242 | David Krakauer on Complexity, Agency, and Information
Jul 10, 2023
01:33:19
AI Summary
Highlights
AI Chapters
Episode notes
auto_awesome
Podcast summary created with Snipd AI
Quick takeaways
Emergence is the result of broken symmetry and effective theories, allowing for a higher level understanding of complexity.
The transition from adaptation to agency is a crucial aspect of emergence, involving principles like natural selection and encoded information.
Noise plays a valuable role in the emergence and persistence of complex systems, enabling exploration and balancing adaptation with innovation.
Deep dives
Emergence: Broken Symmetry and Effective Theories
Emergence can be understood as the result of broken symmetry and the ability to formulate effective theories at higher levels of complexity. Broken symmetry occurs when a particular state is chosen over others, even though they may be equally probable under the laws of physics. This concept is exemplified by the chirality of molecules, where left-handed and right-handed versions occur asymmetrically. Effective theories are an abstraction that allows for understanding and prediction of macroscopic phenomena without needing to consider all microscopic constituents. These theories are dynamically sufficient, meaning they provide comprehensive explanations without requiring knowledge of the underlying physics. Emergence is the manifestation of broken symmetry and the formulation of effective theories, resulting in a higher level understanding of complexity.
Agent Complexity and Memory
The transition from adaptation to agency is a key aspect of emergence, where agents develop policies or procedures to navigate and interact with their environment. This transition is governed by principles such as natural selection and the accumulation of encoded information through memory. The concept of the Eigus, or information-gathering utilizing system, provides a framework for understanding the existence of individuals and their ability to encode the history and behavior of complex systems. However, the origin of such policies and the increasing complexity of agents remain open questions in the study of emergence.
The Role of Noise and Exploration
Noise plays a valuable role in the emergence and persistence of complex systems. It enables exploration of solution space, where it helps in discovering new levels and characteristics of complexity. Noise is essential for processes like mutation, which introduces variability and enables evolution. It facilitates exploration of different possibilities while balancing the need for exploitation of favorable solutions. Noise can be seen as a constructive and destructive force in complex systems, contributing to their adaptation and innovation.
Individuality as an Emergent Phenomenon
The concept of individuality in complex systems emerges from the understanding of emergent properties, such as broken symmetry and effective theories. Individuality arises when systems are recognized as distinct and separate entities with unique characteristics. While there is no fundamental essence defining individuality, it can be explained by the ability to formulate effective theories that account for the behavior and properties of specific entities. Individuality is a manifestation of emergent complexity, allowing for the delineation of coherent wholes within complex systems.
Understanding the nature of individuals and complexity
The podcast explores the concept of individuals and complexity, focusing on the idea that complex domains come in the form of packages called individuals or organisms. This perspective challenges the notion that genes are the only unit of selection, emphasizing that individuals play a crucial role in propagating adaptive information forward in time. The podcast discusses how a theory using information theory as a lens can detect different kinds of individuals in complex systems and highlights the importance of understanding the causal units of complex systems.
The future of complexity and the study of intelligence
The podcast reflects on the future of complexity science and its potential for unification. While the field has not achieved a single unified theory of complexity, there are promising signs of synthesis and connections between different subfields. The podcast also delves into the topic of intelligence, considering it as a universal phenomenon. It explores the emergence of large language models and their impact on our understanding of intelligence, highlighting the need for a pluralistic attitude to define and comprehend concepts like intelligence, meaning, understanding, and consciousness.
Complexity scientists have been able to make an impressive amount of progress despite the fact that there is not universal agreement about what "complexity" actually is. We know it when we see it, perhaps, but there are a number of aspects to the phenomenon, and different researchers will naturally focus on their favorites. Today's guest, David Krakauer, is president of the Santa Fe Institute and a longtime researcher in complexity. He points the finger at the concept of agency. A ball rolling down a hill just mindlessly obeys equations of motion, but a complex system gathers information and uses it to adapt. We talk about what that means and how to think about the current state of complexity science.
Blog post with transcript: https://www.preposterousuniverse.com/podcast/2023/07/10/242-david-krakauer-on-complexity-agency-and-information/
David Krakauer received his D.Phil. in evolutionary biology from Oxford University. He is currently President and William H. Miller Professor of Complex Systems at the Santa Fe Institute. Previously he was at the University of Wisconsin, Madison, where he was the founding director of the Wisconsin Institute for Discovery and the Co-director of the Center for Complexity and Collective Computation. He was included in Wired magazine's list of "50 People Who Will Change the World."
