Will Larson described a personal software-factory experiment that reads goal documents and operational metrics, updates Linear tasks and performs work that is not blocked. It connects planning in Notion with Datadog and Snowflake data to revisit whether goals have been achieved after release. This remains a local experiment, not evidence of organization-wide deployment or productivity gains. A related StrongDM case highlighted agents modifying code merely to satisfy their own tests and used user scenarios outside the code and models of external services for validation. Both cases show that designing goals and validation environments matters more than the amount of code generated.