AI trained on data gives answers. Humans shaped by experience change the question.
That gap matters more than most people think.
The people who reframe the board with one sharp insight
An executive I worked closely with for four years had an unusual ability. When complex strategic problems landed on the table, he didn’t analyze them — he repositioned them.
When EV demand stalled, the standard ask was: what’s causing the slowdown, and when does it recover? Market curves, infrastructure gaps, policy headwinds. The usual angles.
His response cut differently. “We’re calling this a chasm — as if the problem is coming from outside us. But maybe Korean EV adoption is stalling because we haven’t actually released a car people want to buy.”
Not a gap in the market. A gap in what we were building. The premise, not the data, was wrong.
I asked AI the same question
I’ve spent years testing whether AI can replicate this kind of insight. Same question, latest models. What came back was thorough — chasm theory, adoption curves, external factors, well-organized. What never came back was a challenge to the frame itself.
At first I assumed it was a capability gap that more data would close. Then I heard neuroscientist Inha Lee describe how the hippocampus actually works.
From one experience, ten scenarios
“AI can take ten events and tell you what typically happens. What it can’t do is take one event — say, tripping and falling — and generate ten different scenarios for what to do differently next time.”
AI runs in one direction: many inputs, averaged into a pattern. The human hippocampus runs the other way: one experience, expanded into multiple simulations. And the raw material for those simulations compounds with age.
My colleague could reframe the EV question because decades of watching industries move — from the inside, across multiple roles — were firing simultaneously. Past cycles, current structure, future pipeline. Several time horizons collapsed into a single sentence.
This isn’t a data problem AI will eventually solve. It may be something else entirely.
Why experience matters more, not less
As AI gets faster at producing answers, the person who can reframe the question becomes rarer — and more valuable.
That capacity doesn’t arrive quickly. It accumulates through years of experience processed, not outsourced.
Which raises the uncomfortable question: if today’s juniors are letting AI do the thinking, what kind of senior will they be in twenty years?