四个 Gemini 4 Argon 智能体在无人类协助下通关 Minecraft
Four Gemini 4 Argon agents beat Minecraft without any human help
- 1四个 Gemini 4 Argon 智能体首次在无人类输入下通关 Minecraft，自主击杀末影龙
- 2Google DeepMind 团队未预设角色、策略或游戏专用工具，智能体自行分工、互报威胁并共享记忆
- 3全程 27 小时 14 分钟，共编写 2,783 个 JavaScript 程序，交换 229 条聊天消息，死亡 605 次
- 4按 2027 年生效的 Gemini 4 Argon 标准价计算，整场运行花费 1,620 美元
通用大模型智能体靠自组织协作完成了以往需从零训练专用 AI 的复杂任务，长时程运行的算力成本也有了可核算的标价。
- 影响：长时程多智能体任务有了具体的成本参照，27 小时 1,620 美元可作为预算量级；无需人工设计分工的协作模式，为流程自动化提供了不同于定制开发的路线。
- 参考：可考虑选一个流程长、规则清晰的场景试点多智能体协作，先按 27 小时约 1,620 美元的量级估算 token 预算；值得盘点现有自动化方案里哪些仍依赖人工设计分工。
- 依据：结果来自 Google DeepMind 团队实验（Faldor et al.），由 The Decoder 转述，暂无独立验证；1,620 美元按 2027 年才生效的标准价计算，非当前价格。
Oct 18, 2026
Four AI agents running on Google's Gemini 4 Argon have beaten Minecraft for the first time without human input, slaying the Ender Dragon in a fully autonomous run. The Google Deepmind team dropped the agents into a randomly generated world without assigning roles, strategies, or game-specific tools. They split up tasks on their own, warned each other about threats, recovered items from dead teammates, and logged mistakes in a shared memory. Over 27 hours and 14 minutes, the four agents wrote a total of 2,783 JavaScript programs, exchanged 229 chat messages, and died 605 times.
The entire session cost $1,620 at the standard Gemini 4 Argon price that takes effect in 2027. Minecraft is just the latest example of agentic AI models beating video games that used to require purpose-built AI systems trained from scratch just a few years ago.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
来源：The Decoder：AI News · the-decoder.com