The Ghost in My Code

There are moments at work that bring a profound sense of completion. For me, it's the decisive click of the "Send" button after pouring all my effort into a product.

Last week, that feeling came with a proposal, the culmination of a full week's work. It was a crystallization of dry data analysis and strategic ideas that I believed were compelling enough. The first recipient, as a safety precaution, was my boss, the CEO, for his review before it officially went on its way to win over the client.

A few minutes later, a Slack message popped up. It wasn't praise or feedback on the content. He asked a direct question: "What percentage of this was AI-assisted, Thuan?"

Without a moment's hesitation, I typed back immediately: "About 30%. I used it to brainstorm ideas and for English grammar checks."

To me, it was an effective and perfectly normal conversation. AI is a tool, a collaborator, and acknowledging its existence is as natural as admitting you used Google for research or a computer to type a document.

But then it struck me that my comfort seemed to be an exception. I recalled team meetings, code reviews. I noticed the hesitations, the evasive answers, the awkward silence when someone asked, "This code is brilliant, how did you come up with such a clever algorithm?"

It's like there’s an invisible "ghost" in our lines of code, in our reports—a silent collaborator no one dares to name. Why are we so reluctant to acknowledge such a powerful assistant? Where does that hesitation come from? Today, being the weekend, while waiting for my ride to Saigon, I decided to sit down and try to "peel back" the psychological layers behind this "ghost."

Peeling Back the Layers - Decoding the Silence

When you look closely at this silence, it’s not mere modesty. It’s a complex of very human fears, wrapped in several layers.

First Layer: Imposter Syndrome and the Fear of Being "Exposed." The AI "ghost" embodies the greatest fear of the "imposter"—a common psychological syndrome among knowledge workers: that one day, someone will point at their work and say, "You didn't do this." Admitting to using AI feels like handing them the evidence for that very accusation. It raises questions about one's true value. Every line of code suggested by AI is a whisper: "See? You would've been stuck without me." Silence, therefore, becomes a defense mechanism to protect one's sense of self-worth.

Second Layer: The Craftsman’s Pride and the Fading Sense of Accomplishment. I have a side hobby: woodworking. In my small workshop, alongside my hand chisels and planes, I have a CNC machine. It helps me cut complex parts with near-perfect precision, creating products that would take immense time and effort by hand. But this very machine sometimes puts me in debates with my friends who are traditional woodworkers. They argue that machine-made products lack "soul," the touch of the hand, and the intentional imperfections that only a true craftsman can create. For them, value isn't just in the final chair, but in the calluses on their hands and the sawdust in their hair.

This debate echoes strangely in our digital world. Programmers, writers, and creators share a similar pride in their "craft"—in the process of grappling with a problem, searching through the night, and that "eureka" moment of finding the solution on their own.

And then the AI "ghost" appears, just like the CNC machine in my workshop. With its efficiency, AI has inadvertently stripped away the joy and pride of that process. It turns the craftsman into an operator. The product still gets made, perhaps even better, but the pride in the creative journey is eroded. Silence, in this case, is sometimes a way to hold onto the feeling that they are still the "craftsman," not just a "button-pusher."

Third Layer: The Silent "Valuation" and the Fear of Obsolescence. In a corporate environment, this fear is very real. Many worry: "If I admit to using AI for 30% of my work, will my boss silently conclude that 30% of my capabilities are replaceable?" They fear that transparency will become data for others to "re-evaluate" their competence. They're afraid of being categorized in the "needs AI to function" group, which might be seen as less valuable and more easily replaced than the "doesn't need AI" group. This is about job security, a defense against an uncertain future.

Fourth Layer: The Comfort Trap – When the "Life Jacket" Makes Us Too Lazy to "Swim." Just as a life jacket cannot replace the skill of swimming, some people are hesitant not out of shame, but from a deeper fear: dependency. They worry that if they rely too heavily on this "ghost," their core skills—logical thinking, deep debugging, original creativity—will gradually dull. Silence is sometimes a form of self-discipline, a reminder to "try to solve it yourself first," to ensure they're still "swimming" and not just floating aimlessly thanks to the "life jacket."

The Inevitable Shift: From "Cheating" to "Competitive Advantage"

These fears are real and very human. But history shows that every tool that amplifies human intellect goes through a similar cycle of perception. The pocket calculator was once considered "cheating" in math class. Google and Stack Overflow were once seen as "shortcuts" in the programming world.

We are at the very cusp of a similar shift with AI. When a technology becomes powerful enough, widespread enough, and proves its superior effectiveness, an inevitable reversal of perception will occur.

Hesitation and concealment will gradually fade, giving way to a new race. The question will no longer be "Do you use AI?" but rather: "How well do you use AI?"

Proficiently using AI will no longer be a support function; it will become a sought-after core skill in itself. It's not just about typing a few questions, but a whole art and science, including:

• The Art of Questioning (Prompt Engineering): Knowing how to "converse" with the machine to extract the most accurate, creative, and insightful answers.
• The Power of Vetting (Critical Validation): The ability to quickly evaluate, debug, and refine what AI produces. Knowing what is "gold" and what is "garbage" in the sea of information it provides.
• An Integration Mindset: Knowing how to seamlessly combine AI's output into a larger system or a complex workflow to generate the highest value.

In the very near future, job descriptions will include the line: "Proficiency in using Copilot, ChatGPT-4, or similar large language models required." Your transparency about using AI for 30% of your work will no longer be a weakness, but could be proof that you possess a skill of the future.

Conclusion: Transparency is Strength

Let's return to the conversation with my boss. My comfort in answering "30%" came from having reframed the game and my role in it. My value doesn't lie in typing every single word of the proposal, but in my ability to architect a strategic solution and take ultimate responsibility for its quality.

In this new landscape, transparency is strength. Answering "30%" isn't a confession; it's a declaration: "I not only get the job done, but I know how to use the most advanced tools to optimize performance and quality. I am the master of my tools, not a slave to old habits."

We must shift from the mindset of a craftsman who only knows their own chisel and hammer. Let's see ourselves as conductors. A conductor's power isn't in playing the violin or the horn, but in their ability to listen, coordinate, and harmonize all the musicians (including the AI musician) to create a complete and magnificent symphony. Your conductor's baton is your capacity for critical thinking, validation, and integration.

The "ghost" in our code is, in reality, the silhouette of the future coming into focus. Instead of fearing or avoiding it, let's learn to converse with it, master it, and turn it into our most powerful collaborator.

Because in this era, those who don't know how to use AI proficiently are the ones truly falling behind. Your comfort today will be your sharpest competitive edge tomorrow.