Re:Work #7: How expertise gets created
Between AI agents, faster development cycles, and the growing pressure to do more with leaner teams, one thing feels increasingly clear: the next challenge of knowledge work is not just how work gets done faster, but how people still become great at it.
The shift
Everyone wants senior engineers, senior lawyers, senior consultants, and senior operators.
But that skips over a harder question: How does someone become senior in the first place?
A lot of early-career work has historically been where people built judgment.
You learned by doing the messy parts. You saw patterns. You made mistakes. You watched someone more experienced make tradeoffs in real time.
That work was not always efficient. But it was formative.
If AI automates, accelerates, or abstracts away more of that work, organizations have to be much more intentional about how experience gets created.
What I’m seeing
- Organizations get very focused on which tasks AI can optimize, but less focused on which tasks quietly teach people how the work really works.
- Leaders ask how to scale expert output, but not always how to scale expert judgment.
- A growing gap between access to answers and access to the pattern recognition behind those answers.
Moving forward
Organizations need to stop treating AI adoption as only a productivity challenge.
It is also a learning design challenge.
That means:
- Preserving the messy parts of work that build judgment, even when they look inefficient.
- Making expert reasoning more visible, not just expert output more reusable.
- Creating new apprenticeship models that combine human coaching, AI support, and deliberate exposure to real tradeoffs.
AI may make some work faster. But speed alone does not create expertise.
The companies that win the next era of work will not be the ones that simply replace junior tasks with agents.
They will be the ones that redesign how people learn, how judgment is transferred, and how expertise compounds across the organization.
Less “how do we automate the work?”
More “how do we preserve and scale the path to mastery?”
Question
As AI takes on more entry-level work, what experiences do you think we need to protect, redesign, or create so people can still build real judgment?