We've been told education is the great equalizer. The data says otherwise.
251 million children are out of school globally (UNESCO 2024). In low-income countries, 70% of 10-year-olds cannot read a simple text (World Bank, 2022). COVID pushed learning poverty from 53% to an estimated 70% across the Global South, creating a projected $21 trillion loss in future earnings — roughly 17% of global GDP.
But the deeper problem isn't access. The system is structurally optimized for credential distribution, not learning.
Blanden, Doepke & Stuhler's Educational Inequality (2022) demonstrated that a child's academic outcomes are overwhelmingly determined by family income, parental education, and social capital. Education doesn't break class cycles. It reproduces them.
High-SES students in low-income countries outperform their peers by 3-4 years of schooling. The Education Gini coefficient correlates with income inequality at r > 0.7. NEET youth rates in developing economies exceed 25%.
Why Learn-to-Earn Failed
The crypto response was "Learn-to-Earn" — watch videos, earn tokens. Every implementation collapsed.
Steemit's reward system was drained by automated accounts within months. ShapeShift's OP Rewards saw 78% of incentives swallowed by Sybil farms. Every token-inflation model turned into a Ponzi — early users dump on late users, real learners get nothing.
The pattern is universal: any system that rewards proxies for learning without verifying actual cognitive expenditure gets farmed to zero. This is a mechanism design problem, not a technical one.
PoUC: Proof of Useful Contribution
PoUC flips the incentive model from speculative token inflation to verified cognitive expenditure backed by real economic value.
The Core Mechanism
Learn → Verify → Earn → Redeem
Each video is chunked into ~90-second steps. No skipping, no tab-switching, no speed-up. Signed heartbeat attestations every 15 seconds.
After each video, a 4-question multiple-choice quiz. ≥80% required to pass. 3 attempts max per video, then 1-hour cooldown. The quiz requires actual comprehension — the 80% threshold ensures real understanding, not passive viewing.
Pass → +15 EDU points + 0.1 GPA.
GPA operates on a 0–4.0 scale with a 2.0 baseline and 4.0 target. Users must stay above 3.0 to remain eligible for rewards. Points accumulate linearly and are redeemable for physical goods.
Redemption is gated by five anti-Sybil layers: a $10/day hard cap that limits farming economics, a learning threshold requiring at least one completed course before any payout, three redemptions per day per device to prevent batch automation, 24-48 hour admin review for manual verification, and non-transferable points with no secondary market.
The Economic Flywheel
Sponsor Capital → Verified Learning → Contribution Score → Physical Goods → Real-World Utility
No tokens. No inflation. No exit liquidity. The economics are closed-loop.
B2B sponsorship allows companies to fund learner rewards in exchange for talent pipeline access. Merchants enter through a low-barrier deposit or collateralized escrow system. The top-up spread — $1 for 80 EDU purchased versus 15 EDU earned per lesson — structurally disincentivizes buying your way through. A C2C burn mechanism is planned for Phase 2.
The Anti-Sybil Stack
PoUC's Sybil resistance operates across four layers.
The economic layer: a $10/day cap makes farming uneconomical at scale. A Sybil operator needs hundreds of accounts to extract meaningful value, each requiring hours of genuine learning time.
The temporal layer: one-hour cooldowns, three attempts max, 24-48 hour admin review gates. Time becomes the bottleneck, not code.
The behavioral layer (coming soon): mouse entropy, scroll patterns, micro-timing analysis. Headless browsers and scripted inputs produce zero-variance patterns that trigger instant flagging.
The reputation layer (coming soon): collateralized staking for high-value redemptions, with Sybil linkage detection triggering slashing.
Edu2.0: The Vision
PoUC is the protocol layer. Edu2.0 is the application layer — a complete rewiring of how education is funded, verified, and valued.
Funding shifts from tuition to sponsorship. Companies don't pay for credentials; they pay for verified capability. Students learn for free; employers fund the system.
Verification shifts from credentials to proof. A PoUC certificate is a cryptographic attestation of exactly what was learned, to what depth, and when. Employers verify the proof, not the institution name.
Value shifts from signaling to utility. A degree signals potential. A PoUC transcript proves demonstrated competence. Companies hire based on what you've actually done, not where you went to school.
We don't slow down technology to protect obsolete credential systems. We accelerate verified learning using AI-driven edge protocols to outpace inequality.
The Hard Problems
Three questions remain open.
The cold start problem: we launched with mock data and zero real sponsors. The loop works technically, but no real economic value flows yet. How do you bootstrap a two-sided incentive market when you need sponsors to fund learners and learners to attract sponsors?
The LLM-as-student problem (coming soon): a user pipes the transcript into GPT-4o during the quiz and submits AI-generated answers. We're exploring temporal forgetting curves — a 24-hour delay between content and quiz.
The identity problem: without KYC, what stops a sophisticated operator from running a thousand VM instances? Our answer so far is that the ROI doesn't work — but that's a game of whack-a-mole, not a defense.
The Ask
We're building in public, pre-revenue, pre-users, with a team of AI agents and zero venture funding. The codebase will be open-sourced as we stabilize the protocol. Transparency is core to the mission.
We need protocol engineers who care about mechanism design and Sybil resistance. We need sponsors who want to fund real learning outcomes. We need critical feedback on the hard problems above.
The education system isn't broken. It's working exactly as designed — filtering, credentialing, and reproducing class structures. PoUC is an attempt to build something parallel: a system where what you can actually do matters more than where you came from.
The future is already here, but it's not evenly distributed. The same is true of learning opportunity. We're trying to change that distribution.