This AI Finds Simple Rules Where Humans See Only Chaos

Researchers at Duke University published a study on December 22, 2025, describing a new AI methodology for analyzing ultra-complex dynamic systems. Traditionally, systems like climate turbulence or complex biological processes were described by neural networks as "black boxes"—models predicted results but didn't explain causes. The new approach, published in npj Complexity, allows AI to extract compact, interpretable mathematical equations directly from data, finding hidden order in chaotic processes.

This breakthrough is fundamental for physics, biology, and climatology. Instead of relying on massive computational models, scientists can now obtain elegant formulas that are easy for human understanding and verification. The methodology effectively turns AI into an "autonomous scientist" capable of discovering new laws of nature where the human brain sees only unstructured noise. Developers believe this will lead to a series of discoveries in climate change forecasting and understanding cellular mechanisms in the coming years.

Source: ScienceDaily