AI's Impact on Flow State

How to measure AI developer productivity in 2025 | Nicole Forsgren

Topic 3 of 18

Published Oct 19, 2025 · Clip 8:37–12:03 (3:25)

The host shares personal experience with flow state in coding and notes AI's potential to disrupt it via interruptions from agents and code review. Nicole frames DevEx around flow state, cognitive load, and feedback loops, noting AI shifts coding from line-by-line to prompting and reviewing, which can interrupt but also enable flow through toolchains. She describes senior engineers using AI agents in parallel for tasks, maintaining high-level flow on goals while machines handle details. This allows systematic upfront planning over ad-hoc coding, yielding production-ready code faster.