Data Dissolved: A Protocol for Ending Data Feudalism

Data Dissolved proposes a structural remedy to data feudalism by applying the same thermodynamic principles used to dissolve capital. The protocol meters behaviour in a physical unit — the attention‑second — and issues decaying vouchers that platforms must spend to target, rank, recommend, or train models. A single statutory lever enforces the transition: any dataset used for behavioural prediction must be released to a public data commons after 90 days, unless its holder demonstrates continuous, transparent, and consensual use. Decay kills compounding; queues replace auctions; expired value flows into the Public Algorithm Fund to train open, civic‑grade models. The result is a digital ecosystem where hoarding becomes impossible, manipulation becomes unprofitable, and childhood becomes safe by construction. The architecture is the argument.