# Research, Rewritten

> Using AI to Produce Knowledge You Can Trust. Deep research gives you plausible. This book gives you reliable. It wires general-purpose AI into every step of truth-seeking work and says where you can hand it to AI, where you must gate it yourself, and where nobody knows yet.

Site: https://hallieren.github.io/research-rewritten/  
Whole book in one file: https://hallieren.github.io/research-rewritten/llms-full.txt

## Chapters

- [Preface · This Book Was Put on Trial](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch-preface.md): You have probably just read a report written by AI, thought it looked better than the last one you wrote by hand, and stopped with your hand over the forward button. That one second of stopping is
- [Start Here · A Two-Hour Win](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch00-start-here.md): This chapter skips the argument and gets results first. Two hours from now you will hold an experiment number that makes your heart race, and a list that calms you back down. The two together are the
- [Chapter 1 · After Coding, Research](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch01.md): An AI research report can dazzle you within ten minutes and embarrass you after you forward it. Those are two faces of the same thing. This chapter states its true shape, and states what this book
- [Chapter 2 · A Map Stolen from Paradigm Shifts](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch02.md): The change you are watching has run the same script at least five times, and the audience's notes are still around. This chapter's job is to steal those notes back and bind them into a map that every
- [Chapter 3 · The Truth-Seeking Workflow and the Autonomy Ladder](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch03.md): This chapter's ladder. No ladder line. The ladder itself is born in this chapter. From Chapter 4 on, every chapter opens with a ladder line that tells you how high this step can safely climb right
- [Chapter 4 · Master a Field](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch04.md): This chapter's ladder. At this step AI sits stably at assistant level. Scanning, distilling, and tracking can all be handed off. The collaborator door is half open, and having AI find what the
- [Chapter 5 · Questions and Hypotheses](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch05.md): This chapter's ladder. At this step AI is stable at assistant level. Batch-generating candidate questions, candidate hypotheses, and counterexample angles can all be handed off. But this step has no
- [Chapter 6 · Turn an Idea into a Falsifiable Test Plan](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch06.md): This chapter's ladder. At this step AI is stable at assistant level. Drafting the plan and enumerating confounders can both be handed off. Collaborator level is still exploring. AI drafts a plan
- [Chapter 7 · Execution](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch07.md): This chapter's ladder. Execution is the step AI climbs highest on among the seven. Collaborator level is solid, and the door to autonomous researcher level opens a crack here, which counts as local
- [Chapter 8 · Read the Results, Catch the Errors](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch08.md): This chapter's ladder. At this step AI downgrades on purpose. Batch splitting, recomputing, pattern scanning can be handed off. Those jobs have cheap ground truth, a recomputed number comes out right
- [Chapter 9 · Delivery](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch09.md): This chapter's ladder. At the delivery step AI sits steadily at assistant level, and this is one of the steps whose price collapsed hardest in the whole book. First drafts, audience rewrites, figure
- [Chapter 10 · Red Team](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch10.md): This chapter's ladder. On the red team step AI sits steadily at assistant level, and this is the step of the seven that gets underrated hardest. Having it attack your conclusion costs almost nothing
- [Chapter 11 · Failure Modes Unique to Research](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch11.md): This chapter's ladder. This chapter cuts across all seven steps. Every failure mode grows on a step of its own, so the chapter cannot be filed under any single step. Recognizing failure modes is
- [Chapter 12 · The Verification Workflow](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch12.md): This chapter's ladder. Verification cuts across all seven steps and does not occupy one of them, so this line labels the checking task itself. Mechanical checks, citation existence, number tracing,
- [Chapter 13 · An Honest Map](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch13.md): This chapter's ladder. This chapter is outside the seven steps, so it carries no ladder line. Chapter 3, section 3.5 marks the capability water line, how high AI can stand at each step. This chapter
- [Chapter 14 · The Researcher's New Craft](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch14.md): This chapter's ladder. This chapter does not occupy one of the seven steps. It turns the whole ladder around and looks at it from the human side. The four levels are not only AI's climbing route,
- [Chapter 15 · Make It a Habit and a Capability](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch15.md): This chapter's ladder. No ladder line this time. This chapter is not one of the seven steps. The ladder measures how high AI climbs inside each process step. This chapter is about the level you have
- [Chapter 16 · Coda · How This Book Stays Current](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/chapters/ch16.md): This chapter's ladder. Not marked. The coda sits outside the seven steps.

