a16z Podcast | Cryptonetworks as Emerging Economies (Done Right?)

Network Architecture and Value Layers

• Layer One vs. Layer Two Distinction: Layer one prioritizes security and machine work (clearing/settlement), while layer two handles human work requiring judgment (e.g., content curation, governance), making value harder to model due to the difficulty of quantifying human input.
• Cost Dynamics: Moving from layer one to layer two reduces capital costs associated with machine infrastructure but introduces harder-to-model costs for human judgment, relying on "invisible hand" market mechanisms to determine value.
• Incentive Shifts: Value capture mechanisms evolve up the stack, shifting from securing the network (layer one) to incentivizing economic actors to provide services (layer two) cheaper or better than centralized alternatives.
• Access Token Model (Taxi Medallion): A model where a "work token" grants the right to supply service (fixed supply, creates scarcity) while a separate token serves as payment, allowing the work token value to be approximated via discounted cash flow models.
• Dual Token Risks: Separating access/work tokens from currency tokens risks replicating modern capitalism's income inequality, as early accumulators of capital concentrate as the network grows, contrasting with the crypto ethos of combining currency and capital into single assets to distribute wealth.
• Passive vs. Active Participation: While separation allows users to avoid risk (paying in fiat), it risks capital concentration; conversely, requiring work tokens for participation forces active staking, though delegation markets (e.g., hedge funds owning "medallions" hiring drivers) can still create passive capital concentration.
• Risk Distribution: Participants argue that risk is necessary to capture upside value; preventing users from taking risks (e.g., via dual token systems) may cheat them of capital appreciation, whereas a single token model ensures users participate in value creation as the network scales.
• User Sophistication: Early adopters (supply side) can tolerate risk and complexity, but as the network mainstreams, systems may need to abstract complexity for the demand side, allowing users to pay in fiat while supply sides retain exposure to the native asset's capital appreciation.

Governance Evolution and Economic Analogies

• Crypto as Emerging Economies: Crypto networks are modeled as economies with a currency, executive branch (dev team), legislative system (blockchain/consensus), and supply/demand sides, requiring similar checks against corruption and fiscal mismanagement as national economies.
• Historical Tech Value Shift: The evolution of technology value follows a pattern: Hardware (1950s-60s) → Software (1970s-80s) → Data (1990s-2000s) → Governance (current/future), as open blockchains challenge the proprietary data business model just as the Internet challenged Microsoft.
• Value Accrual Trajectory: Value may migrate up the protocol stack; while Layer 1s (computational substrate) may eventually become commoditized like cloud infrastructure (AWS), middleware and applications (consumer-facing) may capture higher scale and value.
• Marginal Cost Principle: Value accrues to the layer with the highest marginal cost; as a network grows, the cost and value of governing it increase, potentially driving token value up if the token supply remains fixed.
• Protocol Taxation Mechanisms: Networks like Decred (10%) and Zcash (20-30%) utilize implicit taxes via block rewards allocated to developer pools, functioning similarly to a government funding public infrastructure through taxation to bootstrap network value.
• Ossification vs. Fluidity: General base layers (like IP or Bitcoin) benefit from ossification and "rough consensus" for stability, whereas complex applications built on top require formal, dynamic governance to adapt to user needs and specialized expertise.

Governance Mechanisms and Risks

• Rough Consensus vs. On-Chain Governance: "Rough consensus and running code" suits general, deterministic protocols (Layer 1), while applications requiring specialization and dynamic change benefit from formal on-chain governance to enforce decisions without requiring network-wide software upgrades.
• Power Tokens: The discussion reframes "governance tokens" as "power tokens," representing the right to change the rules of the game; the value of this power increases as the network and the rules become more valuable.
• Checks and Balances: Multi-stakeholder operations (developers, miners, users) offer checks against single-token control; informal hard forks provide a backup mechanism where "governance by defection" can occur if formal processes fail.
• Voter Apathy and Delegation: On-chain governance faces low participation due to voter apathy, leading to the potential sale of voting power to professional managers, which mirrors the separation of capital and currency and risks centralizing control.
• Sybil Attacks and Identity: On-chain governance is vulnerable to identity spoofing (multiple keys), but mechanisms like account age, token history, and wallet consolidation can create "on-chain reputation" to discourage sybil attacks and reward long-term holders.
• Governance Design Flexibility: Governance systems can be designed with non-linear curves where power is amplified for aggregated tokens, allowing for sophisticated rules beyond "one token, one vote" to align incentives and prevent manipulation.
• Enforceability: Formal governance allows for a canonical group of contracts to execute decisions, avoiding the need for every end-user to download new software or manage forks, provided the token holders have aligned interests with end-users.