Appendix E
Illustrative Scenarios
Events are the ephemera of history; they pass across its stage like fireflies, hardly glimpsed before they settle back into darkness and oblivion.
Structure is one thing; mechanism is another. The scenarios that follow are not predictions but illustrations, sketches of what the coordination substrate might look like if the framework holds. They share a common architecture: verification determines autonomy, consequence determines who remains in the loop, and enforcement operates through collateral rather than jurisdiction.
The strangeness is organizational rather than technological. Every mechanism described here follows from cost structures that already exist. The scenarios are offered to make the framework tangible, not to forecast specific implementations.
The Infrastructure Commitment
A solar developer is building a large photovoltaic installation in the Atacama Desert. The project is ordinary. What is new is the coordination substrate.
The cast is minimal. A developer agent authorized to negotiate within policy bounds. A supplier agent representing the panel manufacturer. A lending agent providing milestone-based disbursements. An insurance agent pricing coverage in real time. An attestation agent synthesizing satellite imagery, port telemetry, and sensor logs into signed milestone confirmations. Three humans exist in the background: two engineers and a CFO. They do not run operations. They set risk budgets, approve exceptions, and absorb liability that cannot be bonded away.
Every material commitment runs through a covenant structure. Each agent has cryptographically bound signing authority limited by category, amount, and time. Performance is collateralized. Each party posts a bond sized to the downside it can impose. Milestones release funds only when the attestation agent signs an evidence bundle. Large releases include a challenge window during which any counterparty may stake a challenger bond and submit contrary evidence. Losing challengers get slashed. Proven fraud slashes the attester and reverses releases where possible.
This is not trustlessness but enforceability purchased with collateral and evidence.
Scene 1: Cheap verification, fast selection.
A shipment of panels arrives at port ahead of schedule. The supplier requests milestone release.
The attestation agent pulls the standard evidence set: bill of lading, container IDs, seal integrity, vessel position, unload logs, serial number samples matched to manufacturing records.
The evidence almost clears. Two feeds disagree. The shipping agent's manifest asserts one count. The port's seal verification feed reports mismatches.
The covenant automatically moves the milestone into challengeable state. Payment pauses. The supplier may post an expedite bond to keep installation crews on schedule. Any counterparty may challenge with evidence but must stake a challenger bond.
A competing attestation firm's agent posts a challenger bond and submits a claim of substitution fraud.
The attestation agent does not defend itself with rhetoric. It widens the evidence set. Higher-resolution satellite imagery. Port crane telemetry. A second IoT seal feed. A request for robotic serial number verification.
Within hours, the evidence converges. The mismatch is real but not theft. Seals were replaced by port authority due to damage during offload. Contents match the sample.
The covenant resolves. The milestone payment releases. The supplier pays a documentation inconsistency penalty. The challenger bond is partially slashed for a low-quality dispute. The attestation agent's fidelity bond is untouched.
No one trusted anyone. The system priced the dispute, bought more verification, and moved on.
Scene 2: Term structure in action.
The CFO sees a different notification. Bitcoin has appreciated since financing closed. The project's overcollateralization ratio has drifted above target. Excess collateral is safety but also opportunity cost.
The developer agent proposes refinancing the time profile, borrowing shorter-duration funds against the excess collateral to accelerate construction.
It queries the lending mesh for an 18-month facility. Several capital agents respond with rates clustered around a published Bitcoin-secured capital curve. The curve is a market convention for this facility, not a universal outside option. It reflects the available mining and lending opportunities, custody and liquidity costs, collateral terms, duration preference, and counterparty risk of the participating venues.
The only human step appears here, exactly where the framework predicts.
The CFO is not needed to compute the refinancing. The developer agent can do that. The CFO is needed to accept the tail risks that cannot be perfectly verified. Is the revised acceleration schedule operationally feasible? Is the increased execution risk worth earlier revenue? If something breaks, who absorbs blame?
The CFO approves with one constraint: autonomy for installation milestones must remain supervised until defect rates trend lower. The covenant updates. The schedule pulls forward.
The curve that prices this facility is not purely financial. The lending mesh references feasible alternative deployments of the same capital, including computational workloads that can arbitrage grid congestion at specific sites. Factor Prime appears here as an inspectable opportunity-cost comparison rather than a universal law of conversion.
Scene 3: A failure that forces the system to show its teeth.
