In the context of AI, energy demand has become a critical barrier to progress in the US, where data centers are struggling to meet growing electricity needs. China is rapidly expanding its renewable energy capacity, with 429 GW added in 2024, significantly outpacing the US. This disparity could jeopardize the US’s position in both energy and AI innovation as reliance on inefficient coal plants increases.
To address the energy strains from AI, the article suggests flexibility from data centers in energy consumption during peak times, which could alleviate pressure on the grid. Studies indicate that even slight reductions in demand could allow for considerable new capacity without requiring extensive new infrastructure. However, the uncertainty surrounding future AI energy demands complicates planning, with predictions varying dramatically from double to quadruple current usage. The global trend towards renewables is evident, and unless the US adapts, it risks falling behind as countries like China emerge as leaders in both green energy and AI technologies.
👉 Pročitaj original: MIT Technology Review – AI
