Colleagues often ask me if I’m worried about the growth of AI and if it will take my job away. Honestly, not really. What worries me is not that AI is capable of writing code or even complete solutions, but how quickly people are willing to hand over responsibility to it.

I’ve worked with people who treat AI agents like servants that do everything for them. One agent writes code, another reviews it, then yet another posts on Slack that it’s done, so the first one can deploy the change to production. I get that it’s a lot easier to deliver work, but we must still remember that systems have to be maintained.

Around a year ago, when ChatGPT started becoming more common, I noticed that a lot of content on LinkedIn was clearly being written by AI. At first I didn’t mind it, but I got overwhelmed by it pretty quickly, because it was flat and lacked substance, personality, and emotion. Then the same thing started happening in the comments. AI was praising other AI-written posts.

These examples show how easily people hand responsibility over to AI agents. I think that’s a misuse of one of the best opportunities AI gives us: the chance to learn faster and grow faster.

We should use AI as a companion and partner to make our work better. “Let’s do it together” is a much better approach than “do it for me.”

Let’s say I need to write a Cloud Run function that accesses the Kubernetes cluster to restart a workload.

I start with the overall idea. I’m still the owner of the task. I should drive the architecture. I should define how to access the cluster securely. I should point an agent to a secret with a proper kubeconfig or service account token. Then it’s my responsibility to look at the code, review it, and potentially ask AI to split the code into separate files/functions, etc.

I always wanted to make a videogame. This is what got me into IT in the first place. But every time I tried to learn game engines, physics and all other moving parts, I felt overwhelmed. With AI, I was able to come up with a step-by-step plan for how to approach the subject and make the learning curve feel much less steep.

I started with a simple prompt: “I want to create a top-down pixel art RPG game. I know nothing about Godot. Please help me create a plan so by the end of it, I’ll be able to build my dream game in Godot. Do not write any code for me.”

The result was great. The AI helped me shape a full roadmap, starting with a simple Pong clone, then moving on to Breakout, a side-scroller, a Zelda-like game, and finally the RPG itself, with a clear idea of what I would learn at each step.

But that’s not all. It also helped me create a 14-day plan for each small game, with clear goals and deliverables. For example: “On Day 2, we’ll work on collisions.”

I want to emphasize the word “helped.” It didn’t do it for me. The point is for me to grow, not the model :)

This way not only the AI is learning, but more importantly, we do.

That’s why I’m not especially afraid of AI taking my place. What worries me more is people giving up the parts of the work that matter most: thinking, understanding, and taking responsibility. AI is at its best when it helps us grow into the work, not step away from it.

– MK