The Professional Who Was Too Good at His Job
When expertise becomes a filter instead of a doorway.
A contemporary POrable inspired by Edward de Bono's thinking methods and informed by research on expertise, cognitive fixation, and generative AI at work.
1. The Story
Arun had spent twenty years becoming the person everyone trusted.
When a project was in trouble, he found the flaw. When a proposal looked promising, he spotted the hidden risk. When an enthusiastic colleague announced a breakthrough, Arun knew which questions would expose its weaknesses.
His reputation was deserved.
He had saved his organization from expensive mistakes.
Then AI entered the workplace.
Colleagues began bringing him unfamiliar proposals: automated workflows, AI-assisted customer responses, new ways to analyze information, and processes that once required days but might now take hours.
Arun examined them carefully.
"This won't handle exceptions."
"The information could be unreliable."
"We cannot compromise quality."
Often, he was right.
Yet something unexpected began happening.
Some of the ideas he rejected returned weeks later in improved forms. Their creators had redesigned them, introduced human checks, narrowed their scope, or found a different use for the underlying principle.
Arun noticed that the original proposals had indeed been flawed.
But the possibilities behind them had not necessarily been flawed.
One afternoon, a colleague asked him a question he had never considered:
"Arun, when you find something wrong with an idea, do you also look for what might be right inside it?"
He had no immediate answer.
2. The Invisible Pattern
Arun's difficulty was not a lack of intelligence, imagination, or willingness to learn.
It was something subtler.
His experience had trained him to recognize familiar patterns with extraordinary efficiency.
That was his strength.
But the same strength sometimes led him to classify unfamiliar proposals before exploring what they might become.
He was excellent at answering:
He rarely asked:
The distinction mattered.
The first question evaluated an existing idea.
The second opened a search for alternatives.
Arun had become exceptionally good at the first without noticing how little time he gave the second.
3. The PO Interruption
At the next innovation meeting, Arun tried an experiment.
He wrote a deliberately provocative statement at the top of his notebook:
PO: The expert must help an idea become possible before deciding whether it is acceptable.
Taken literally, the statement was unreasonable.
Experts could not postpone every judgment. Some ideas involved safety, privacy, legal obligations, or serious operational risks. Such concerns required immediate attention.
But PO was not a command to accept an unreasonable proposition.
It was a device for moving thinking away from an established pattern.
Arun began exploring what the provocation might reveal.
What if an expert's first contribution to a new idea was not a verdict?
What if it was an attempt to identify the useful principle hidden inside the proposal?
What if expert judgment could be applied twice: first to help develop a possibility, and then to evaluate it?
A different role for expertise began to emerge.
4. The Moment of Movement
A colleague proposed using AI to draft responses to complex customer complaints.
Arun immediately saw the weaknesses.
The system could misunderstand a customer's situation. It could produce inaccurate explanations. It could promise something the organization could not deliver.
His familiar response was ready:
"This is too risky."
But he remembered the provocation.
Instead, he asked:
The team explored the question.
Perhaps AI should not send responses.
Perhaps it could prepare a draft for a trained employee to review.
Perhaps it could identify missing information before a response was written.
Perhaps it could help employees compare a draft with approved policy, while leaving the final decision to a human.
The original proposal had not survived.
Something more useful had emerged from it.
Arun's expertise had not become less important.
It had become essential in a different way.
He could now help the team distinguish a promising possibility from an unsafe implementation.
And he could do so without ending the exploration prematurely.
5. The Discovery
Arun realized that he had been treating ideas as finished objects.
Either they worked or they did not.
But many new ideas were not finished objects.
They were starting points.
An incomplete idea could contain a valuable principle. An impractical proposal could reveal a different application. A technically incorrect suggestion could provoke a better question.
His experience helped him recognize the defects.
Now he was learning to recognize the possibilities those defects might conceal.
6. The Research Behind the Story
6.1 Why Expertise Can Create Blind Spots
Research by Bilalić, McLeod and Gobet (2008), studying expert chess players, found that familiar solutions could interfere with the discovery of better alternatives.
This phenomenon is often discussed in connection with the Einstellung effect: prior knowledge or a familiar solution can bias attention and problem-solving.
The implication is not that expertise is harmful. It is that expertise can sometimes make one approach so readily available that other approaches receive insufficient consideration.
