2 points | by willXare 14 hours ago ago
2 comments
The human approval needs to be cryptographically tied to the agent's specific state action. We approach this via verifiable identity for the agent itself and continuous monitoring. X @stuartmccomb
The loop starts with one human reviewing AI output and ends with six people reviewing the review process.
The human approval needs to be cryptographically tied to the agent's specific state action. We approach this via verifiable identity for the agent itself and continuous monitoring. X @stuartmccomb
The loop starts with one human reviewing AI output and ends with six people reviewing the review process.