Human in the Loop Is Not a Plan
Designing evaluation that scales with autonomy
A book by Alpesh Nakrani
Subscribe to get the complete book. Your PDF downloads immediately.
- Pages
- 87
- Chapters
- 8
- Format
- PDF · 18.9 MB
What's in the book
Putting a human in the loop feels safe and scales terribly. The reviewer becomes the bottleneck, then a rubber stamp, then a liability. This book is about building evaluation that keeps pace with autonomy: sampling, escalation, and the small set of decisions a person should never delegate.
'A human reviews it' is where most AI systems quietly fail. What real evaluation looks like when the machine outpaces the reviewer.
Table of contents
PDF page- FMFront Matter: Human in the Loop Is Not a Plan—
- 01The Review Queue Collapse13
- 02Human in the Loop Is an Incomplete Sentence20
- 03The LOOP-SAFE Framework28
- 04Risk-Tiered Review and the Capacity Math35
- 05Rubrics, Calibration, and Disagreement44
- 06Sampling, Evals, and Continuous Learning52
- 07Autonomy Boundaries and Evaluation Strength60
- 08The Operating System for Scalable Oversight68
- ENDConclusion: Design the Loop or Remove the Claim76
- AAppendix A: Source Index79
- BAppendix B: Glossary82
The full chapters, illustrations, and reference material are included in the downloadable PDF.
Human in the Loop Is Not a Plan
The complete edition,
ready to read offline.
Get your free copy.
Subscribe to Alpesh’s book and writing updates. Unsubscribe anytime.
Your download is ready.
You’re subscribed. Your PDF download should start automatically.
Download PDF againIf it doesn’t start, use the button above.