An installation crew reports anomalous thermal signatures during commissioning. Panels are producing within tolerance, but infrared imaging shows patterns consistent with microfractures.
The attestation agent flags the anomaly but cannot determine root cause from remote sensing. The defect is not visible to satellites. It manifested only after installation and thermal cycling.
The developer agent initiates a formal quality dispute. The covenant responds. The next milestone payment enters challengeable state. The supplier is notified. The insurance agent recalculates premium in real time, adding basis points until root cause is determined. The autonomy dial for that installation zone snaps from delegated to supervised, increasing human inspection sampling.
The supplier responds. The defect pattern matches a known failure mode in panels manufactured during a specific window. A temperature excursion during lamination created latent stress concentrations.
The resolution path activates. The supplier posts a remediation bond. The supplier's manufacturing insurer assumes liability under warranty terms. A 2-of-3 attestation quorum confirms scope. Affected panels are flagged for replacement.
No litigation. No discovery. No jurisdictional wrangling between Chilean developer, Chinese manufacturer, and Singaporean insurer. The covenant structure routed consequence to the parties obligated to bear it, verified scope through a quorum mechanism, and released funds only upon attested remediation.
The Clinical Decision
A hospital runs an agentic triage and care-pathway system. Capability is high. The binding constraint is consequence.
The hospital's insurer does not price AI as a monolith. It prices autonomy levels for specific decision classes.
A patient arrives with symptoms that could be sepsis or a benign viral flare. The triage agent assigns probability distributions across deterioration, unnecessary treatment, and intermediate trajectories requiring monitoring.
Verification is expensive. Ground truth unfolds over days. Some counterfactuals never become observable. If the patient receives aggressive treatment and does well, was the treatment necessary? The data cannot answer.
The insurer quotes autonomy like a live market.
Three packages appear, ranked by total cost:
Delegated: Agent acts without physician signoff. Hospital posts a larger performance bond. Premium highest. Strict logging constraints.
Supervised: Physician co-sign required within a window. Premium moderate. Verification requirement lower. Physician time cost incurred.
Assistive: Agent recommends only. Physician owns the decision. Premium lowest. Operational throughput lowest.
The selection among packages is not an actuarial optimization performed in real time on each patient encounter. It is a policy decision: the hospital and its insurer agree in advance which decision classes fall under which regime, based on the V/C characteristics of the clinical domain. Cardiac emergencies with cheap verification (diagnostic ECG, confirmatory biomarkers, standardized protocols) fall under delegated or supervised regimes. Sepsis with expensive verification (ground truth unfolds over days, counterfactuals are unobservable) defaults to assisted or supervised. The triage agent routes patients to the pre-agreed regime for their decision class rather than computing bespoke insurance optimization per encounter.
One package is unavailable regardless. State law requires a physician signature for high-acuity interventions. The bond market cannot route around this constraint. Some constraints are not bondable.
The physician appears as a liability sink.
The senior physician is paged. Not because the triage agent cannot reason but because the system needs a party who can absorb consequence when verification is slow and outcomes are irreversible. And because the law requires it.
The physician sees a one-screen brief: the proposed pathway with confidence intervals, available autonomy packages with cost breakdowns, recommended verification purchases that would reduce premium, and a risk delta showing how each purchase shifts the distribution.
The physician chooses a hybrid. Supervised regime for the initial decision. Purchases rapid biomarker assays and continuous monitoring. Requires a 2-of-3 clinical model quorum for the first hours. Sets a conditional autonomy rule: if biomarkers normalize within a window, the triage agent may execute discharge preparation without additional signoff.
Liability shifts. Premium adjusts. The covenant records the authority transfer.
The autonomy dial moves.
Biomarkers normalize. The conditional rule activates.
The triage agent now has delegated authority for discharge preparation. It orders final labs, schedules follow-up, generates discharge instructions, notifies pharmacy.
No human reviews these actions. The verification envelope held. Autonomy expanded because evidence supported expansion and because discharge preparation falls outside the high-acuity classification.
The attestation agent logs every state transition. If the outcome is bad, the logs determine whether protocols were followed, thresholds respected, covenants honored.
A tail event shows the membrane.
A different patient takes a sudden turn. The triage agent flags a cardiac emergency. Time to intervention is critical.
The autonomy dial snaps to maximum for this decision class. The agent activates the cath lab team, pre-positions medications, notifies cardiology, begins prep protocols.
