AI scientists can’t optimize worsening energy efficiency: Why behind-the-scenes AI is costing you

Energy Consumption of AI - AI's electricity usage is unprecedented, with ChatGPT using nearly 10 times more electricity per query than a Google search [2] - AI requires constant computations, immense processing power, vast memory, and cooling, consuming energy on a scale rivaling entire cities [3] - Bloomberg Intelligence estimates the AI industry will grow 42% by 2030, becoming a $13 trillion market [7] - The surge in energy demand from AI is compared to the biggest surge since the invention of air conditioning [7] Impact on Utility Prices - Utility prices have skyrocketed by 267% in regions dominated by data centers [4] - Missouri saw the largest increase this year with costs rising more than 38%, followed by North Dakota at over 33%, New Jersey at 28%, and Iowa at 27% [6] - The Trump administration's bill provides over $1 billion in federal funding for unregulated AI expansion, potentially worsening the situation [6] - Utility companies are expanding power grids and upgrading infrastructure to handle AI demand, with costs passed to consumers [8] Regulatory and Legislative Responses - Newly elected governors in Virginia and New Jersey are demanding that PJM Interconnection cut costs for consumers, threatening to withdraw if they don't comply [8][9] - Oregon passed the Power Act to help utilities negotiate fairer deals with data centers and crypto miners [9] - Nevada's clean transition tariff forces Google to pay 100% of the costs for new generation, protecting ratepayers [10] - Pennsylvania regulators are pushing for data centers to pay their fair share for grid usage [10]