Optimistic Execution’ of Off-chain Voting

  • Commentary
  • September 26, 2020

Aragon, a DAO infrastructure and subjective oracle provider, recently proposed a new method for off-chain voting with on-chain execution, dubbed ‘optimistic execution.’ The method is intended to address both on-chain voting’s high execution costs (as Aragon observes, on-chain voting on Ethereum can cost up to $30 to cast a vote) and the counterparty risks present in current multi-signature contract workarounds.

For the off-chain voting component, Aragon’s optimistic execution would integrate Snapshot, an application that allows token holders to conduct verifiable polls off-line and stores a record of votes cast in IPFS. To bridge the on-chain execution gap, Aragon proposes that some token holders submit the action approved by the off-chain vote, along with slashable collateral, to the project’s (Aragon) DAO, where the action can be executed on-chain after a default timelock. To prevent actors from submitting illegitimate actions for execution, the DAO would allow anyone to dispute the action during the timelock period, with a subjective oracle, such as Aragon Court, determining whether the action submitted was the same as the action approved by the off-chain vote. (Presumably as determined by using Snapshot’s voting record.) In the case of a successful dispute, the submitter’s collateral is slashed and the action aborted.

Cost of voting is among factors thought to drive down participation and any method that might positively affect participation rates has our attention. That said, Aragon’s proposal leaves us with several open questions. While optimistic execution could lower voting costs, on-chain is less distributedly paid for compared to current methods—in place of all participating stakeholders spending $30 to vote and execute an action, one stakeholder bears the risks and opportunity costs of locking a (presumably substantial) sum of forfeitable collateral. DAOs considering adopting optimistic execution may need to create incentives for members to bear such costs, or adopt practices for sharing the risk. Whether any liability for executing actions on-chain would fall onto the submitter or Aragon jurors is also worth understanding more.

Finally, the method also introduces two project independent actors, Snapshot and Aragon, into the voting and execution process. Whether Snapshot’s approach to creating a faithful voting record is indeed secure, tamper-proof, and readily accessible to jurors, and whether Aragon Court actively monitors actions submitted to DAOs at a frequency satisfactory to a project’s comfort level are all areas projects considering optimistic voting will want to examine more closely.