The US saw a 4% increase in electricity use this year compared to last, largely attributed to the rapid growth of data centers supporting artificial intelligence (AI) usage. However, with coal-fired power seeing a 20% surge, concerns about AI’s environmental impact grow.
While academics can test individual AI models’ energy needs, real-world scenarios are harder to assess. Google, on the other hand, offers valuable insights through its analysis of search queries, revealing that energy consumption has dropped by 33 times over the past year. This shift highlights the need to consider all factors in AI’s environmental impact, including hardware manufacturing, storage, cooling, and training requirements.
Source: https://arstechnica.com/ai/2025/08/google-says-it-dropped-the-energy-cost-of-ai-queries-by-33x-in-one-year