Measuring AI productivity by output volume alone may hurt the whole org

blaizedsouza · x · 2026-09-25

Allen Holub argues that assessing AI's productivity gains purely by output volume—ignoring the full product development flow from idea to customer delivery—does more harm than good, to the point where an organization might be better off dropping AI entirely. He draws on Goldratt's The Goal, where factory robotic automation damaged overall throughput, as a direct analogy for AI code generators. A single output-volume metric is worthless on its own; you have to consider the entire system.

Related event: Veteran engineer warns against measuring AI productivity by output volume(2 posts)→

Original post →

More from Companies & People

Companies & People channel →