LeituraPesquisaRadarFramework de investimento
Entrar / Cadastrar
← Voltar para a leitura
NotíciasTópico · Gemini通关MCMaterial1 min de leituraEditorial DisconfirmAI

Google DeepMind Reports Four Gemini Agents Autonomously Beat Minecraft

Google DeepMind team shows four Gemini 4 Argon-based AI agents beat Minecraft's Ender Dragon in 27 hours without human intervention, demonstrating long-horizon planning and collaboration.

2026-10-20 11:30 GMT+8

Google DeepMind researchers recently published results showing that four AI agents based on the Gemini 4 Argon model successfully defeated the Ender Dragon in the sandbox game Minecraft without real-time human intervention. This marks the first time an AI has achieved end-to-end autonomous completion under conditions with no pre-scouted maps and no manually written custom tools.

The experiment lasted 27 hours and 14 minutes, during which the AI agents spent approximately $520 in compute tokens, relying entirely on the open-source Mineflayer library to write and execute 2,783 JavaScript code snippets to control character actions. Researchers observed via dashboard only, without interference.

The AI exhibited "emergent behavior," spontaneously evolving division of labor, with some agents crafting tools, others gathering resources, and others building defensive structures. Despite the final victory, the team died 605 times, and 22% of the executed code failed. The experiment demonstrates potential for AI collaboration in complex multi-step tasks, offering reference value for real-world software engineering automation.

Fontes

Entrar / Cadastrar
LeituraPesquisaRadarFramework de investimento
technews.tw
↗
Tópico

Gemini通关MC

美国谷歌旗下DeepMind团队近日宣布,四个基于Gemini 4 Argon模型的AI代理在无人类干预下,耗时27小时14分钟在《我的世界》中击败最终Boss“末影龙”。这是史上首次通用型AI代理仅依靠开源库Mineflayer,通过编写…

4 etapas0 peçasAtualizado 2026-10-21 03:51

Como chegamos aqui

  1. 2026-10-15Google DeepMind提交论文《A team of Gemini 4 Argon agents beats Minecraft》,记录四个AI代理自主通关过程。post.smzdm.com ↗
  2. 2026-10-18The Decoder报道Gemini 4 Argon代理在无人类帮助下击败Minecraft,强调其自主分工与协作能力。the-decoder.com ↗
  3. 2026-10-19Engadget详细披露实验数据:27.23小时、2783个程序、605次死亡,以及约810至1620美元的成本估算。engadget.com ↗

O que ainda está indefinido

  • 这种长周期自主规划能力能否有效迁移到真实的企业级软件开发任务中?
  • 随着Gemini 4 Argon入门期促销结束,标准定价下的算力成本是否会阻碍此类多代理协作的普及?
  • 在更复杂的非结构化环境中,AI代理如何进一步降低22%的代码执行错误率?
Ler o tópico completo →
Continuar pesquisando →