Bucket Foundation — the reform thesis, grounded in the evidence
This document is the bridge from diagnosis to mission. It reads the quantitative atlas (docs/EDUCATION_PROBLEMS.md) and the four structural deep-dives (docs/deep/01–04) and asks the only question that matters for a foundation: given the actual problems, where does Bucket's open-knowledge thesis measurably move the needle — and where does it not? Honesty about the second half is what makes the first half credible.
1. The diagnosis, in one page
Three measurable crises (atlas, sourced):
- Learning — 48.3% global learning poverty (Sub-Saharan Africa 86.5% vs Europe 8.6%).
- Access — 51.2M primary + 61.2M lower-secondary children out of school.
- Financing — world spends 3.6% of GDP, below the 4% floor; 92 of ~200 countries underfund.
Four structural problems the indicators can't measure (deep-dives, sourced):
- The industrial model optimizes for sorting, compliance, and credentialing — not learning. One age-batched, bell-timed, standardized machine, exported to every tier regardless of context.
- Systems teach content but not learning-to-learn — the cognitive science of how to learn is settled (spacing, retrieval), yet 84% of students reread and 72% wrongly believe massing beats spacing. Philosophy, epistemology, and agency are near-absent. The capacities the knowledge+AI age most needs are the ones schooling least teaches.
- Systems ignore the body — sleep, light, movement, and nutrition are large, well-evidenced levers on cognition (later start times, exercise, iodine ≈ 13.5 IQ points, lead) that the factory schedule structurally violates (early starts, fluorescent boxes, no movement).
- Systems cap the top while failing the bottom — age-batched pacing and teach-to-the-middle put a ceiling on students who could super-excel (SMPY: ability predicts with no right-tail plateau; acceleration g≈0.70, helps the top without harming others), even as the floor goes unmet.
The through-line: schooling is logistics for sorting a population; it is not a system for maximizing knowledge — its access, its production, its validation, or its reach for those who can go furthest.
2. The wedge: Bucket reforms the knowledge layer, not the schooling-logistics layer
Bucket is a foundation, not a ministry of education. It cannot fix per-pupil funding, teacher pay, student debt, or the access/financing crisis in low-income states — those are state-capacity problems, and pretending otherwise would be dishonest. What Bucket can own is the layer underneath all of it: how knowledge is accessed, produced, validated, and paid for. Map the levers to the problems:
| Bucket lever | Problem it attacks (from the evidence) |
|---|---|
| Free-to-read primary research (open access, the canon) | Paywalls/credentialism gating frontier knowledge; <1% of people ever read primary research; the access-to-the-boundaries problem for students who could super-excel |
| Paid-to-cite, author-routed (feed402/x402) | Publisher capture of citation value; the broken incentive that funnels research rewards to journals not authors |
| research-atlas + the metascience papers | "Who gets to do research" — funding concentration, the PhD/grant pipeline's short-termism (our own published findings) |
| The 40 research tools + the research agent | Learning-to-learn + self-directed learning at the frontier: an AI amplifier that lets a motivated person do research, not just consume it |
| The canon as an open frontier | The ceiling problem — open access to the full boundaries of knowledge so the few who can do genius work aren't capped by an age-batched curriculum |
3. What Bucket explicitly does NOT claim to solve
The K-12 funding model, teacher shortages/pay, the low-income access + financing crisis, school infrastructure, and the floor-level learning-poverty emergency are real, larger, and outside a knowledge-layer foundation's reach. Bucket's contribution is complementary, not a substitute: open knowledge is necessary but not sufficient — a reachable frontier is useless to a learner never taught to direct their own learning, and irrelevant to a child who isn't in school at all. Claiming otherwise would be the kind of overreach this whole atlas was built to avoid.
4. The honest synthesis
The industrial education system was a real democratizing achievement (87.7% global literacy) whose model is now mismatched to the knowledge age. Reforming the schooling-logistics layer is a job for states. Reforming the knowledge layer — open access to it, fair production and validation of it, and an un-capped frontier for self-directed learners and the few who can extend it — is the job Bucket is built for. Everything in this atlas is the evidence base for that focus, including the explicit boundaries of it.
Sources: `docs/EDUCATION_PROBLEMS.md` + `docs/deep/01–04` (every number traceable to World Bank / UNESCO / OECD / NCES / NSF / the peer-reviewed learning-science literature).