Measuring the Wrong Things Faster Than Ever

Tokenmaxxing. Bloated PRs. Endless tiny commits. We already know how easily developer productivity metrics can be gamed. So why are token spend and AI-generated lines of code becoming the default way to measure the impact of agentic coding tools? As AI become part of everyday work for most developers, engineering leaders are under pressure to prove impact. Many organizations are falling back on familiar, easy-to-measure metrics, even when those metrics miss the real value AI creates. At the same time, AI is shifting from automation to augmentation: not just helping developers code faster, but changing how teams think, collaborate, and make decisions. That means the old models for measuring productivity and ROI may no longer apply. Join Laura Tacho for a research-backed look at how AI is reshaping developer productivity.