AI Is Not a Product. It's Infrastructure.
And that changes everything — for hospitals, for boards, and for the policymakers who shape the system around them.
Most organizations are still asking the wrong questions about AI. The dominant narrative treats it as a new software category — something to procure from a vendor, deploy in a department, and measure ROI on. That view has produced thousands of pilots and surprisingly modest impact: healthcare adopts AI at a higher rate than most sectors, yet nearly two-thirds of organizations haven't scaled it past the pilot stage.
That is not a technology problem — it is an organizational one. This keynote makes the case that AI is enterprise infrastructure, much as electricity once was for the factories of the Industrial Revolution — and that once something becomes infrastructure, the question stops being "should we adopt it?" and becomes "how do we build our future around it?"
What this keynote covers
01
From pilot to infrastructure
Why governing AI as a software line item — the way facilities manages a vendor contract, not the way a utility runs electricity — is what stalls transformation at the pilot stage.
02
The frameworks that work
Path framing, path narrating, path stretching; ADKAR for clinician adoption; and why hybrid, "ambidextrous" governance outperforms centralized or decentralized models by 41% in adoption and 26% in cost efficiency.
03
Six priorities for AI as infrastructure
Safety-centred governance, designing with front-line teams, built-in transparency, continuous monitoring, an honest narrative, and systematic AI literacy.
Further reading
Who it's for
“When something becomes infrastructure, the questions change. You no longer ask, should we adopt it? You ask, how do we build our future around it?”
Bring this keynote to your event
Tailored to your audience, in-person or virtual.