AI Three Questions②: Scientific Inquiry | WAIC 'Questions of Science' Series to Explore New Frontiers in AI and Scientific Integration}

WAIC 2025 — World Artificial Intelligence Conference

Forum: July 26-28, 2025

Exhibition: July 26-29, 2025

Locations include: Expo Center, Expo Pavilion, Xuhui West Bund, etc.

To deepen the core themes of AI, this year WAIC prominently introduces the “Three Questions of AI” — addressing cutting-edge topics in mathematics, science, and modeling. The mathematical question involves axioms and formulas to build cognitive frameworks; the scientific question roots in empirical evidence to explore nature; and the modeling question integrates both, transforming abstractions into practical applications.

The three questions are interconnected: mathematics quantifies science, science gives meaning to mathematics, and models realize intelligence. Their synergy fosters innovation across diverse fields.

The upcoming “Questions of Science” series at WAIC aims to focus on key propositions in the integration of AI and science, providing a high-quality academic platform. Whether for AI scholars or industry professionals, this event promises profound insights and frontier discussions.

AI and Human Scientists: Collaboration of Rational Analysis and Creative Inspiration

In scientific research, human scientists rely on intuition and inspiration to pioneer new fields, but face limitations with large data and complex analysis. AI excels in data processing and logical reasoning. How can AI’s analytical power complement human intuition for breakthroughs? This is crucial for improving research efficiency and fostering major scientific advances in the AI era. For example, in drug discovery, AI can analyze vast biological data to identify potential targets, while human scientists validate feasibility, significantly shortening development cycles.

Data and Models: Systematic Integration to Overcome Uncertainty Barriers

Scientific research involves multimodal data with significant differences across fields like biology and physics, akin to cross-language communication. How to align representations of different scientific data modalities is fundamental for cross-disciplinary collaboration and knowledge innovation. In healthcare, aligning medical imaging with clinical data supports precise diagnosis and personalized treatment. Building world models for causal inference and bridging semantic gaps between physical and digital worlds are vital for future societal development, enabling accurate predictions in climate, finance, and beyond, and facilitating seamless virtual-physical integration in smart factories, urban management, and transportation.

Computational Boundaries: Exploring Quantum and Classical Synergies

The boundary between quantum and classical computing in large-scale scientific simulation is a core issue. Quantum computing offers exponential speedups for specific tasks, while classical computing excels in stability and versatility. Clarifying their collaboration can open new paths for hybrid architectures, enabling breakthroughs in materials science, cryptography, and fundamental physics, especially under current resource constraints, and accelerating scientific discovery.

Life Sciences: Holistic Data-Driven Breakthroughs

In life sciences, AI’s role in generating original hypotheses, constructing virtual cells and organs, conducting virtual experiments, and advancing high-throughput systems is crucial for tackling major diseases and improving health. AI-driven breeding also offers innovative solutions for global food security.

Materials Science: Multi-Dimensional Analytical Innovation

Materials science is undergoing transformative change with AI. Recognizing and analyzing high-dimensional data, evolving equations, and cross-scale systems are key to understanding material properties. AI enhances atmospheric modeling, simulates stellar evolution, accelerates particle physics discoveries, and predicts material behaviors like superconductivity, pushing the boundaries of traditional research.

In materials development, graph neural networks and databases enable automated material screening, property prediction, and process optimization, revolutionizing R&D and manufacturing processes.

In this “Questions of Science” series, young researchers will engage in deep dialogues, exploring innovative ideas and solutions to key scientific challenges, providing forward-looking insights into AI and science integration. If you wish to understand how AI reshapes scientific research and witness frontier discussions, this series invites your participation. Let’s explore the infinite possibilities of AI and science together.

Original article link: https://mp.weixin.qq.com/s/2grPxBNrYyHH-YUJ9-PxLQ