← Projects

Active experiment

Reliable personal systems

Experiments in LLM-assisted memory and personal operations that remain useful after the first week.

AIAutomationExperiments
PROJECT / PERS
Current stageActive experiment
What I’m learningWhere automation reduces cognitive load—and where it quietly creates a new maintenance burden.

The question

Can a personal system powered by language models become more dependable over time rather than accumulating invisible failure?

The experiment focuses on observability, graceful degradation, and making the system easy to correct when its model of reality drifts.