Summary
Arturo Nunez, a former Unity engineer with nearly a decade of experience, presents Nereu — an AI-native game engine designed to let anyone build games by describing what they want in plain language rather than writing code or configuring complex editor panels. In the live demo, an AI assistant named Bibi interprets natural-language prompts to spawn assets, configure physics, set up animations, and attach cameras, reducing a multi-hour traditional workflow to a few minutes of conversational iteration.
The technical foundation is an “Asset Tag System” (ATS) inspired by the Entity Component System pattern from game development, where objects are described with intent-based tags (“player,” “drivable,” “double-jump”) and runtime systems automatically wire the appropriate components. The AI assistant uses scene context plus asset-library metadata to resolve ambiguous prompts and suggest matching resources.
Nunez argues that the bottleneck in game creation has never been creativity but the steep technical learning curve — 3D modeling, rendering, physics, animation — that forces small teams to spread expertise impossibly thin. Nereu’s bet is that a natural-language interface sitting on top of a well-structured tag system can democratize game-making the same way no-code tools democratized web publishing, letting creators focus entirely on the game idea itself rather than the engine underneath it.
📺 Source: AI Engineer · Published August 18, 2026
🏷️ Format: Showcase
