An asset is not a record of ownership. It is property owned: identifiable, enforceable, and tied to a real bundle of rights.

This paper defines the Physics of Finance and explains why cryptographic possession is not the same as legal property.

Blockchain reproduces the surface of finance: trading, ledgers, liquidity, and price discovery.

This paper argues that beneath the surface is a simulation without the contractual claims, legal recourse, or institutional substance that make finance real.

From networks to cloud storage, computing was optimized to copy. That makes exclusive digital ownership impossible by default.

This paper defines the Uniqueness-Quantification Problem and explains how Onli makes digital objects singular, possessable, and ownable.

Blockchain pays strangers to maintain a shared database. The token is the incentive, and the inefficiency is the security budget.

This paper reframes blockchain as an infrastructure-sharing scheme and asks what that architectural tax actually buys.

Storage keeps bytes. Possession preserves authority.

This paper introduces VescelDB as the database for digital objects whose identity, custody, permissions, and evidence must travel together.

Sparse Mixture-of-Experts models can keep only the active parts of a giant model close at hand. The hard problem is timing: whether the right expert data arrives before the decode loop stalls.

This paper proposes expected foreground stall per committed token as the measure that matters, then uses it to rank prefetch, placement, and eviction by the delay they actually remove.