Governments as learning systems

Treat expertise as a function of environmental regularity and feedback speed: build government units that explicitly optimize for short feedback loops, rapid training, and error correction rather than seniority and prestige. This reframes reform as designing institutions around measurable information dynamics (how often you see consequences), not just rules or personnel changes. — If politics were reorganized around feedback‑rich learning units, many chronic policy failures could be addressed without partisan fights over values — by changing how decisions are tested and updated.