This is part 2 of a six-part series exploring AI, professional identity and the changing value of human expertise.
This article was originally written in English. If your browser is set to Dutch, it may be translated automatically. Please note that automated translations may contain grammatical errors or inaccuracies.
When expertise takes on a new meaning
Professional identity has always been shaped by scarcity. We become known for capabilities that other people cannot easily reproduce.
Writing a compelling report.
Synthesising large amounts of information.
Finding the right evidence.
Producing a clear presentation under pressure.
Turning a messy discussion into a coherent strategy.
These capabilities take time to develop. They become more than tasks. They become evidence of expertise, and part of how people understand their professional value.
Then AI arrives and, within minutes, produces something that looks remarkably similar.
That can create professional confusion followed by an uncomfortable question:
If AI can produce in five minutes something that looks like what took me twenty years to learn, what exactly am I bringing?
When the once-difficult becomes ordinary
Generative AI is particularly relevant to knowledge work because it can perform activities involving writing, summarising, research and the communication of information. Studies of real-world AI use show that these are among the most common workplace activities for which people seek AI assistance (Tomlinson et al., 2025).
That does not mean entire professions disappear. Jobs, as anyone who has ever had one knows, are made up of many tasks, relationships and responsibilities. But it does mean that some visible signs of professional competence are becoming easier to reproduce.
Imagine these situations:
A polished first draft once demonstrated writing ability.
A well-structured presentation once demonstrated the ability to organise information and tell a story.
A concise summary once demonstrated that someone had read and understood a large body of material.
Now, each of these outputs can be generated quickly.
The output may be imperfect. It may contain weak reasoning, factual errors or generic recommendations. But it can look convincing enough to change how people perceive the value of producing it.
AI changes what is scarce
This is the distinction I keep returning to:
AI can help us produce answers faster. Experience helps us ask better questions.
AI can make the production of text, images, analyses and presentations faster and cheaper. But abundance in production creates a new problem: someone must decide what is worth keeping.
Which information is correct? Which recommendation fits this organisation? What is missing? What should be challenged? What is technically possible but ethically wrong? What sounds persuasive but will fail in practice? When more content can be produced, the ability to evaluate it becomes more important, not less.
In this dynamic, the scarce resource begins to change.
It moves away from producing information and towards interpreting it. Away from creating the first answer and towards recognising the better one. Away from presentation and towards judgment, context, trust and responsibility.
Fluency or expertise?
AI is good at producing fluent language. Fluency, however, is not the same as understanding.
A strategy document may be logically organised but politically naïve.
A change plan may follow established models while overlooking unwritten stakeholder dynamics.
A leadership message may sound polished but strike entirely the wrong emotional note.
A risk analysis may identify common risks while missing the one everybody inside the organisation knows but nobody has documented.
This is where professional experience matters.
Experienced professionals carry context that may never have been formally recorded. They recognise patterns, read the room, remember previous attempts and understand which apparently minor detail could derail an otherwise sensible proposal.
Their value does not reside only in the document they produce. It resides in what they notice before, during and after producing it.
A different test for professional identity
This transition may be difficult because organisations have often rewarded visible output.
How many reports were completed? How quickly was the presentation delivered? How polished was the proposal?
AI challenges the value of those signals. And it offers an opportunity to define expertise more honestly. Perhaps expertise was never primarily about producing the artefact. Perhaps the artefact was simply how expertise became visible.
If writing, researching and structuring information become abundant, organisations will need to become better at recognising what remains scarce:
- Sound judgment.
- Institutional memory.
- Ethical reasoning.
- Relational intelligence.
- The courage to disagree.
- The wisdom to know when no available option is good enough.
AI is changing how expertise creates value. It also gives us an opportunity to recognise the experience, judgement and understanding behind good work.
And that leaves us with a useful question: If information is becoming abundant, what should we learn to value as scarce? #
Sources
- Eloundou, T., Manning, S., Mishkin, P., & Rock, D. (2023). GPTs are GPTs: An early look at the labour market impact potential of large language models. arXiv.
- Tomlinson, K., Jaffe, S., Wang, W., Counts, S., & Suri, S. (2025). Working with AI: Measuring the applicability of generative AI to occupations. arXiv.
