软餐获悉，谷歌对外演示了将在其搜索引擎中集成的生成式 AI。该功能首先在美国推出。谷歌演示了如何基于植物的照片和搜索查询 “如何给它浇水？”，生成式 AI 会提供具体答案。这项功能仅适用于安卓上的 Google Lens 应用程序，且仅支持美国英语。谷歌没有它依赖哪个语言模型，Google 高级产品经理 Lou Wang 明确表示不是 Gemini 模型。他表示，并非每次搜索都会使用生成式人工智能。生成式 AI 旨在从不同的来源和不同类型的内容（如文章、视频和照片）中收集答案，提取知识并进行总结。
除了生成式 AI，谷歌还推出了 “圈选搜索”（Circle to Search）功能。用户可以长按主页按钮或手势导航栏，然后呼叫 Google，并在屏幕上圈选文本或对象。然后，该部分将成为搜索查询的一部分，除了剪切的图像外，还可以包含文本。圈选搜索将在全球范围内推出，但目前仅适用于 Google Pixel 8 手机和本周发布的三星 Galaxy S24 手机。
Google demonstrated the integration of generative AI in its search engine. This feature is first being launched in the United States. Google showcased how, based on a photo of a plant and the search query “how to water it,” generative AI can provide specific answers. This functionality is only available on the Google Lens application for Android and supports only U.S. English. Google did not disclose which language model it relies on, and Senior Director of Product Management, Lou Wang, explicitly stated that it is not the Gemini model. He explained that generative AI is not used for every search. Its purpose is to gather answers from various sources and different types of content such as articles, videos, and photos, extract knowledge, and summarize it.
In addition to generative AI, Google also introduced the “Circle to Search” feature. Users can long-press the home button or the gesture navigation bar to activate Google and then circle text or objects on the screen. This selected portion becomes part of the search query and can include both clipped images and text. Circle to Search will be rolled out globally, but currently, it is only available on Google Pixel 8 phones and the recently unveiled Samsung Galaxy S24 phones.
微信扫一扫