One engineering leader wrote about an engineer on the team whose role changed once AI arrived. He stopped leading the project. Instead every major decision went through him, and his job was to break it: what happens when the retry fires twice, when the data is stale, when the acknowledgment times out.
The AI-generated work went up. So did the scrutiny, and that mix beat either extreme.
When making gets cheap, the person who asks how it fails is the one you cannot replace.
Kay
P.S. Course 0 is where these ideas become practice: https://systemthinkinglab.ai/course-0