AI 改游戏世界：能改对，但改不像
让 AI 在游戏里生成世界已经不难，难的是改一个现成的世界——比如把 Minecraft 里的村庄改成要塞，但保留村民和地形。这篇论文把「改世界」拆成四个深度：改属性、改实体、改动力学、改系统，越深越难。最强的 AI 智能体在 110 个任务里能解决 78.2%，但几乎所有失败案例都是「改完能运行，但行为不对」——房子建好了，门却打不开。视觉一致性更是硬伤，所有配置的视觉通过率都低于 50%。这不是你明天能玩到的功能，但它说明：AI 改游戏世界的能力已经接近「能干活」，离「干得漂亮」还有距离。
📄 原文摘要(英文)
Interactive world models are increasingly capable of generating environments and acting within them, yet deliberately editing an existing executable world remains underexplored. We formulate world editing as intervening on an existing world while preserving properties that should remain unchanged, and introduce intervention depth as an axis describing how strongly an edit couples world entities, dynamics, and systems. We instantiate this capability through industry-grade game modding and introduce IGMWorld, together with IGMBench, a benchmark of 110 tasks and over 1.1K executable state and behavioral criteria across Minecraft and Terraria. The tasks span property, entity, dynamics, and system interventions and are evaluated through deterministic executability, behavioral, preservation, and visual checks. Frontier coding agents already exhibit substantial world-editing capability: the strongest configuration solves 78.2% of tasks under a strict task-level criterion, while criterion-level performance reaches 94.8%. Reliability generally decreases with intervention depth, and this pattern persists even among tasks with similar numbers of evaluation criteria. Most failed edits still build and load successfully, suggesting that the main difficulty is making the edited world behave as requested. Visual consistency remains a separate weakness, with all evaluated configurations below 50% joint visual pass rate. These results show that world editing is a distinct capability from world generation and interaction, and that executable games provide a practical testbed for studying it.