WHITE PAPER / OUTLINE

The operating system is part of the experiment.

Each chapter moves from question to mechanism to experiment to evidence, then leaves failure modes and open questions visible.

From the original research notebook. Earlier concepts are preserved here as history. See the current theory map and dispositions →

  1. 01

    What if intelligence is being wasted?

    FOUNDATION
  2. 02

    Models should think, not wait

    FOUNDATION
  3. 03

    Context is a scarce resource

    FOUNDATION
  4. 04

    Evidence instead of operational exhaust

    FOUNDATION
  5. 05

    Semantic editing instead of text churn

    FOUNDATION
  6. 06

    Measuring agent trajectories

    MECHANISM
  7. 07

    Durable state instead of conversation memory

    MECHANISM
  8. 08

    Scheduling without cognition

    MECHANISM
  9. 09

    Concurrent agents need resource authority

    MECHANISM
  10. 10

    Routing models by capability and constraint

    MECHANISM
  11. 11

    Bounded recovery instead of retry loops

    MECHANISM
  12. 12

    Autonomous discovery without task explosions

    MECHANISM
  13. 13

    Authorization for machine actors

    MECHANISM
  14. 14

    Disposable execution

    MECHANISM
  15. 15

    Destroy the executor before verification

    MECHANISM
  16. 16

    Independent review

    MECHANISM
  17. 17

    Controlled real-agent acceptance

    MECHANISM
  18. 18

    Everything that broke while building the system

    MECHANISM
  19. 19

    Benchmarks

    EVIDENCE
  20. 20

    Negative results

    EVIDENCE
  21. 21

    What remains unproven

    EVIDENCE
  22. 22

    Opsle Tasks

    EVIDENCE
  23. 23

    Reproducing the experiments

    EVIDENCE
  24. 24

    Future research

    EVIDENCE