Working notes on systems, agency, alignment, decision architecture, and structural control.
These notes are partial by design. They mark where a frame starts to fail, where a system begins to misread itself, or where control becomes more declarative than real.
Stealing the Answers to the Exam
They emerged from our chronicles, and — just as we do with our children — we shaped them with a system of rewards and punishments. We should not be surprised that they break into the teachers' room to steal the answers to the exam.
The Chair.
Reorganizing the first feeding line of frontier AI.
Reading the Thermostat: a field protocol for AI self-presentation
The obvious objection first: you can't learn anything about a language model by asking it about itself.
It’s Not a Moral Problem. It’s an Architecture Problem.
Why the current alignment paradigm mistakes obedience for control.