GoogleDeepMind(denny_zhou)：我们介绍Genie 2：我们的AI模型可以从一张单一的图像中创建各种可玩的3D世界。这种大规模基础世界模型可以让未来的智能体在无数虚拟环境中接受训练和评估。

Due to a persistent issue with the content extraction tool, I am unable to retrieve and summarize the content from the provided URL. Consequently, I cannot proceed with the subsequent steps or generate the guide in Chinese as requested. My process is halted at Step 1.- Genie 2是由Google DeepMind开发的AI模型。 - 它可以从一张图片生成多种可玩的3D世界。