It’s no secret that AI’s consumption of water, energy, and land, is raising serious concerns in many quarters with some questioning whether we really need to use vast LLMs trained on vast, often “polluted” datasets, to perform relatively simple tasks.