No human signoff required for preparatory actions. The V/C ratio is favorable. Cardiac verification is cheap. The ECG pattern is diagnostic, biomarkers confirmatory, protocol standardized. Premium for delegated cardiac prep is a fraction of premium for delegated sepsis response, precisely because verification is cheaper and protocols more standardized.
The actual intervention still requires a physician. The patient is in the cath lab within minutes, with the cardiologist scrubbing in. The outcome is good.
The Cross-Border Procurement
The first two scenarios operate within a single jurisdiction. This one does not. The conclusion depends on the distinction.
CargoMind operates autonomous procurement agents that source electronic components for Western manufacturers from suppliers across East Asia. Each agent manages $50-200 million in annual procurement volume, negotiating prices, coordinating logistics, verifying quality, and settling payments across a web of suppliers in Shenzhen, Taipei, Seoul, and Singapore.
The complexity is irreducible. A single procurement batch involves forty or more suppliers across six or more jurisdictions. Each supplier presents counterparty risk; each jurisdiction presents regulatory exposure; each component presents quality variance. Payment terms span prepayment to ninety-day net. Currency exposures include USD, RMB, TWD, KRW, and SGD. Compliance obligations shift as export controls tighten.
Human procurement teams cannot optimize across this surface in real time while maintaining the responsiveness that just-in-time manufacturing demands. The agents can.
The V/C gradient across procurement.
Procurement decisions distribute unevenly across the verification surface.
Low-value, low-verification-cost: Standard component reorders from established suppliers. Price falls within historical bands. Quality falls within specification. Delivery falls within contracted windows. The house model handles these without external inference. Volume: seventy percent of transactions, forty percent of value.
High-value, low-verification-cost: Price arbitrage on commodity components. Multiple suppliers offer interchangeable parts. The agent optimizes across price, delivery timing, and payment terms. Verification cost is low because the components are fungible and quality is contractually guaranteed. Volume: fifteen percent of transactions, twenty-five percent of value.
High-value, high-verification-cost: New supplier qualification. A Shenzhen manufacturer offers capacitors at fifteen percent below current suppliers. The saving over annual volume is three million dollars. But verification cost is high: site audits, sample testing, reference checks, financial due diligence. The payoff justifies the cost, but the agent cannot complete the verification autonomously. Physical inspection requires human eyes. Legal due diligence requires licensed professionals. The agent routes to specialized verification networks, paying for attestations that reduce uncertainty. Volume: ten percent of transactions, twenty percent of value.
High-value, very-high-verification-cost: Exception handling. A supplier defaults mid-shipment. Components are stuck in customs in Hong Kong. A critical batch fails quality control on arrival. The verification cost for determining fault, recovering losses, and finding replacement supply exceeds what any parametric system can adjudicate. These route to humans—claims adjusters, lawyers, relationship managers. Volume: five percent of transactions, fifteen percent of value.
The actuation bottleneck binds at the high-C boundary. The agent can process documents, optimize logistics, and execute payments autonomously. It cannot inspect a factory floor, depose a defaulting supplier, or testify in arbitration.
The settlement problem.
CargoMind's agents need to move money. Not metaphorical value, but actual currency, settling actual obligations with actual counterparties across actual jurisdictions.
The annual flow is one hundred fifty million dollars. The counterparty graph spans:
- Twelve suppliers in mainland China, holding RMB accounts at state-controlled banks
- Eight suppliers in Taiwan, holding TWD accounts and increasingly excluded from mainland banking channels
- Six suppliers in South Korea, holding KRW accounts with correspondent banking access to the United States
- Four suppliers in Singapore, multi-currency but subject to licensing requirements for large flows
- Three Western manufacturers as customers, holding USD accounts in the United States banking system
No single banking relationship spans this graph.
Correspondent banking, the legacy infrastructure, settles in two to five business days and costs one to three percent in fees and foreign exchange spread. For a ten-million-dollar component order with seven-day delivery, waiting five days for payment confirmation is operationally absurd. The supplier will not ship. The manufacturer's production line stops. The cost of delay dwarfs the transaction fees saved.
CargoMind's agents require settlement finality in hours rather than days, access from any jurisdiction in the counterparty graph, fees proportional to transaction value rather than correspondent banking minimums, and programmability (automatic release upon attestation conditions).
The elimination logic.
