There comes a phase where daily coding activities start to feel repetitive. Not because of a loss of interest, but because a new question emerges: is a programmer’s role limited to just writing code? From that reflection, I decided to step away from routine coding and begin experimenting with a different approach—building an AI Assistant as a space for highlighting exploration and learning.
This experiment was built using Openclaw, a framework that enables more flexible and modular AI agent development. The goal was not to create a perfect product, but to understand how AI can interact, respond to context, and genuinely assist users. The process involved a lot of trial and error, from configuring agents and response flows to managing limits and performance.
What makes building an AI Assistant interesting is how different it feels compared to conventional coding. I was no longer focused solely on “does this code run,” but rather on “does this response make sense,” “is this AI helpful,” and “how does the user feel during interaction.” At this point, the role of a programmer shifts into that of a behavior and experience designer.
This experimental AI Assistant, named Selnoviktech AI Assistant, was then opened to the public as an open testing space. The goal is simple: to observe how AI behaves in real-world usage, receive direct feedback from users, and continuously improve from there.
Let’s playground: https://t.me/selnoviktech_ai_assistant_bot
For me, stopping coding does not mean stopping creation—it marks a step toward understanding the future of technology from a broader perspective.