山姆·奥尔特曼：AI将在2030年前超越人类智力

Core Viewpoint - OpenAI CEO Sam Altman predicts that by the end of this decade, specifically by 2030, AI will surpass human intelligence in capabilities, completing tasks that humans cannot accomplish [2][3]. Group 1: AI Development Predictions - Altman expresses confidence that by 2030, extraordinary AI models capable of performing tasks beyond human abilities will be developed, and he would be surprised if this does not happen [2]. - He anticipates that models developed as early as 2026 could be "quite astonishing" in their capabilities [2]. - Altman's prediction is more conservative compared to other AI leaders, with some believing AI will outperform humans in nearly all aspects by 2027 [3]. Group 2: Economic Impact of AI - Altman envisions that in the near future, 30% to 40% of tasks in the current economy could be performed by AI [3]. - He emphasizes the importance of understanding human needs and desires in the context of AI's capabilities, suggesting that while AI will be a powerful tool, human oversight will remain crucial [4]. Group 3: Infrastructure and Limitations - The operation of ChatGPT is currently limited by the substantial energy requirements and the need for large data centers [4]. - OpenAI is collaborating with Oracle and SoftBank to build a significant data center complex in Texas, which is part of a larger infrastructure project aimed at supporting AI development [4]. - Despite ongoing infrastructure development, Altman notes that current efforts are still insufficient to meet the demands of ChatGPT [4]. Group 4: Future of AI and Scientific Discovery - Altman believes that in a few years, AI will be capable of making scientific discoveries that humans cannot achieve alone, which he refers to as "superintelligence" [5].