Get ready for a controversial take on AI.
Every week, I read dozens of posts urging professionals to use AI more. The consensus is clear: if you’re not leveraging AI, you’re falling behind. And they’re not wrong. AI is a powerful tool. Refusing to learn it is like refusing to use a calculator in an accounting job, it’s not bold, it’s foolish.
But here's a question no one seems to be asking: How much AI is too much?
Bear with me. That’s what this post is about: understanding the hidden risks of overreliance on AI. Because while most of the internet is preaching acceleration, some of us are quietly watching professionals dull the very skills that once set them apart.
The Cognitive Cost of Letting AI Think for You
It’s not just a hunch. Research shows that relying on AI too heavily can reduce cognitive engagement.
A recent study published in MDPI’s “Societies” journal looked at more than 600 participants and found a significant negative correlation between high-frequency AI use and critical thinking ability. The more often people turned to AI for answers, the less effort they made to analyze or reflect independently. This effect is known as cognitive offloading. The brain stops trying, because the machine always delivers.
An even more telling experiment from MIT’s Media Lab gave participants a writing task. One group used ChatGPT, another used Google, and a third worked without any external tools. The group using ChatGPT showed lower memory performance, less executive control, and weaker originality. Not only were their essays less effective, they retained less of what they learned.
This should raise alarms, especially for young professionals still developing their craft. The more often you skip the mental reps, the less capable you become over time.
Let me put it this way.
Say your goal is to exercise, so you build a habit of running five miles a day. Then one day, you buy a car. To make things easier, you start driving five miles instead. Did you still get your exercise? No. You got to a destination, but that was never the point. The value was in the work, not the outcome.
This is the problem with unchecked AI use. You might get a result. But you’ve skipped the work that builds skill: creativity, logic, problem-solving.
And just like physical muscles, if you stop using them, those abilities atrophy.
Use AI, But Know When Not To
The challenge now isn’t whether to use AI. That’s settled. You should. But knowing when not to? That’s the new edge.
Professionals who will thrive long-term aren’t those who use AI for everything. They’re the ones who treat AI like a tool, not a crutch. They’ll still do deep work. They’ll keep writing from scratch, debugging without prompts, solving problems in their own head before asking for help.
And over time, they’ll keep their edge sharp while the rest dulls from disuse.
Here’s what that looks like in practice:
- Skip the AI when you're not under time pressure. If a task will take 15 minutes on your own, do it. Build the mental reps.
- Deliberately create AI-free days. Once a week, commit to solving without autocomplete, templates, or generated answers.
- Use AI as a second brain, not the first. Try first. Then use the tool to check or expand your thinking.
Ask Yourself This Every Day
Could I still do this if I didn’t have AI?
If the answer is no, that’s your cue to take the harder path, or to use AI in a way that lets you understand the reasoning behind the answer, learn with it, don’t let it take the reins. Not always. Not for everything. But often enough to stay sharp.
AI is here to stay. But so is the human brain if we let it work.
