Our approach to energy innovation and AI’s environmental footprint
Google published a technical report on the environmental impact of AI, including calculating the cost of training models (not just individual queries). They reported energy use much lower than many earlier estimates, partly due to improvements in efficiency.
Over a 12-month period, while delivering higher-quality responses, the median energy consumption and carbon footprint per Gemini Apps text prompt decreased by factors of 33x and 44x, respectively. Based on our recent analysis, we found that our work on efficiency is proving effective and the energy consumed per median prompt is equivalent to watching television for less than nine seconds.