The Truth About Developer Productivity in the AI Age (IT'S A TRAP)

Is developer productivity being measured all wrong in the age of AI? While artificial intelligence tools like claude code make coding with ai visibly faster, many organizations are falling into the trap of measuring the wrong things. Steve breaks down why tracking pull request throughput or ai prompt counts is often a "weak construct" that fails to predict real software engineering success. In this video, you’ll learn: - Goodhart’s Law: Why a measure ceases to be a good measure the moment it becomes a target. - DORA Metrics vs. Activity Metrics: Why the dora / accelerate metrics remain the "high watermark" for measuring software delivery. - The Trap of Output: Why shipping 60% more pull requests doesn't necessarily mean better business outcomes. - Productivity Tips for Teams: Why the team, not the individual, is the true unit of delivery. - Steve on X ➡️ https://x.com/SteveSmith_Tech Steve on LinkedIn ➡️ https://www.linkedin.com/in/stevesmithtech/ Steve on bluesky ➡️ https://bsky.app/p