Onboarding is a door. Upboarding is the bloodstream.
In May 2026, the two loudest voices in AI quietly changed their tune. The executives who had spent a year warning that AI would erase white-collar work now say the disruption they feared hasn’t arrived — that automation may expand the work people do rather than delete it. The market read it as a walk-back. We read it as a concession: the debate has moved onto the only ground that ever mattered.
Because the honest answer to “will jobs go away?” was always “wrong question.” Jobs are bundles of tasks. AI automates tasks, which recomposes jobs rather than deleting them wholesale. The headcount may even rise. What changes — violently, unevenly, continuously — is the composition of the work. And once you see that, the interesting question isn’t how many jobs survive. It’s how an organization keeps its people matched to work that keeps moving.
Onboarding solved a problem that no longer exists
Onboarding moves a person from Outside to Inside — once. Documentation, procedural ramp, the first ninety days. AI has crushed the cost of all of it. But cheapness isn’t the point. The point is that onboarding was built for a world where roles sat still. When a role sat still, one good crossing was enough.
Roles don’t sit still anymore. AI eats the lower rungs of everyone’s work continuously, so the floor under every role keeps rising. A one-time crossing into a fixed role is almost beside the point when the role won’t stay fixed.
The value didn’t disappear when onboarding commoditized. It moved to the thing that never ends.
Upboarding is the thing that never ends
We call it Upboarding: continuous Inside-to-Inside movement that elevates people into the higher-value work each change opens up. Not a one-time ramp. Not a training program with a finish line. A permanent internal circulation — toward judgment, toward orchestrating the machine rather than competing with it, toward the novel problems the floor keeps exposing.
The optimists and the pessimists have, without noticing, converged on the same mechanism: automate most of a job, and people do what’s left. Fine. Upboarding is the function that decides what “what’s left” becomes — and keeps deciding, as the floor keeps rising.
Why this is an operating-model problem, not an HR program
Here is the part that matters for whoever is writing the check. Whether AI displacement becomes upboarding (people move up the value chain) or just offboarding (people move out) is not a fact about the technology. The same AI capability, dropped into two different operating models, produces elevation in one and layoffs in the other.
So “jobs versus no jobs” was never decided by AI. It’s decided by the architecture you run AI inside. That’s why Upboarding lives in the operating model — as MOM 303, Upboarding & Workflow Modernization — and not in a training budget. A program has an end state. An operating model runs forever. Which is exactly how long the work will keep changing.
