On the morning of October 30, at the first AI-driven science seminar of Tianqiao Brain Science Research Institute, Shanda Group and Chen Tianqiao, the founder of Tianqiao Brain Science Research Institute, a world-renowned innovative entrepreneur and philanthropist, announced that they would invest US$1 billion in computing power to support innovative artificial intelligence research by scientists around the world.
Chen Tianqiao believes that the ultimate value of AI is discovery. Discovery intelligence can proactively build testable theoretical models about the world, propose falsifiable hypotheses, and continuously revise its understanding framework through interaction with the world and self-reflection. This is the true sense of general artificial intelligence. It can ask questions rather than just answer them, understand patterns rather than just predict results, transcend imitation, and have the essential ability to create and discover these intelligences. The meaning of AGI is no longer "replacing humans" but "evolving humans."
Chen Tianqiao pointed out that building discovery intelligence requires building five capabilities, which together form a viable, discovery-oriented intelligent closed loop:
First, neurodynamic structures—sustained, self-organizing activity that keeps the system active over time.
Second, long-term memory – flexible storage and selective forgetting to build knowledge and formulate hypotheses.
Third, causal inference mechanisms—mechanisms for inferring beyond training distributions.
Fourth, the world model – an internal, unified simulation used to predict the future and mentally test ideas.
Fifth, metacognition and intrinsic motivation systems—uncertainty awareness, attention control, and curiosity-driven exploration.
To help scientists around the world advance discovery-based intelligence research, Chen Tianqiao announced a number of supports specifically targeted at young scientists. Specifically include:
First, Benchmark—a comprehensive assessment suite across neurodynamics, memory, causality, world models, and metacognition, with discoverability as a core metric.
Second, structural computing power - invest $1 billion in computing power to prioritize clusters that support structural experiments (memory systems, causal architectures, neurodynamic hypotheses).
Third, the PI incubator - provides an independent path for doctoral students and postdocs to establish laboratories named after themselves, lead teams, and pursue bold ideas without waiting for traditional timelines, while establishing R&D centers around the world.