Cheaper AI does not mean greener AI

Summary

The article argues that while advancements are making AI models cheaper to train and deploy, this does not necessarily lead to lower environmental impact. Lower costs can drive increased usage and proliferation of AI applications, potentially resulting in greater overall energy consumption and carbon emissions, highlighting the need for sustainable AI practices beyond just cost reduction.