Subvolts take Explore how brain principles could build more reliable AI, moving beyond current prediction models. Nov 19, 2025 - Developers
- Learn about alternative AI architectures and learning mechanisms beyond current large language models.
- Business
- Understand how new AI approaches might offer more reliable solutions for high-stakes workflows and autonomous agents.
Features a leading neuroscientist discussing a foundational theory with potential to reshape AI architectures beyond current dominant models.
#302 Karl Friston: How the Free Energy Principle Could Rewrite AI
Audio streams from Craig S. Smith. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Eye on AI's #302 episode features neuroscientist Karl Friston discussing how his Free Energy Principle could guide AI development beyond current transformer models. Host Craig Smith explores Friston's ideas on active inference, brain-inspired generative models, and Verses AI's Axiom architecture. They cover how Axiom uses probabilistic beliefs and message passing for learning, addresses transformer limitations, and grounds AI meaning through embodiment and uncertainty.
This episode explains how a scientist's idea about how brains work might create smarter AI. Instead of just guessing the next word, this new AI approach learns like a brain, understanding uncertainty and acting in the real world. This could lead to more reliable AI for important tasks.
This episode offers a deep dive into a theoretical framework, the Free Energy Principle, that could fundamentally alter AI architecture. It's for researchers and developers seeking alternatives to current large language models, particularly those interested in more robust, brain-like AI. The discussion on active inference and explicit uncertainty directly contrasts with the next-token prediction of models like OpenAI's GPT and Google's Gemini, suggesting a path toward more reliable AI for high-stakes applications.
- HostCraig S. Smith
- GuestKarl Fristonneuroscientist, Verses AI
- Guest: Karl Friston
- Title: How the Free Energy Principle Could Rewrite AI
- Show: Eye on AI
- Episode Number: #302
- Length: 63 minutes
- Architecture Mentioned: Axiom
This episode is sponsored by AGNTCY. Unlock agents at scale with an open Internet of Agents.
Visit https://agntcy.org/ and add your support.
How could Karl Friston's Free Energy Principle become a blueprint for the future of AI?
In this episode of Eye on AI, host Craig Smith sits down with Karl Friston, the neuroscientist behind the Free Energy Principle and advisor to Verses AI, to explore how active inference and brain inspired generative models might move us beyond transformer based systems. They unpack how Axiom, Verses' new architecture, uses probabilistic beliefs and message passing to build agents that learn like brains instead of just predicting the next token.
We look at why transformers face scaling and reliability limits, how Free Energy unifies prediction, perception, and action, and what it means for an AI system to carry explicit uncertainty instead of overconfident guesses. Learn how active inference supports continual learning without catastrophic forgetting, how structure learning lets models grow and prune themselves, and why embodiment and interaction with the real world are essential for grounding language and meaning.
You will also hear how Axiom can sit beside or beneath large language models, how explicit uncertainty can reduce hallucinations in high stakes workflows, and where these ideas are already being tested in areas like logistics, robotics, and autonomous agents. By the end of the episode, you will have a clearer picture of how Karl Friston's Free Energy blueprint could reshape AI architectures, from enterprise planning systems to embodied agents that understand and act in the world.
Stay Updated: Craig Smith on X: https://x.com/craigss Eye on A.I. on X: https://x.com/EyeOn_AI
- Free Energy Principle
- Karl Friston
- A scientific idea explaining how living things maintain themselves by minimizing surprise.
- active inference
- Karl Friston
- A way of thinking about how brains make predictions and take actions to reduce uncertainty.
- transformer based systems
- Craig Smith
- The current AI models, like ChatGPT, that are good at predicting the next word in a sequence.
- catastrophic forgetting
- Craig Smith
- When an AI model learns new information and forgets what it learned before.
55 min42 min51 min34 minUkraine's Secret Weapon: The Points System Winning the Drone War | Andrii Hrytseniuk
WithAndrii Hrytseniuk
Questions people ask
- How does the Free Energy Principle relate to AI?
- It offers a blueprint for AI by explaining how brains learn and act. This principle could guide the development of AI systems that learn like brains, moving beyond current transformer models.
- What is Axiom and how does it differ from current AI?
- Axiom is Verses AI's new architecture using probabilistic beliefs and message passing. It aims to build agents that learn like brains, carrying explicit uncertainty rather than just predicting the next token.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Craig S. Smith.