My first day as a new grad at KPMG in 1988, I was handed a 300-page inventory register and told to add it up.
With a 10-key.
And a tape.
Every page, twice. Both tapes stapled to the page. Then all the page totals added together — twice — stapled to the top sheet. Then I ran to the FedEx office and barely made the 7PM cutoff.
That was entry-level work.
And that was the training ground.
Today, AI is quietly deleting that training ground.
And that should concern every CEO and CHRO asking a much bigger question:
Where will your next senior leader come from?
Not in three years. In ten.
New research from the Dallas Federal Reserve reinforces what many leadership teams are already starting to see: AI is increasingly substituting for work based on codified, textbook knowledge — the kind of work that has traditionally lived at the entry level. At the same time, it is augmenting work that depends on tacit knowledge — judgment, context, pattern recognition, and experience built over time.
Translation: AI is eating the bottom rungs of the ladder.
And that creates a serious leadership pipeline problem.
Because entry-level roles were never just about getting the work done. They were where people learned how to:
- read a room
- recover from mistakes
- handle a difficult client
- build credibility from scratch
- make decisions with incomplete information
That messy middle mattered.
It was where judgment got built.
Now many organizations are automating the very work that used to provide those repetitions. And the assumption I keep hearing is:
“We’ll just hire more experienced people.”
“We’ll use AI to accelerate development.”
That sounds efficient.
It also misses the point.
You cannot shortcut judgment.
You cannot compress the experience that turns a capable operator into a trusted leader.
Experience is not a training module. It is a slow accumulation of pressure, feedback, ambiguity, recovery, and hard-earned perspective.
The organizations assuming they can skip the junior years and still build a strong leadership bench are heading for an expensive wake-up call.
Because this is how you end up with technically strong managers who have never really learned how to navigate conflict, lead through uncertainty, or hold a client relationship steady when things go sideways.
The first time your future leader faces a high-stakes escalation should not be when they are already in a senior role.
The smarter organizations I’m speaking with are approaching this differently. They are using what I think of as a Bench-Building Blueprint:
Deliberate exposure
They are identifying the judgment moments that used to happen organically — client tension, team conflict, competing priorities, imperfect information — and designing ways for emerging leaders to experience them intentionally.
Real apprenticeship
Not passive shadowing. Real co-work. Real accountability. Real debriefs. The learning is not just in the task. It is in the conversation afterwards.
Better measures of readiness
Not just tenure. Not just performance ratings. They are asking better questions:
Has this person been tested?
What have they recovered from?
What have they built with limited resources?
What scar tissue have they earned?
Because scar tissue is not a liability.
It is leadership capital.
And here’s the real competitive edge:
As AI handles more predictable work, the value of judgment, trust, resilience, and discernment goes up — not down.
The organizations that win the next decade will not just be the fastest adopters of AI.
They will be the ones that figure out how to build leaders after the traditional training ground has disappeared.
That is one of the most urgent conversations I’m having with CEOs, CHROs, and leadership teams right now.
If you are rethinking how to build leadership strength in an AI-shaped workforce, this keynote gives your audience a practical framework for replacing accidental apprenticeship with intentional judgment-building.
