Neuroevolution: Harnessing Creativity in AI Agent Design

A comprehensive, cutting-edge introduction to neuroevolution that equips AI professionals and advanced students to harness evolutionary creativity for innovative AI solutions.

Neuroevolution—the fusion of neural networks with evolutionary algorithms—is a powerful approach to AI model design. Providing a comprehensive, cutting-edge introduction to the field, this textbook methodically moves from fundamental principles to state-of-the-art techniques, highlighting how evolutionary processes can creatively design neural network architectures and behaviors beyond what conventional methods achieve.

The book starts with core concepts of evolutionary computation and basic neuroevolution examples, followed by key design considerations like genetic encodings and the critical role of diversity. It then covers evolving intelligent agents,  from low-level control through intelligent behavior to high-level multi-agent decision-making strategies. Next, it discusses integrating neuroevolution with other modern AI methods such as large language models  and reinforcement  learning. Finally, it offers insights into biological evolution, examining evolved neural structures, developmental processes, coevolution of bodies and brains, and the evolution of behaviors and language. Written by pioneers of the field, this timely book shows how  the creativity of evolution can be harnessed in AI, automating the discovery of novel neural network solutions that improve adaptability and performance in complex, uncertain environments.

• Provides a structured, step-by-step path from basics to advanced topics
• Details essential mathematical and conceptual foundations 
• Unifies decades of research in one place, explaining key algorithms with illustrative examples
• Explores how neuroevolution interfaces with contemporary AI advances
• Includes extensive real-life case studies across robotics, games, art, and decision-making
• Features a comprehensive online supplement with demos, tutorials, lecture slides and videos, exercises and code

Sebastian Risi is a Professor at the IT University of Copenhagen, where he directs the Creative AI Lab, and a Research Scientist at Sakana AI. Yujin Tang is a Research Scientist at Sakana AI, with a research focus on machine learning and evolutionary algorithms. David Ha is a cofounder and CEO of Sakana AI. Risto Miikkulainen is a Professor of Computer Science at the University of Texas at Austin and VP of AI Research at Cognizant AI Labs.