sagepub.com/doi/10.1177/25148486251332087

  4. Roussilhe, Gauthier, « AI data centres: between energy requisition and environmental eclipse », February 2026, URL : https://gauthierroussilhe.com/en/articles/ai-data- centres-between-energy-requisition-and-environmental-eclipse

  5. Roussilhe, Gauthier, « How to use computing power faster: on the weird economics of semiconductors and GenAI », July 2025, URL : https://gauthierroussilhe.com/en/articles/how-to-use-computing-power-faster

Teaching units

Unit no.Unit type Teaching unitsTeachers Number of hoursCredits
20123Cours séminaire Beyond AI hype: its impacts on local ecologies Loup CELLARD122

Teachers

  • Loup CELLARD

In the curricula

Contacts and resources

Contact
JEROME Amélie