前中国首富盛大陈天桥：宣布投10亿美元算力支持发现式智能！共同形成一个有生命力、面向发现的智能闭环

Core Insights - The core message of the news is that Chen Tianqiao, founder of the Tianqiao Brain Science Research Institute, announced a $1 billion investment in computational power to support global scientists in innovative artificial intelligence research [1][8]. Group 1: AI and Discovery - Chen Tianqiao believes that the ultimate value of AI lies in discovery, emphasizing that discovery-oriented intelligence can actively construct testable world models and propose falsifiable hypotheses, which is essential for true general artificial intelligence (AGI) [3]. - This form of intelligence is characterized by its ability to ask questions rather than just answer them, understand patterns instead of merely predicting outcomes, and possess the inherent capability for creativity and discovery [3]. Group 2: Components of Discovery-Oriented Intelligence - To build discovery-oriented intelligence, five capabilities are necessary: 1. Neural dynamic structure, which ensures continuous and self-organizing activity to keep the system active over time [4]. 2. Long-term memory, allowing flexible storage and selective forgetting to establish knowledge and form hypotheses [5]. 3. Causal reasoning mechanisms that enable inferences beyond training distributions [6]. 4. World models that serve as an internal, unified simulation for future predictions and psychological testing of ideas [6]. 5. Metacognition and intrinsic motivation systems that foster awareness of uncertainty, attention control, and curiosity-driven exploration [7]. Group 3: Support for Young Scientists - To assist global scientists in advancing discovery-oriented intelligence research, several initiatives specifically targeting young scientists were announced: 1. Benchmark testing, a comprehensive assessment suite across neural dynamics, memory, causality, world models, and metacognition, with discoverability as a core metric [8]. 2. Structural computational power, with a $1 billion investment prioritized for supporting structural experiments in memory systems, causal architectures, and neural dynamic hypotheses [8]. 3. A PI incubator that provides independent pathways for PhD students and postdocs to establish their own laboratories, lead teams, and pursue bold ideas without waiting for traditional timelines, alongside the establishment of global R&D centers [9].