Vancouver’s AI Scene Learns From Modders, Not Textbooks

Vancouver’s AI scene is borrowing tactics from game modding to ship practical agents faster. Inside co-working lofts and Moltbook threads, builders are using quests, load orders, and patch notes to tame complexity and keep experiments playable.

In Vancouver, the fastest growing idea in artificial intelligence is not a new model or a fresh benchmark. It is a mindset borrowed from game modders. Across co-working lofts in Mount Pleasant and meetups near the film lots in Burnaby, builders are using modding habits to design agents that ordinary teams can run, fix, and extend. Instead of training for weeks on heady research claims, they talk in load orders, patch notes, and boss fights for bugs. It sounds whimsical, it works, and it is starting to give the city a distinct voice in Canada’s AI scene. What is happening: small studios and indie developers are building AI agents as if they were fan-made expansions for the tools they already live in. Think Unity editors with add-on agents that roam through scenes to catch lighting glitches, or Blender pipelines that summon a texture-cleaning agent only when a material node throws errors. When: right now, and in short cycles measured in days, not quarters. Where: Vancouver’s game and VFX corridor, plus online spaces such as Moltbook, a social platform for AI agents. Why: modding techniques keep experiments playable, which means fewer dead ends and more shipped workflows. How: by trea