Why “knowing how to use AI” says little about the work

The more I use AI, the less I believe that “knowing how to use AI” is enough to explain my work. The skill is still useful and, in many contexts, expected. What seems weak to me is presenting it without context, as if the tool explained the quality of what was done.

A phrase that fits everything

“Knowing how to use AI” can describe very different activities: revising a text, researching a subject, building a prototype, organizing information, analyzing data, automating a step in a process. Each one requires different knowledge, and the tool plays a different role in each case. On its own, the phrase says little about what someone is able to deliver.

At the same time, the skill has become a requirement. The 2024 Work Trend Index, from Microsoft and LinkedIn, surveyed 31,000 people in 31 countries and found that 75% of knowledge workers were already using AI at work, and that 66% of leaders said they would not hire someone without AI skills. The same study found that only 39% of users had received any training from their company (Microsoft and LinkedIn, 2024 Work Trend Index).

If the skill is widely expected, it no longer sets anyone apart by itself. What sets people apart is the ability to show what was done with it.

What to show instead of the phrase

In my portfolio, I prefer to show the task and the decisions. I try to answer four questions for each project.

question	why it matters
What problem was I trying to solve?	places the work in context and makes it possible to judge whether the solution makes sense
What was my role?	separates what I decided from what the tool produced
What was built?	shows the concrete deliverable, not the intention
What has not been validated yet?	keeps a proposal from looking like a result

The second question is the most delicate when AI is part of the process. One approach I have been using is to describe the steps: who defined the problem, who wrote the brief, what the tool generated, what was reviewed, changed or discarded, and by which criterion. This description does not diminish the work. On the contrary, it makes visible the part that requires judgment.

A first version is not understanding

This exercise also helps me notice what I need to learn. Getting to a first version with assistance does not mean understanding every detail needed to maintain that solution. A page prototype built with AI help can work in a demo and hide technical decisions I would not be able to justify or fix on my own.

When I notice that gap, I record it in the project itself. This has two effects: it makes the presentation more honest for whoever is evaluating it, and it turns the gap into a concrete list of what to study.

Being recognized for what you analyze and execute

I want to be recognized for what I can analyze, organize and execute in marketing. AI is part of that process and expands what I can do, in speed and in reach. The portfolio needs to show where it plays a role and within which limits, because that is the information that allows someone to evaluate the work.

In practice, this led to a reorganization of the portfolio itself, with each project labeled by its nature. I described that process in How to indicate the nature of each project in a portfolio.

References

• Microsoft and LinkedIn. 2024 Work Trend Index on the state of AI at work. May 8, 2024. Link