From Skeptic to Superpower: Real‑World AI Coding Workflows That Scale | BRK229
Engineers describe scaling real-world AI coding via autonomous agent flows that break features into parallel task waves.
“within 5 to 10 years 80% of the simple code will be uh done by AI”
At Microsoft Build, engineers from energy-management company Think shared their production AI coding workflow, which uses autonomous agents to break feature work into dependency-graphed tasks executed in parallel waves. They drew parallels between current generative-AI enterprise adoption and the earlier Kubernetes/cloud-native curve, but the session is largely an anecdotal practitioner demo rather than a major industry announcement.