02 JOURNEY
I tried many approaches, wrote a great deal of code, and encountered even more failures. MIRA kept changing because the problems kept becoming clearer.
Each stage records what I was trying to build at the time, how I approached it, what actually existed, what broke, and why the next stage began.
The past is not rewritten as though the direction had been obvious from the beginning.
Dates and states follow the repository lineage. Archived work remains evidence, not current authority.
Zero explored whether language, memory, relationships, and world events could bend later behavior. Its diary preserves both closed loops and corrections; it is historical evidence rather than the current runtime.
Read 30 Zero records →Original ECT tested how causal components reinforce or cancel, how residue ages, and how failed or shadow branches can reopen recovery paths without taking over the main flow.
Read 30 Original records →Canonical MIRA froze a smaller set of runtime invariants and evidence rules so later experiments could distinguish a changed hypothesis from a changed execution engine.
Read 30 Canonical records →The current work asks whether observation can become reusable experience without hiding the answer in an executor. Narrow successes, stopped runs, information loss, and open residuals are recorded in the same ledger.
RESEARCH NOTES
Each published note keeps the research question, hypothesis, setup, observation, interpretation, limitations, and next question apart.
Open the research notes