Erik Dane's (2010) work on cognitive entrenchment further examines the tension between specialized expertise and cognitive flexibility.
Importantly, cognitive entrenchment is not inevitable. Expertise can also provide the knowledge needed to recognize, refine, and implement innovative ideas.
6.2 What Changes When AI Enters the Picture
In Generative AI at Work, Brynjolfsson, Li and Raymond (2025) studied AI assistance among 5,172 customer-support agents.
They reported an average productivity improvement of approximately 15%, with particularly large gains among less experienced and lower-skilled workers.
The study does not show that experienced professionals are less creative or inherently resistant to AI.
It shows something more relevant: AI can change how knowledge and experience translate into workplace performance, and its benefits are not distributed uniformly.
Research by Dell'Acqua and colleagues on the jagged technological frontier likewise demonstrates that AI assistance can improve performance on some knowledge-work tasks while creating difficulties on others.
Consequently, the professional challenge is not simply to adopt AI faster.
6.3 What Edward de Bono Contributes
In The Mechanism of Mind, Edward de Bono developed a conceptual account of the mind as a pattern-forming information system.
This provides a useful theoretical lens for understanding why established perceptions can guide subsequent thinking.
In PO: Beyond Yes and No, de Bono introduced PO as a means of moving beyond immediate acceptance or rejection toward new possibilities.
His treatment of Challenge for Change is especially relevant here.
The challenge is not necessarily an assertion that the existing way is wrong.
It is an invitation to explore whether another way might be possible.
The PO statement in this story is an original contemporary application of that method, not a quotation from de Bono.
Nor should de Bono's conceptual account be treated as a complete neuroscientific explanation of expertise. The independent studies provide more specific evidence for the psychological and workplace claims.
7. From Possibility to Accomplishment
Arun introduced a small change to the way he reviewed unfamiliar proposals.
He separated exploration from final evaluation, while retaining immediate safeguards for serious risks.
He called it the Two-Pass Review.
Identify the useful principle, question the assumptions, and generate alternatives.
Ask what could be changed to make the idea more useful or feasible.
Examine feasibility, evidence, cost, reliability, safety, accountability, and unintended consequences.
Reject, redesign, or test the resulting proposal as appropriate.
Arun did not stop finding flaws.
He stopped assuming that finding a flaw meant the exploration had to end.
8. Try It Yourself: A Seven-Day PO Experiment
Choose one unfamiliar idea that you would normally dismiss.
For seven days, use the following questions before reaching a final conclusion, provided no immediate safety or compliance issue requires rejection.
- What familiar assumption is shaping my reaction?
- What useful principle might exist inside this idea?
- What alternative form could make that principle workable?
- What evidence or small, low-risk test would help me evaluate it?
Compare the resulting alternatives against the original proposal using criteria chosen in advance.
You may discover a better approach.
You may also confirm that the original idea should be rejected.
Both outcomes are useful if the exploration improves the quality of your decision.
9. The POrable Insight
Expertise should not merely tell us what cannot work. It should also help us discover what might.
The aim is not less judgment. It is better-timed judgment, informed by deliberate exploration.
Arun was still the person everyone trusted when something went wrong.
But now people also approached him when something new might be possible.
His expertise had not changed.
The direction in which he applied it had.
10. Sources and Further Reading
This POrable draws on Edward de Bono's thinking methods and independent research on expertise, cognitive fixation, and AI-assisted work.
- Edward de Bono (1969). The Mechanism of Mind. Conceptual foundation for pattern-forming information systems and perception. Exact passage-level verification remains outstanding.
- Edward de Bono (1972). PO: Beyond Yes and No. Relevant sections include first- and second-stage thinking, PO as a change tool, and Challenge for Change.
- Bilalić, M., McLeod, P., & Gobet, F. (2008). Why Good Thoughts Block Better Ones: The Mechanism of the Pernicious Einstellung (Set) Effect. Cognition. Read research
- Dane, E. (2010). Reconsidering the Trade-off Between Expertise and Flexibility: A Cognitive Entrenchment Perspective. Academy of Management Review. Read research
- Brynjolfsson, E., Li, D., & Raymond, L. R. (2025). Generative AI at Work. Quarterly Journal of Economics. Read research
- Dell'Acqua, F., et al. (2026). Navigating the Jagged Technological Frontier. Organization Science. Read research
