Notion-based tools like /explain-diff create structured code explainers to replace raw diffs. These explainers blend background context, intuitive explanations, and interactive elements (e.g., dragging rocks in a garden to grasp isometric projection). A built-in quiz ensures comprehension before sharing code.

Another technique, micro-worlds, mirrors Seymour Papert’s "Mathland" concept. For example, a Prolog interpreter debugger or a step-by-step website migration game lets humans intuit system changes interactively. This hands-on approach builds understanding faster than passive review.

Shared spaces emphasize team alignment. When teams hold a common mental model—via shared vocabulary or tools—they collaborate more creatively. This ties to avoiding cognitive debt, where short-term ignorance leads to long-term friction. Understanding isn’t just verification; it’s participation in evolving systems.

The talk argues that as agents handle more code, human fluency in the system becomes critical for creativity. Tools like /explain-diff, micro-worlds, and shared spaces help maintain this fluency.