The Cycle of Conferences on Artificial Intelligence, entitled “Electricity in the 21st century”, organized by the Royal Galician Academy of Sciences (RAGC) in collaboration with the Diputación da Coruña, comes to an end today, with the conference given by the CITIC researcher Verónica Bolón, head of the Computer Science and Artificial Intelligence area of the Universidade da Coruña.

In a world where sustainability became a global priority, the integration of artificial intelligence (AI) in the search for a greener and more sustainable future is presented as a fundamental evolution. This talk will explore the transformative power of AI in the context of environmental sustainability and how this combination can fundamentally change the way we address ecological challenges.

The term “green AI” encompasses two distinct approaches, as Verónica Bolón explains. On the one hand, it refers to the application of AI techniques to address problems related to climate change, such as optimizing energy consumption. On the other hand, it refers to the development of algorithms that “are more inclusive and environmentally friendly, not only by producing new results without increasing computational cost, but also by ensuring that any researcher with a conventional computer has the opportunity to conduct cutting-edge research without the need to resort to expensive cloud servers.”

Typical research in artificial intelligence, sometimes referred to as “red AI”, aims to obtain state-of-the-art results at the cost of using large computational power, usually through a large amount of training data and numerous experiments. In her talk, the CITIC researcher will address why the energy consumption of some AI algorithms is a legitimate concern. The largest and most powerful AI models, such as deep neural networks, require a substantial amount of energy for training and operation. This not only carries cost implications, but can also contribute to increased carbon emissions and the depletion of finite energy resources.

In addition, key strategies and advances in the quest for “green AI” will be examined. This includes the development of more energy efficient algorithms, the optimization of specialized hardware for AI tasks, and the adoption of renewable energy sources to power the data centers running these algorithms.