The Green Artificial Intelligence Model #

The demand for resources driven by Artificial Intelligence (AI) continues to surge, whether it is due to vertical scaling, where Machine Learning (ML) Models are growing in size (e.g., Large Language Models (LLMs), Diffusion Models), or horizontal scaling, involving the deployment of smaller models on millions of devices (e.g., TikTok filters, etc.). Additionally, AI is enabling a multitude of new applications, leading to a proliferation of AI applications. Given the immense potential of AI in addressing climate change, the solution is not to cease its use but rather to employ it in a smart and resource-efficient manner. Various initiatives are already underway to tackle this challenge, such as frameworks that render CO₂ emissions transparent, more computationally efficient algorithms, and energy-efficient hardware, among others. However, what’s currently lacking is a comprehensive model that encompasses the entire AI lifecycle. This is where the Green Artificial Intelligence Model comes into play.

• Introduction
• Quantifiying the Environmental Footprint
• Influence Factors
• Interplay of Influence Factors
• Discussion