Education and Healthcare Suck for the Same Reasons

Metrics, Management, and Goodhart’s Law

• Many criticize the mantra “if you can’t measure it, you can’t manage it” as reductive and harmful when overapplied.
• Others argue metrics are philosophically necessary: if something cannot in any way be detected, it cannot be managed.
• Several point out that complex work (software, teaching, medicine) resists simple metrics; attempts are easily gamed and can distort behavior.
• A recurring theme: metrics are useful prompts and proxies, but never sufficient on their own; “metrics-supremacy” is seen as dangerous.
• Some suggest involving frontline practitioners in choosing which metrics to optimize, rotating them regularly to reduce myopia.

Healthcare Practice, Documentation, and AI

• Multiple commenters note doctors spending more time typing than listening; record-keeping and billing workflows are seen as crowding out empathy.
• Some argue the core issue is underinvestment in people (scribes, admin support), not record-keeping itself.
• AI scribes are highlighted as one of the few current LLM uses clinicians actually like, reportedly improving visits by freeing attention for patients.
• There is disagreement on what to measure: patient-centric metrics (time to appointment, time with doctor, perceived adequacy of attention) vs. hard outcomes like mortality, which are noisy and lagging.

Education Funding, Outcomes, and Inequality

• Strong disagreement on whether “more funding” is the key fix.
• Several claim the U.S. already spends heavily per student, with flat test scores and poor outcomes in many districts, implying money is not the main constraint.
• Others push back, citing structural inequality, distribution of funds, curriculum quality, and student backgrounds; they reject framing “bad kids” as the core problem.
• Examples are given of wealthy districts with high spending but declining outcomes, and poor states with low spending but strong test performance.

Standardization, Scale, and Trust

• Some see standardization as an unavoidable response to scale; others blame deeper issues: loss of trust in professionals and fear of failure driving control systems.
• One view: both healthcare and education are distorted because payers and “customers” differ (insurers vs. patients; parents vs. children), leading institutions to optimize for third-party metrics and incentives.

Alternative Models and Role of AI

• Ideas floated: self-directed learning pods, community-funded clinics, income-linked school funding, lifelong satisfaction surveys.
• Skeptics doubt such models can scale beyond niches.
• Several note LLMs might make high-quality one-on-one tutoring widely accessible, pushing schools and doctors toward roles emphasizing character development and bedside manner rather than information delivery.