aifollow.news
Search
Google 发布 Genie 4，可用文字实时修改 AI 生成的互动场景
Automatically verified and published · Generated and evidence-checked automatically; not reviewed by a human.
Google announced Genie 4, a world model it says can generate more realistic interactive experiences, let users change them in real time with natural-language prompts, and generate audio. Experiences can last about two minutes, up from one minute with Genie 3. Google suggests film pre-visualization, game prototyping, and robot training scenarios as potential uses.
Google 今天发布了用于通过 AI 生成交互式体验的 Genie 4 世界模型。该公司表示，这一模型有 4 项主要更新：它能生成更逼真的世界，允许用户通过自然语言提示实时更新体验，能够生成音频，并在底层使用 Gemini 3.8 Flash 来辅助模拟。它还能创建时长约 2 分钟的体验，而此前的 Genie 3 模型只能创建 1 分钟的体验。Google 将 Genie 4 定位为一款工具，可用于帮助电影制作人进行预可视化、制作游戏创意原型，以及为机器人创建训练场景。但……请到 The Verge 阅读全文。
Name a person, company or topic. We find the sources and gather every new development in your own tab.