Evolution might not be a series of "happy accidents" after all, but a system that actually learns where to mutate.
April 16, 2026
Original Paper
An abstract model of nonrandom, non-Lamarckian mutation in evolution using a multivariate estimation-of-distribution algorithm
arXiv · 2604.12884
AI-generated illustration
The Takeaway
For over a century, the bedrock of biology has been that mutations are totally random—nature throws darts in the dark, and some just happen to hit the bullseye. This paper presents a model where mutations are actually "structured" by information that the genome has accumulated over millions of years. It suggests that life has an internal logic that guides where changes happen, making evolution far more efficient than we ever dreamed. It’s like finding out the darts weren't being thrown randomly, but by a player who was slowly learning how to aim. This would fundamentally change our understanding of how every species on Earth—including us—came to be.
From the abstract
At the fundamental conceptual level, two alternatives have traditionally been considered for how mutations arise and how evolution happens: 1) random mutation and natural selection, and 2) Lamarckism. Recently, the theory of Interaction-based Evolution (IBE) has been proposed, according to which mutations are neither random nor Lamarckian, but are influenced by information accumulating internally in the genome over generations. Based on the estimation-of-distribution algorithms framework, we pre