With the ever-growing adoption of AI-based systems, the carbon footprint of AI is no longer negligible. According to recent figures, training a state-of-the-art NLP model can generate as much carbon as five cars during their entire lifespan (including fuel). The situation is only expected to worsen as Large Language Models (LLMs) are becoming of widespread adoption.
According to the interview, instead of solely chasing new models for the sake of speed and accuracy, researchers and practitioners should also start using the numerous existing solutions to mitigate the environmental impact of AI.
The statement was based on the recently published review on Green AI, where numerous efficient and effective Green AI solutions were collected from the literature.
The feature of the STLab research efforts in the newspaper highlights the underlying vision of the group in developing research that lies at the intersection of academic rigor and industrial relevance, and communicating the research findings to a wider audience.
The full New York Times article is available at this link.