Share on :

OBJECTIVE 2050: AI for the environment

16 January 2024 On the Research side
Published by Patrick BOUILLET
Viewed 90 times

Digital World - Objective 2050 is a monthly podcast focusing on technological innovation for tomorrow's world.

This all-new episode is dedicated to the role of artificial intelligence (AI) in the fight against climate change. Dejan Glavas, Professor of Sustainable Finance and Director of ESSCA's "AI for Sustainability" institute, explores the positive impact of AI, but also the environmental challenges posed by its use.

Exploring the different types of positive impact of AI

In the first part of this podcast, Dejan Glavas explains in detail the different types of positive impact AI can have on our environment.

Concrete examples such as water treatment, resource management and modeling to forecast climatic phenomena illustrate how AI is becoming an effective tool in our environmental arsenal, enabling greater precision, responsiveness and therefore efficiency.

In this section, Professor Glavas also discusses the use of AI in the financial sector, and in particular in non-financial reporting, through a few concrete examples.

The "but": rising environmental costs

Like many innovations, however, AI has less exciting aspects: its implementation and use necessarily entail negative consequences.

As Professor Glavas points out, AI is very energy-, resource- and hardware-intensive. Its impact on the environment is growing as its use expands, raising legitimate concerns.

Dejan Glavas gives us the keys to a better understanding of the ecological impact of AI, with relevant orders of magnitude and comparative examples.

Did you know, for example, that a ChatGPT prompt requires ten times more energy than a simple Google search?

Striking a delicate balance between precision and sobriety

The challenge is to strike a balance as quickly as possible between the quality of AI models and their consumption of energy and resources.

We need to direct our efforts towards a more sober use of AI, where model accuracy and quality do not compromise environmental sustainability.

Listen to the podcast

No comment

Log in to post comment. Log in.