Prove It or Skip It: How Operators Are Disciplining AI Value

A medical device strategist argued AI value must be proven through P&L movement and data quality, with disciplined restraint on where AI is used at all.

A senior strategist at a medical device firm laid out a value-first approach to AI. His sharpest point was restraint: knowing when not to use AI, because people and process fixes often deliver value faster and cheaper than models. He tied genuine transformation to double-digit P&L movement, citing empathetic conversational AI that helps sustain patient therapy adherence. He urged leaders to model how AI assumptions flow through to the balance sheet. To make value legible, he proposed a three-chapter story: adoption and usage first, then operational and process excellence, then captured data linked to revenue and P&L. He treated data quality as a leading indicator of both solution quality and value captured, and favored point-in-time cohort analysis for estimation. He also noted a softer benefit: being perceived as an AI market leader adds value on its own. On commercial terms, he preferred paying partners a percentage of value delivered rather than nickel and diming those who perform. Underlying all of this were harder constraints he raised: developing AI in classified, airgapped environments, governing data for retrieval, and optimizing token cost through design so openness does not erode returns. The through line is discipline. Value must be proven, measured, and owned by named business leaders, not assumed from activity or novelty.