Week 12
AI, authorship, and the capstone
Build a complete writing workflow in which AI improves rigor without owning the result, by running the full system on one real essay.
1 min read
In short
Build a complete writing workflow in which AI improves rigor without owning the result.
Run the full system on one real essay, with AI limited to questions, diagnosis, research, reader simulation, and local edits.
Read first
Read these modules before you start:
Then review the genre playbooks, Modules 37 to 44 .
Run the full system on one real essay
Try it
Choose one real essay of 1,500 to 3,000 words. Run the full system:
Private Thinking Grill.
Author-written Thinking Essay or adequate private synthesis.
Reader Reconstruction.
Intended Emphasis Contract.
Author-written Reader Essay.
Reader-Response Audit.
Cognitive and evidence audit.
Source-fidelity check, where relevant.
Prose and Unslop pass.
Read aloud and final proof.
Keep AI out of the author's role
During the capstone, AI may ask questions, diagnose, research with sources, simulate a reader, and locally edit approved passages. It may not write the full essay or manufacture lived examples.
You succeed when
The final claim can be traced back to private evidence and reasoning.
Reader Architecture materially shaped selection and order.
The countercase changed or confirmed the boundary.
Emphasis matches intention.
The prose remains recognisably yours.
You can explain why every major section exists.