I, scientist with Balazs Kegl Glen Reuschling
Oct 8, 2025
Glen Reuschling, a retired electronics engineer, dives deep into the intriguing world of embodied AI and robotics. He discusses why embodied AI lags behind software-driven solutions and the philosophical implications of patterns in existence. The conversation shifts to the nature of randomness, exploring whether it's a conscious choice rather than mere chance. Glen also examines how engineered systems can achieve efficiency akin to insects through nested agency. Their dialogue challenges traditional views on agency, computation, and the design of intelligent systems.
AI Snips
Chapters
Books
Transcript
Episode notes
Hardware Perspective Drives Embodied AI View
- Glen links nested agency and embodied intelligence to his background in hardware and medical device design.
- He argues physical constraints and repair costs make embodied AI slower but crucial for true intelligence.
Patterns Arise From Dynamics Not Just Forms
- Fractals and cellular automata both produce recurring patterns but arise from different dynamics.
- Glen prefers cellular-automata-style bottom-up generation as a model for biological agency.
Novel Biological Code Emerges From Anomalies
- New biological 'code' arises when agents notice anomalies and build internal models to act on them.
- Relevance realization explains how organisms discover and instantiate novel patterns.
