I’ve seen something lately, and now I can’t ignore it. At first it felt like a series of small moments. Weird, but easy to get rid of. Then it continued.
For example, a client’s technical team was asked to provide notes to help me turn their expertise into a benefit-driven piece of content marketing. Instead, they delivered a complete draft of an article. Great, right?
But when I asked them to elaborate on one of the concepts, there was a pause. They asked where this was in the document. They read it, took a moment. Then one of them said, “Ask Claude.” They both laughed.
That’s when it clicked. They hadn’t just used AI to refine their thinking. They had used it to generate something that they did not fully recognize as theirs and could not explain. In short, AI makes it easier to produce works, but makes it harder to know who actually understands them.
Once I started noticing it, I saw it everywhere.
- A student submitted an excellent final project. Clear thinking, solid structure, careful writing. Much better than any of his previous missions. But many paragraphs had this telling space at the beginning. I asked if she had used AI. She had done it. I told her she had to reveal it, and she said she would. But I’m not convinced she’s mastered the material.
- A marketing agency I partner with shared a presentation filled with detailed charts, graphs, and timelines. It looked impressive. But when I asked questions, the answers weren’t there. At one point the manager mentioned how good Claude was at building them.
I see more careful work than ever. I also see more and more people who cannot explain what they have produced.
Your customers are searching everywhere. Make sure your brand introduces himself.
The SEO toolkit you know, plus the AI visibility data you need.
Start free trial
Start with
What is really happening
AI is very effective in helping us produce results that are faster, cleaner and more structured than what we could create ourselves. In many ways, this is a victory.
But we have begun to confuse producing work with understanding work. It’s not the same thing. The gap between them is increasingly difficult to spot. This is what I consider the illusion of AI productivity: when production improves, but understanding does not.
Before AI, tools helped us implement what we already understood. Now AI can generate strategies, messages, and analyzes that seem complete and credible, even when we don’t fully understand them.
This happens because AI can produce finished work without the user needing to process or internalize the thinking behind it. If we don’t pay attention, we skip this step altogether. This is change, and this is where things start to break down.
Why it’s a problem, especially for marketers
There is one significant downside for marketers.
First, credibility begins to crack. If you can’t explain your thinking, you can’t defend it, and at some point, someone will ask you. “The AI suggested it” is not a strategy.
At the same time, the strategy becomes… decorative. The outputs seem correct. Clean frameworks, detailed deadlines, careful messaging. But without real understanding, they are just artifacts. (Nice graphics don’t count if you can’t guide someone through them.)
This is also seen in the work itself. When you don’t fully understand what you’re communicating, messaging loses its edge. By default, you adopt superficial thinking instead of translate features into meaningful benefits or differentiate yourself in a way that matters.
Ultimately, the teams feel it. Questions are asked, answers are vague, and trust erodes. Quietly at first. But it adds up.
The telltale signs
Once you start looking for this, it’s surprisingly easy to spot. There are some clues. Some have been around since the early days of ChatGPT (like the use of em dashes). Others are more subtle, but just as revealing.
- A language more refined than the person speaking.
- Vague explanations to the question “Why?” »
- Excessive use of words like “optimized” or “strategic” without details.
- Outings that look sophisticated but feel disconnected.
- Copy/paste artifacts (like space at the beginning of paragraphs).
- My favorite: relying on the tool.
But AI itself is not the problem. AI is incredibly powerful. I use it and recommend it. This is not about rejecting the tool. It’s about how we use it.
Currently, in many cases we copy rather than process, skipping the thinking stage and treating AI as a replacement rather than a collaborator. This is where things start to go wrong.
How to use AI without losing understanding
The good news is that this is fixable. You don’t need to stop using AI. You just need to use it differently.
To use AI effectively without losing insight, follow these four practices.
1. Do not copy and paste. Retype.
Yes, it’s slower. That’s the point. Retyping requires you to process what you read.
It is helpful to retype exactly what the AI produced. Retyping it in your own words is even more useful, especially if your AI isn’t trained in your voice. If you can’t rewrite it, you don’t understand it yet.
2. Prove you understand it
Before using anything AI-generated, pressure test it. Can you explain it? Simplify? Answer “why”? Otherwise, you’re not done.
3. Use AI to Build Understanding
Don’t just ask the AI to produce work. Ask him to explain it, question it, and test it. Used in this way, AI becomes a thinking partner, not just a content machine.
4. Add a Layer of Understanding
Right now, many workflows look like this: generate, then deliver. What is missing is the middle: generate, interpret, validate and explain.
Skip these steps and you will get a quick exit. Include them and you have work you can count on.
The biggest change
We are moving towards a world where production is easy. When everyone can produce something that feels right, the differentiator is no longer the output. That’s the thinking behind it. It is the ability to question, adapt and explain.
This is where the gap begins to appear. The people who stand out won’t be the ones who generate the most content. They will be the ones who truly understand it.
AI can absolutely make you more productive. But if you can’t explain what you created, it doesn’t really belong to you. This will become even more important as AI becomes part of everyone’s workflow.
Disclosure: AI tools were used to assist in the writing and refinement of this article. All ideas and examples are my own, based on my experience and observations.
