I’ve been thinking about why the companies with the most AI training hours are often the ones with the least to show for it.
Ninety percent of companies report investing in AI. Fewer than 40% report meaningful bottom-line impact. The usual explanation is that they picked the wrong tools, or started too late, or didn’t get executive buy-in. But I keep seeing a different pattern — organizations that did everything right by the conventional playbook and still can’t convert capability into output. They bought the seats, ran the workshops, hired the consultants. And their people are still doing most of the work manually, checking boxes on AI adoption metrics while the actual leverage sits unused.
The problem isn’t adoption. It’s that the skill the economy actually needs doesn’t have a name, doesn’t have a curriculum, and doesn’t work like any workforce skill that’s come before it.
Picture a bubble. The air inside is everything AI agents can do reliably. The air outside is everything that still requires a human. The surface — that thin membrane between the two — is where the interesting work happens: deciding what to delegate, how to verify, where to intervene, when to trust. That surface is where the value concentrates. What most people miss is that when the bubble inflates, the surface area increases. Every capability jump creates more boundary to operate at, not less. More seams, more judgment calls, more decisions about where human attention creates value.
Every prior workforce skill — literacy, numeracy, computer literacy, coding — was a destination. You reached it, you had it, you moved on. But the skill of working at the surface of this bubble has no fixed destination, because the surface keeps expanding outward. You can’t learn it once. You can only learn to stay on it as it moves. And the infrastructure we’ve built to teach workforce skills assumes the target stands still.
That mismatch is the most expensive gap in the global workforce right now. This briefing is about what the skill actually is, how it decomposes into practicable components, and what it means for how you hire, structure teams, and allocate your own attention.
This briefing covers:
The five operations. Frontier operations decomposes into boundary sensing, seam design, failure model maintenance, capability forecasting, and leverage calibration — five things a person does simultaneously, not sequentially.
Why the gap compounds. A person who started building this skill six months ago isn’t six months ahead. She’s operating in a different technological era — and the distance is widening.
Who gets this and who doesn’t. Japan, South Korea, and Germany need this capacity most and are least equipped to build it. Singapore is the outlier that proves the model.
Team of One vs. Team of Five. The two organizational units emerging in frontier-capable companies — when to use each, and what connects them.
What to hire for. The interview questions that identify a frontier operator, and why traditional signals are nearly useless.
What to do Monday. Concrete implications for individual contributors, managers, executives, and policymakers.
The skill I’m describing has a name: frontier operations.
