Is “AgentOps” becoming the next DevOps?

As AI agents move into production, do we need a dedicated operational layer for monitoring, debugging, and reliability? Or are logs and dashboards enough?

Curious how others see this evolving.