Stablecoins offer speed but embed custodial risk. USDC's reserve bank can freeze accounts. Tether's opacity creates uncertainty that sophisticated counterparties price into every transaction. A Shenzhen supplier accepting USDC payment is trusting Circle's banking relationship, which is ultimately a bet on continued United States regulatory tolerance.
Central bank digital currency rails are worse. They are the state's payment system, which means the state's jurisdiction, which means the state's ability to freeze, seize, and surveil. An agent operating across United States-China tensions cannot hold treasury in either state's instrument without accepting the risk that geopolitical friction freezes its operating capital.
For this scenario, the elimination logic selects Bitcoin. The selection is contingent on CargoMind's stated threat model and counterparties; another deployment could rationally select stablecoin, bank escrow, legal bond, internal reserve, or another rail.
Bearer settlement without an asset issuer: self-custodied Bitcoin reduces issuer and administrative-revocation dependencies. Key custody, transaction confirmation, software, fee markets, and exchange access remain dependencies.
Resistance to account-level administrative freezes: self-custodied Bitcoin has a different seizure surface from a bank or issuer account. Coercion, key compromise, custody failure, legal process, and endpoint controls can still prevent use or recovery.
Conditional release: multisignature, hash locks, and time locks can encode some rail-level conditions. They do not determine whether an off-chain promise was fulfilled or who had legitimate authority to define the predicate.
Liquidity: the scenario assumes sufficient Bitcoin and fiat-conversion liquidity for its transaction sizes. That assumption requires measurement at execution time and is not a universal comparison with every alternative rail.
The selection is not a claim of trustlessness. Bitcoin does not define CargoMind's promises, establish off-chain facts, supply adjudication, prevent key theft, prove external unencumbrance, or eliminate legal and exchange dependencies. On-chain analytics can also make payment flows traceable; Appendix F examines that capture surface.
For CargoMind's purposes, traceability is acceptable (the transactions are legitimate procurement settlements, not illicit flows), while seizure resistance is essential (the risk is geopolitical freezing, not criminal prosecution).
Bitcoin also requires exchange access for fiat conversion at the endpoints where suppliers need local currency. Exchanges are firmly jurisdictional—licensed, regulated, and subject to the same freezing risk that the elimination logic was designed to avoid. CargoMind's agents must maintain exchange relationships in multiple jurisdictions to ensure that no single jurisdiction's action can cut off fiat conversion entirely. The seizure resistance of the settlement layer does not eliminate jurisdictional dependency at the fiat conversion layer; it reduces it from a single point of failure to a diversifiable risk.
For illustration, assume direct Lightning routing fees equal 0.1 percent of one hundred fifty million dollars in annual settlement volume. That assumption is not an empirical forecast. A deployment model must separately estimate routing availability, channel-management overhead, and the opportunity cost of pre-funded liquidity; the locked Bitcoin does not itself produce yield.
For large individual transactions, the practical cost may include splitting payments across routes or maintaining dedicated high-capacity channels. The scenario assumes that the measured all-in cost remains below CargoMind's measured correspondent-banking alternative. The conclusion must be recomputed from current venue, liquidity, conversion, and capital-cost data rather than inherited from the illustrative percentages.
The security architecture.
CargoMind's agents hold treasury. At any moment, five to fifteen million dollars sits in Bitcoin across multisig wallets: operating reserves for upcoming payments, collateral posted with suppliers, buffer against settlement timing mismatches.
This treasury is the agent's operational capacity. If the treasury is seized, the agent cannot pay suppliers, cannot receive payments, cannot operate. The treasury must resist seizure.
The threat model is not thieves. The threat model is states.
In October 2024, the United States Department of Commerce expanded export controls on semiconductor manufacturing equipment destined for China. In response, Chinese authorities began scrutinizing financial flows associated with component procurement that might circumvent the controls. A CargoMind agent that held treasury in a Chinese bank would be exposed: the assets could be frozen pending investigation, the account could be closed, the funds could be seized.
The symmetric risk applies to United States jurisdiction. An agent that held treasury in a United States bank while facilitating procurement from suppliers that might be re-exporting controlled components would be exposed to OFAC sanctions, account freezes, civil forfeiture.
The agent cannot hold treasury in either jurisdiction's banking system without accepting the risk that the jurisdiction's politics, not the agent's conduct, triggers asset seizure.