## Templates

- [Chapter 1 templates · The "Is This Book for You" Self-Test](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch01-templates.md): How to use. Answer the seven questions below honestly, one by one, and score each by the rule it gives. Total time about one minute.
- [Chapter 2 templates · The Transfer Map](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch02-templates.md): This appendix goes with section 2.5 in the chapter (the transfer map) and section 2.6 (the three failure condition questions).
- [Chapter 3 templates · Ladder Self-Rating Sheet + Seven-Step Workflow Check Card](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch03-templates.md): This appendix is the fillable version of the two maps in Chapter 3.
- [Chapter 4 templates · Controversy Map Prompt + Paper Card + Frontier-Layer Subscription Rules + Coverage Checklist](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch04-templates.md): How to use. The four tools match the three-layer intake workflow of Chapter 4. The map layer uses Template 1, the skeleton layer Template 2, the frontier layer Template 3, and Template 4 runs before
- [Chapter 5 templates · Five-Step Question-Sharpening Prompt Set + Question-Sharpening Card](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch05-templates.md): How to use. Run the five prompts in order, and everything they produce lands on the question-sharpening card at the end. Budget one hour for the whole run. If you run well over, go back and read the
- [Chapter 6 templates · Test Plan Template + Confounder Checklist + Plan Red-Team Prompt](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch06-templates.md): This appendix is the complete fillable version of the three tools in Chapter 6.
- [Chapter 7 templates · Minimal Harness Checklist + Ledger and Resume Patterns + "Have AI Audit the Harness" Prompt](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch07-templates.md): This appendix is the complete fillable version of the three tools in Chapter 7.
- [Chapter 8 templates · Result Interrogation Checklist + Preregistered/Post-hoc Side-by-Side Report Template + Surprise-Result Red Flags](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch08-templates.md): This appendix is the complete fillable version of the three tools in Chapter 8.
- [Chapter 9 templates · Two Real Deliverables + Claims List / Technical-Report Skeleton / One-Page Memo Templates + Delivery Prompt Set](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch09-templates.md): This appendix comes in two parts. Part one holds the book case's two real deliverables, the technical-report skeleton (shaped like a workshop paper) and the CTO one-page memo from the small-model
- [Chapter 10 templates · Red-Team Dispatch Brief Template + Four-Attack-Surface Red-Team Prompt Set + Disposition Table](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch10-templates.md): This appendix is the complete fillable version of the three tools in Chapter 10.
- [Chapter 11 templates · Failure-Mode Census Sheet + AI Self-Check Prompt Set](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch11-templates.md): This appendix is the complete fillable version of the two tools in Chapter 11.
- [Chapter 12 templates · Layered Verification Workflow Card + Independent-Channel Check Prompt Set + Verification Budget Sheet](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch12-templates.md): This appendix is the complete fillable version of the three tools in Chapter 12.
- [Chapter 13 templates · The Honest Map Toolkit](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch13-templates.md): How to use. A three-piece kit, used together. First list the rows with Template A, then assign each row a status with rules card B, and last schedule the reviews by cadence table C. Goes with Chapter
- [Chapter 14 templates · Process Self-Check Sheet + Dispatch Brief Template](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch14-templates.md): Prerequisite. You have read the Chapter 14 prose. Template 1 goes with 14.3/14.8 (the craft migration list and the "my process list" exercise), Template 2 goes with 14.5 (dispatch craft).
- [Chapter 15 templates · 30-Day Adoption Plan + Three Process Questions Card + Team Metrics Starter Sheet](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch15-templates.md): This appendix is the complete fillable version of the three tools in Chapter 15.
- [Chapter 16 templates · The One-Hour-a-Month Tracking Recipe Card](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/ch16-templates.md): How to use. This card covers change in the craft of "doing research with AI" itself, not your research field (that is the job of the Chapter 4 frontier layer), and not the row-by-row re-derivation of
- [Template index](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/appendices/template-index.md): The entry point for the templates of all 16 chapters, in chapter order. Each template first appears in its own chapter and is meant to be used with the text.

## Experiment ledger

- [Experiment ledger index](https://raw.githubusercontent.com/hallieren/research-rewritten/main/docs/experiments.md): The crime scene for every number in the book. Each experiment's preregistration entered its repository before any result did. The criteria existed before the results, which is the method Chapter 6