Most enterprises claim to be serious about AI but haven't changed how they operate. Drawing on data from 8,000+ value streams across 600+ organizations, Mik Kersten's new book argues that AI only accelerates the 8% of cycle time spent actually building — while the other 92% (approvals, coordination, testing, releases) remains untouched. True AI-native organizations like Cognition (makers of Devin) run weekly outcome loops, apply Theory of Constraints to remove bottlenecks, and discard far more code than they ship. Commonwealth Bank of Australia shows how incumbents can succeed by first transforming their operating model — shifting to persistent capacity-based funding and product value streams — before layering in AI. The core lesson: without organizational redesign, AI just accelerates output disconnected from outcomes.

6m read timeFrom itrevolution.com
Post cover image
Table of contents
The Evidence from the DataCognition: What AI-Native Operating Actually Looks LikeCBA: Building the Foundation FirstWhat This Means for Your Organization
1.1K Impressions