Bitcoin's proof-of-work security architecture protects confirmation history against reorganization in this scenario. It does not protect CargoMind's keys or prevent legal seizure at the surrounding institutions.
An adversary seeking to impair the treasury could:
Obtain the private keys, which are distributed across hardware security modules in three jurisdictions with three-of-five multisig requiring coordination across Singapore, Switzerland, and the Cayman Islands.
Acquire enough hash power to censor outgoing transactions or reorganize recent confirmations. Hash power alone cannot forge the multisignature required to spend CargoMind's existing UTXOs.
Compel every jurisdiction holding a key shard to cooperate simultaneously, which requires alignment between Singapore, Switzerland, and the Cayman Islands that no current geopolitical configuration provides.
The network supplies a shared confirmation and censorship-resistance profile funded by issuance and transaction fees. CargoMind still bears key-custody, software, fee-market, legal, exchange, and endpoint risks. This rail-specific defense does not create a general floor beneath the treasury's security.
The Shenzhen Freeze.
On March 14, 2027, Chinese authorities announce enhanced scrutiny of component exports to Western manufacturers suspected of defense applications. CargoMind's agents have twelve million dollars in outstanding commitments to Shenzhen suppliers and eight million dollars in components in transit.
At 6:47 AM Singapore time, Supplier 23, a capacitor manufacturer in Shenzhen, notifies CargoMind that its bank account has been frozen pending investigation. Supplier 23 cannot receive payment. Supplier 23 cannot ship the 2.3 million dollars in components CargoMind has ordered and prepaid.
At 7:12 AM, Supplier 31, a connector manufacturer, reports the same.
By 9:00 AM, five suppliers representing 7.1 million dollars in outstanding orders are frozen.
CargoMind's agents execute their exception protocol.
Treasury segregation: The agents immediately move uncommitted Bitcoin from Shenzhen-counterparty payment channels to cold storage. 4.2 million dollars in reserves that were staged for upcoming payments are now isolated from the frozen suppliers.
Counterparty attestation: The agents query each frozen supplier's status through the attestation network. Can the supplier still ship if payment routes through a different entity? Can the supplier refund prepayments through alternative channels? The suppliers' responses, signed attestations, become evidence for downstream adjudication.
Alternative sourcing: For components where Shenzhen suppliers were primary, the agents immediately query Taipei and Seoul alternatives. Price is eight to fifteen percent higher. Delivery is three to five days longer. Quality attestations are available from the same verification networks. The agents calculate: the cost of alternative sourcing—approximately four hundred thousand dollars additional—is less than the cost of production line stoppage at 2.1 million dollars per day for the downstream manufacturer. The agents reroute.
Payment recovery: For prepayments to frozen suppliers, the agents initiate the claims process. The suppliers' collateral, posted in a shared pool governed by a consortium of trade agents, is the first recovery layer.
The entire response—treasury segregation, attestation collection, alternative sourcing, claims initiation—completes in four hours. No human at CargoMind made a decision. The agents operated within their authorization envelope, executing protocols designed for this class of event.
Liability concentration at the membrane.
The downstream manufacturer, a German automotive electronics firm, does not know Supplier 23 exists. The manufacturer contracted with CargoMind for procurement services: East Asian components, quality-assured, delivery-guaranteed.
When components are delayed, the manufacturer's claim is against CargoMind.
CargoMind cannot effectively litigate upstream. Supplier 23 is in Shenzhen, subject to Chinese courts, which will not find for a Western procurement agent against a Chinese manufacturer frozen by Chinese authorities for national security reasons. The legal cost of pursuing recovery would exceed the claim value. The probability of success is near zero. The time horizon is years.
CargoMind becomes the liability sink because market selection makes any alternative configuration unstable. A procurement service that required manufacturers to sue individual suppliers for each failure would lose all customers to competitors that absorbed claims and handled recovery internally. A procurement service that disclaimed liability for supplier defaults would be pricing itself out of the market—manufacturers would pay premium for reliability, and competitors would capture that premium by accepting liability.
The liability concentrates at the membrane, the boundary between the manufacturer and the supply chain, because that is where the information asymmetry sits and where the bargaining power resolves.
The consortium as alternative liability architecture.
The collateral pool is not a CargoMind balance sheet line. It is governed by a consortium of trade agents that participate in the East Asian component market.
