In a recent interview, OpenAI CEO Sam Altman defended the energy costs associated with AI, likening it to the years required to raise a human, which involves substantial energy input. He pointed out that while AI training requires significant energy, it parallels the 20 years and resources needed to educate a human. Altman suggested that AI’s energy efficiency is nearing that of human learning, although he did not cite specific data. According to the MIT Technology Review, by 2028, AI could consume as much electricity as 22% of US households, with data centers projected to use over half their electricity for AI. Additionally, Altman criticized Elon Musk’s idea of establishing data centers in space, deeming it impractical for the current decade. His insights highlight the ongoing debate about AI’s energy consumption and its implications for sustainability in technology.
This summary incorporates relevant SEO terms while addressing the core of the discussion.
Source link
