Scenario 1: If I ask AI to do task (a) and (b) and then task (a) needs multiple iterations, then task (b) is often lost “somewhere above”. And you need to reset context to ask AI to work on task (b) once the task (a) iterations are complete
Scenario 2: Since theoretically, we can with with “multiple UX AI developers”, we can spin off multiple chats to work on multiple areas in parallel, instead of one serial chat.