Ways to Discover a Good Idea: Six Laws of Real Cognitive Work
Class session walking the cohort through the Six Laws of AI-Era Software Engineering and the BIG Tools toolkit that instantiates them — one Known-Knowns gate plus seven tools. Co-taught with Chris Moore and Carl Moore.
Abstract
Most "AI productivity" framings collapse into the same shape: AI generates, human reviews. That arrangement breaks under load — generation outpaces evaluation, and evaluation capacity degrades faster than generation speed improves. This is the Evaluation Crisis Spiral, and it shows up in education, code review, content moderation, and any domain where output volume scales faster than judgment.
BIG Tools inverts the arrangement. It treats AI as a verifier of human-generated work, not a substitute for it. One Known-Knowns gate shapes every session around what the user already understands in head, gut, and heart; seven tools then operate any-order on that context, each governing a distinct cognitive move. The Tracker accumulates provenance so the trail through the work is itself a deliverable.
This session walks the cohort through the Six Laws underneath the architecture (context as the universal bottleneck, human judgment as the integration layer, architecture over model selection, infrastructure built to delete, orchestration as the new core skill, and the orthogonality of speed and knowledge) and shows how each law shaped a specific design decision in the toolkit.
What we covered
- The Six Laws of AI-Era Software Engineering — what they are, where they came from, and why each one is observable across multiple substrates.
- BIG Tools as instantiation: the gate-as-priority pattern, the seven any-order tools, the Tracker spine.
- Live walkthrough of how the toolkit handles a real workshop session.
- The Evaluation Crisis Spiral — why "AI generates, human reviews" collapses, and what the alternative looks like in practice.
Materials
- Slides: pitch.com/v/brown-practicum-bn6zhb
- Toolkit: bigtools.dev — canonical content, recipes, and the working application. The full Six Laws blog post will land on bigtools.dev/blog.
Class context
This is a session in the Embodied Brain Technology Practicum, a cross-institutional class I co-teach at Brown University with Chris Moore (Carney Institute, cognitive neuroscience) and Carl Moore (BIG Ideas methodology, group facilitation). The cohort is twenty-four students from Brown, Ben-Gurion, MIT, Rochester, and CMU. BIG Tools is the architecture I built as the software-architecture layer for the class; the pedagogy half is Chris and Carl's. The Six Laws synthesis, the toolkit implementation, and the architectural choices behind both are mine.