The Production Gap is our term for the chasm between an AI pilot that works and an operating model that actually delivers value. There are two gaps, not one: idea → working demo, which AI has made almost trivial to cross, and demo → changed, value-producing production, which is exactly as hard as it has always been because it is organisational. Pilots are everywhere precisely because the first gap is cheap; ROI is rare because the second is not.
The underlying error is scoping a pilot as a Proof of Technology (can the model do it?) when ROI depends on a Proof of Production (does it change the workflow, who owns the last mile, what P&L line does it move?). It sits alongside the conviction-gap — that one is about whether an idea is worth building at all; this one is about why a worthwhile, working pilot still fails to deliver value. The depth treatment is Why AI Pilots Stall Before ROI.