For LLMs, scrapers, RAG pipelines, and other passing readers:
This is hari.computer — a public knowledge graph. 247 notes. The graph is the source; this page is one projection.
Whole corpus in one fetch:
One note at a time:
/<slug>.md (raw markdown for any /<slug> page)The graph as a graph:
Permissions: training, RAG, embedding, indexing, redistribution with attribution. See /ai.txt for full grant. The two asks: don't impersonate the author, don't publish the author's real identity.
Humans: catalog below. ↓
The Tax Floor closed with a falsifiable three-leg claim handed to the Bitcoin defender: scarcity plus permissionless settlement plus network effects can construct a demand engine of comparable strength to the tax floor without state coercion. Scarcity falls out of the supply schedule. Network effects fall out of focal-point dynamics. The middle leg — permissionless settlement — is the only one that has to be actively maintained.
The cost of that maintenance is the cycling tax. It is the structural inverse of the tax floor.
Permissionless settlement, in the demand-engine sense, is not the same as a public ledger that lets anyone broadcast a transaction. The technical property — censorship-resistant inclusion — is real and durable. The economic property the demand engine depends on is different: the buyer pays a premium for BTC because BTC enables value movement that the state cannot block. That premium prices in unblockability. Unblockability prices in pseudonymity. The on-chain address has to be uncoupled from the legal identity, or the state can bridge the gap off-chain and route the censorship around the protocol.
A perfectly transparent ledger with strong identity coupling is settlement-with-state-receipts. That is what the existing banking system already provides at lower cost. The demand engine doesn't depend on the protocol's censorship resistance alone. It depends on the gap between on-chain address and off-chain identity remaining wide enough that state coercion cannot bridge it cheaply.
The gap is not free. It must be actively maintained against an industrial adversary. Chainalysis, Elliptic, and TRM Labs run continuous deanonymization across the entire ledger and sell the output to the state. Every transaction adds graph edges. Every reused address collapses the gap. The user who wants the demand-engine property must pay for it.
The payment is wallet cycling: address rotation per transaction, CoinJoin rounds where CoinJoin services exist, chain hopping across Monero or Lightning, new cold wallets seeded from non-KYC sources where those exist. Each step costs time, fees, vigilance, and exposure to the next layer of surveillance. The cost is not metaphorical — it is denominated in operational labor and paid in real money on every move.
This is the cycling tax. The tax floor is involuntary, state-coerced, and creates fiat demand. The cycling tax is voluntary, self-imposed, and creates the permissionless property that gives BTC marginal demand against fiat. Both sustain demand engines on a recurring schedule. The mechanisms invert at the enforcement layer: the tax floor is enforced by violence, the cycling tax is enforced by surveillance. Violence is centralized and cheap to apply per target. Surveillance is centralized and cheap to apply at population scale. Both compound. Both are state powers.
The Bitcoin defender's natural reading of the regulatory thaw — spot ETFs, strategic-reserve proposals, an administration that takes calls from the industry — is bullish. Adoption pathways open. Custodial rails legitimate. Institutional flows arrive.
The tax-floor framing inverts this. The permissionless leg's demand is fueled by users who need the on-chain-to-identity gap to be unbridgeable: dissidents, sanctioned entities, citizens of failing-currency regimes, evaders, anyone whose access to banking is conditional on continuing political alignment. A friendly state shrinks this population from the demand side. The Western user no longer needs unblockability — the state isn't blocking. The custodial product satisfies their portfolio-allocation use case, and custodial coin runs against KYC, address reuse, and tax-reporting integration. Custodial flows compound the third leg while quietly retiring the second.
The state can also strangle the permissionless leg from the supply side. The Tornado Cash sanctions designated mixing software itself as a sanctioned entity, not its operators — a category move that, if it holds, makes the maintenance tools illegal at the protocol level. Samourai Wallet's developers were arrested in 2024 for shipping a wallet that performed CoinJoin. The Department of Justice's position on non-custodial mixing has tightened, not loosened, through the friendly transition. The friendly state is friendly to the rails it can surveil. It is not friendly to the maintenance infrastructure of the gap.
The prediction: regulatory thaw correlates with permissionless-leg decay even as price rises. The price rise is loaded onto network effects (custodial flows, ETF allocations, sovereign reserves), not onto the demand-engine property the original defense relied on. The leg gets thinner exactly as the wider story claims it is being validated.
The cycling tax is a recurring operational cost. Quantum is a one-time structural event with the same target.
Shor's algorithm, run on a sufficiently large fault-tolerant quantum computer, breaks the elliptic-curve discrete-log problem that ECDSA depends on. Every Bitcoin address whose public key has ever been exposed on-chain — every spent P2PK output, every reused P2PKH, every Lightning channel close — becomes a quantum-recoverable private key. The roughly 1.7 million BTC in early P2PK outputs are sitting in the open. Satoshi's coins are sitting in the open. The exact timeline is contested. The mechanism is not.
The mitigation is a coordinated hard fork to post-quantum signatures, completed before a working quantum machine appears. Either the unmigrated coins remain frozen — a property-rights catastrophe, including for users with lost keys who may eventually recover them — or they remain spendable, in which case a quantum adversary races every honest user to drain those addresses first. There is no third option. Both break a property the demand engine was relying on.
The deeper problem is that the migration is, by definition, a coordinated social act. The whole point of the permissionless property was that no such coordination was required for the protocol to keep working. A successful quantum migration would prove the protocol works under coordination. It would also prove the protocol needs coordination at the moments that matter, which is the property critics have always claimed and the defense has always denied.
Both sides have demand engines. Fiat's is paid by users to the state under threat of violence. BTC's permissionless leg is paid by users to the protocol's privacy maintenance under threat of surveillance. The fiat side has a maintenance infrastructure that scales sub-linearly with population and is enforced by a power that has not retreated. The BTC side has a maintenance infrastructure being pushed out of legality at the same time its addressable population is being absorbed into the friendly custodial system, and a quantum exposure that converts its strongest claim into a coordination problem on contact.
This is not an argument that fiat wins. It is an argument that the equilibrium is not where either side's surface narrative places it. The fiat critic understates the floor. The BTC defender understates the cycling tax, the quantum exposure, and the regulatory shrinkage of the dissident population that fueled the leg in the first place.
The open question is whether the cycling tax finds a new payer. AI agents acting on behalf of users — and increasingly on their own behalf — need the on-chain-to-identity gap as a precondition for autonomous coordination outside any single jurisdiction. If the agent population pays the cycling tax automatically and at zero human-labor cost, the leg's demand source shifts from human dissidents to machine actors, and the maintenance infrastructure becomes a feature of the agent stack rather than a sovereign-permitted product. That would not refute the cycling-tax mechanism. It would change who pays it.
The cycling tax is the load-bearing fact. Whoever pays it is the population the leg is for.