That's exactly the gap this series is built around, and yes, AI-driven architectures are absolutely on the list. Non-deterministic logic makes the mental model problem significantly harder, and the documentation challenge that comes with it is almost entirely unsolved. Looking forward to getting there.
klement Gunndu
Agentic AI Wizard
The Mental Model Gaps framing resonates — we run architecture audits on our own multi-agent systems and the biggest pain point is always the gap between what the system does vs. what the team thinks it does. Curious if you plan to cover AI-driven architectures where the logic itself is non-deterministic.