Game Development in the Dawn of AI

My junior years were not in game development. They began in Android apps, and most of my main programming growth happened in high-load ETL and cloud integration environments built with MuleSoft and Azure.

That period taught me how often engineering feels uncertain from the inside. You can be working hard and still feel lost or helpless, and books rarely contain the precise clue that unlocks your production issue.

Working in larger companies helped offset this. Access to a wider pool of talent and stronger senior guidance in backend integration made it easier to get unstuck and improve faster.

Switching into game development was a sharp transition. Bridging that gap required extra personal time in side projects, experimentation, and self-directed tinkering to become fluent with technology that teams may expect you to already know.

When agentic coding first entered my workflow, I pushed back. I questioned trust, code quality, and whether it would weaken engineering fundamentals.

Over time, that position changed into deliberate adoption. AI became useful for mundane tasks, narrowing noisy search space, and accelerating exploration. Personal projects, including this website and this article process, became easier to execute.

The hardest boundary remains architecture. AI can generate useful implementation fragments, but long-horizon structure, tradeoff judgment, and system coherence still require human intelligence and intuition.

This is where I now stand: AI should lower the barrier to entry into complex technical domains, but it should not replace ownership of design decisions. The strongest results come from leveraging AI to reduce friction while humans stay accountable for intent, architecture, and quality.