Mathematicians unify evolution across any substrate — biology, AI, chemistry, whatever replicates

Researchers have formalized the mathematical structure underlying any evolutionary system — not just DNA, but any process where things copy themselves, vary, and get selected. This means evolution isn't a biological concept; it's a pattern you can study and predict in artificial systems, chemistry, or AI.

This is foundational math work, not a deployed system or policy change. It doesn't immediately change what anyone builds or how regulators operate. But it does something quieter: it gives researchers a unified language for thinking about replication dynamics across completely different domains. If you're building an AI system that learns through iterated variation and selection, or studying how molecules self-organize, or designing an evolutionary algorithm, you now have a single mathematical framework instead of domain-specific models. The practical payoff is long-term — better predictions about stability, error tolerance, and evolutionary dynamics in any system that replicates.