A recent Business Insider report found that employees at KPMG are being tracked on how often they use AI—and some say the system is already being gamed. The firm isn’t alone. Companies like JPMorgan, Amazon, and Disney are also rolling out dashboards to measure AI adoption, usage frequency, and output.
On paper, it’s about productivity. In practice, it’s creating a new kind of workplace pressure: use AI often, or risk falling behind. And when usage becomes a metric, people start optimizing for the metric—not necessarily for better work.
So where does that leave you if you actually care about doing original, thoughtful work? Here’s the honest answer: you don’t fight the system directly. You outgrow it.
1. Start your thinking offline (or at least off-AI)
Before you open any AI tool, force yourself to sit with the problem first. This is where your real value is formed. When you begin with AI, you’re starting from an average of existing ideas, which makes it much harder to produce something original. By thinking independently first—even if your ideas are rough—you create a point of view that AI can’t replicate. Without this step, your work risks sounding polished but generic, which is exactly what most AI-heavy outputs look like.
2. Use AI as a second brain, not your brain
AI works best when it is responding to you, not replacing you. If you treat it as your primary thinking engine, you slowly lose your ability to question, refine, and direct ideas. But when you approach it as a collaborator—something you brief, challenge, and iterate with—it becomes far more powerful. The difference is subtle but important: one approach makes you dependent, the other makes you sharper.
3. Define what “good” looks like before you generate anything
One of the biggest reasons AI outputs feel generic is that people prompt without a clear standard. If you don’t know what you’re aiming for, you’ll accept whatever comes out. Taking a moment to define what “good” means—whether that’s sharper insight, stronger structure, or a more distinct voice—gives you a filter. It allows you to judge and refine AI outputs instead of passively accepting them.
4. Build your own taste—and protect it
Taste is what separates average work from memorable work. It’s your internal standard for what feels insightful, overused, boring, or fresh. AI tends to flatten things toward what is most common and widely accepted. Without strong taste, it’s easy to accept that average. But when you know what you like—and what you reject—you can push beyond AI’s default tendencies and produce work that actually stands out.
5. Make your thinking visible
Many companies rely on AI dashboards because they don’t have visibility into how work is done. If the only thing they can measure is tool usage, that becomes the proxy for productivity. By making your thinking visible—through drafts, explanations, or structured updates—you give people a clearer way to evaluate your work. This reduces the pressure to “perform” AI usage and shifts attention back to how you actually solve problems.
6. Use AI at the edges, not the center
Not all parts of your work carry equal value. The core—your ideas, your argument, your strategy—is where originality lives. The edges—formatting, restructuring, quick synthesis—are where AI can be most useful. When you let AI take over the center, your work loses its distinctiveness. But when you apply it at the edges, you gain efficiency without sacrificing originality.
7. Focus on outcomes people can’t ignore
AI dashboards measure activity because it’s easy. But in most workplaces, what ultimately matters is impact. A clearer strategy, a sharper piece of writing, a better decision—these are harder to measure but easier to recognize. When your work consistently produces strong outcomes, it creates its own credibility. Over time, that matters far more than any usage metric.
David Adubiina