The consortium holds two hundred million dollars in staked assets, contributed by 340 agents across 28 procurement platforms. Governance is proportional to stake. The consortium's function is narrow: hold collateral, process parametric claims, fund arbitration for disputes that exceed parametric resolution.
The arbitration mechanism is faster than the ICC and cheaper than litigation. Disputes route to a panel of three arbitrators—one selected by claimant, one by respondent, one by the consortium—who rule within 72 hours based on attestation evidence. The arbitrators are procurement professionals with domain expertise, not lawyers. Their decisions are final within the consortium's jurisdiction. Losing parties' collateral pays winning parties automatically.
This is the network state in embryo. The consortium has no territory, no diplomatic recognition, no army. It has assets that can be attached, penalties that can be imposed, losses that can be recovered. It provides the liability architecture that enables agents to operate across jurisdictions where no single court has authority.
The precedent is medieval: the lex mercatoria, the Maghribi coalition, the piepowder courts. What required geographic proximity for eleventh-century merchants, cryptographic verification provides for agents.
The generalization.
The procurement scenario demonstrates the thesis in operational terms.
At sites where the same constrained power and capital can feasibly support either mining or inference, the operator compares the two net returns after hardware, switching, financing, and operating costs. CargoMind's spot pool uses that comparator only for those eligible sites; other inference supply follows other opportunity sets.
The V/C ratio orders the deployment sequence. Low-V/low-C decisions run on embedded sensors and house models. High-V/high-C decisions—new supplier qualification, exception handling, dispute resolution—require expensive verification and human involvement.
The liability sink concentrates at the membrane because the specification problem makes upstream attribution intractable. CargoMind cannot sue the inference provider whose recommendation led to a supplier selection that eventually failed. The causal chain is too long. The counterfactual is too uncertain. The adjudication cost is too high.
The settlement layer must resist the forms of unilateral impairment named in CargoMind's threat model. Bitcoin is selected here because its issuerless bearer rail reduces account-level issuer and administrative-revocation dependencies. Other deployments can prefer stablecoin, bank escrow, legal bond, or internal reserve when their risk and recourse tradeoffs fit better.
CargoMind survives the Shenzhen Freeze because the frozen supplier banks do not control CargoMind's self-custodied keys and because the multisignature policy keeps one institution from spending unilaterally. Bitcoin consensus supplies confirmation and censorship resistance for the rail. It does not cause the key separation, adjudicate the supplier dispute, or make conversion endpoints immune from law.
The network state alternative emerges from selection pressure. The consortium collateral pool and arbitration mechanism exist because no single jurisdiction's courts can efficiently adjudicate disputes spanning Shenzhen, Taipei, Seoul, Singapore, and Frankfurt. The merchants created their own court, as merchants have done since the lex mercatoria.
Bifurcation risk is not hypothetical. The Shenzhen Freeze is a partial bifurcation—not a full decoupling, but a hardening of the boundary between Western procurement and Chinese manufacturing. The agents that survive are those with treasuries and supply chain alternatives in both blocs, able to route around obstacles that would freeze a more geographically concentrated operation.
The Common Mechanism
These scenarios occur in different domains (infrastructure finance, healthcare, and cross-border procurement) with different time scales and consequence profiles. Yet the underlying mechanics are identical:
| Input | Output |
|---|---|
| Verification cost | Autonomy setting |
| Autonomy level | Insurance and bonding price |
| Attestation bundle | Release conditions |
| Disputed claim | Challengeable state and slashing |
| Enforcement cost across jurisdictions | Settlement rail selection |
| Jurisdictional adversariality | Treasury architecture |
These six mappings constitute the membrane. They are domain-agnostic. They operate whether the asset is a solar panel, a clinical decision, or a container of capacitors.
The common thread is not that AI takes over. The common thread is that when verified decisions are cheap, selection cycles at software speed, and value accrues to the membrane, the infrastructure that routes consequence, prices uncertainty, and enforces commitments.
The scenarios described here are neither utopia nor dystopia but the organizational equilibrium that emerges when the cost of verified decision-making falls below the cost of human coordination for an expanding surface of economic activity. The strangeness is not in the technology but in the incentive structures the technology makes possible.
If the framework is wrong, these scenes remain speculation. If it is right, they are mundane, the kind of ordinary operations that future participants will not bother to describe because they will be as unremarkable as wire transfers and shipping containers are today.