Dr. Thomas Parr, co-author of 'Active Inference' book, discusses the unifying Active Inference framework by Prof. Karl Friston, exploring living systems, agency, biology, and consciousness. Topics include resisting entropy, deep questions about causality, modeling, and consciousness, and the complexity of understanding language statistics and hidden causes.
Active Inference unifies physics, biology, and psychology to explore agency and causality.
The dual perspective of Active Inference involves holistic resistance to entropy and mathematical Bayesian mechanics.
Dr. Thomas Parr emphasizes interdisciplinary collaboration and a unified resource for navigating active inference fields.
In active inference, individuals control surprise to maintain homeostasis by optimizing beliefs and approximating surprise through variational inference.
Deep dives
Active Inference: A Comprehensive Overview
Active Inference, as discussed in the podcast, delves into two key perspectives: the high road and the low road. The high road offers a holistic view, emphasizing how organisms resist entropic forces by minimizing free energy. In contrast, the low road delves into the mathematical aspects, exploring Bayesian mechanics, variational inference, and treating perception as perceptual inference. This dual approach converges to highlight the role of action in perception and planning within active inference.
Professor Karl Fristen's Insights on Active Inference
Professor Karl Fristen's preface elucidates that active inference portrays a way to comprehend sentient behavior, where our actions are viewed as perceptual inference and planning as inference to resolve uncertainties. Fristen accentuates how active inference intertwines action and perception, transforming planning into resolving uncertainties about the surrounding environment.
Dr. Thomas Parr's Dual Role and Academic Journey
Dr. Thomas Parr, a postdoctoral scholar and clinician, shares his transition into active inference research under Fristen's mentorship. Reflecting on his collaborative work in writing the active inference book, Parr highlights the interdisciplinary nature of the field and the necessity of a unified resource to navigate the diverse domains active inference encompasses.
The Complex Dynamics of Agency and Predictive Systems
The discussion delves into the complexity of agency within predictive systems like active inference. It navigates the nuances of agent preferences, the role of historical systems dynamics, and the challenge of imbuing artificial agents with autonomy. The dialogues intertwine philosophical ponderings on agency, machine understanding, and the intertwining of internal models with influencing external environments.
Active Influence and Minimizing Surprise
Active influence involves keeping things within a minimally surprising range to maintain parameters like heart rate and blood pressure. By anticipating and controlling surprise, individuals affect their sensory inputs and strive for homeostasis. Dynamics and energy are linked to the system's tendency to move towards more probable states and avoid surprising ones.
Variational Inference and Model Building
Variational inference optimizes approximate posterior distributions to minimize free energy and approximate surprise. By adjusting beliefs about the world, individuals reduce the distance between free energy and surprise, leading to improved model fitting. Model building involves adding or subtracting states and causal factors based on predictions and exploration of model space.
Consciousness and Life as Dynamic Systems
The discussion expands to include concepts like sentience, consciousness, and life as dynamic, reducible systems. Active inference blurs distinctions between living and non-living systems by emphasizing continuous dynamics in understanding biology. The fluidity in defining complex concepts highlights the need for precise mathematical expressions to explore inherent ambiguities.
Thomas Parr and his collaborators wrote a book titled "Active Inference: The Free Energy Principle in Mind, Brain and Behavior" which introduces Active Inference from both a high-level conceptual perspective and a low-level mechanistic, mathematical perspective.
Active inference, developed by the legendary neuroscientist Prof. Karl Friston - is a unifying mathematical framework which frames living systems as agents which minimize surprise and free energy in order to resist entropy and persist over time. It unifies various perspectives from physics, biology, statistics, and psychology - and allows us to explore deep questions about agency, biology, causality, modelling, and consciousness.
Buy Active Inference: The Free Energy Principle in Mind, Brain, and Behavior
https://amzn.to/4dj0iMj
YT version: https://youtu.be/lbb-Si5wa_o
Please support us on Patreon to get access to the private Discord server, bi-weekly calls, early access and ad-free listening.
https://patreon.com/mlst
Chapters should be embedded in the mp3, let me me know if issues
Get the Snipd podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share & Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode