Secrets

I used to think of important secrets in the world as something that people, well, keep secret. They hide it, erase any evidence of it, and chastise people who speak about it. But there’s another type of secret: something that’s widely spoken about, is public, but no one truly understands or appreciates. I have one such secret: Llama.

Llama is a culmination of several beliefs that we’ve developed over time. These are those beliefs in increasing order of specificity.

Just like we have valuable public physical infrastructure like roads and bridges, we need valuable public digital infrastructure.

Smart contracts built on blockchains like Ethereum are going to be one of the most important pieces of digital infrastructure. Ethereum is a decentralized backend for smart contract applications built on it.

Smart contracts are going to continually increase in quantity and complexity.

While some smart contracts will be simple and immutable, many will be expressive and upgradeable.

Just like creating and maintaining public physical infrastructure requires effective governance, creating and maintaining public digital infrastructure will require effective governance and access control.

Governance and access control should be role-based. Each stakeholder should have the power to perform the function they are best suited to do, rather than any one stakeholder having permissions to do everything.

Protocols should decentralize through access control rather than tokens.

Onchain actions should be atomic: users should be able to call a specific function on a target contract. This allows actions to be easily verified, rather than proposal payloads that allow arbitrary execution.

Multi-signature wallets are easy to spin up but are centralized and inflexible. Token voting is more scalable but prone to apathy and 51% attacks. Delegation increases voter participation but leads to protocols run by politicians. Each of these systems are useful for specific things and they should complement each other. We should empower core teams, contributors, and users over politicians, and speculators.

Protocols shouldn’t have just one big treasury but instead have multiple accounts with specific permissions for distinct capital allocation decisions.

A governance system like this should be open-source and permissionless. However, the go-to-market should be enterprise-first. This ensures that the product can build an excellent user experience and develop a sustainable business model. Once there is sufficient adoption, the governance system should cultivate a developer community that builds extensions and scripts that extends the system to a long-tail of use cases.

It feels like my teammates at Llama and I have a window into the future that people may appreciate a few years from now.