谷歌近日在 Search Labs 平台中升级了一项新的“生成式搜索体验”（Search Generative Experience）功能，该功能可根据用户输入内容生成对应图片。这意味着用户现在可以直接在相关搜索页面中进行测试。
例如，用户输入“请画一张戴着厨师帽并正在做早餐的水豚图像”并进行搜索，搜索引擎便会生成 4 张符合描述的图画，用户点选其中任何一张，就能看到它们对应的关键词，也可以编辑这些关键词，来修改画面。
据悉，此类由 AI 生成的图像，都会带有水印，以标注相关图片源于 AI，以防止用户被误导。
谷歌今年 5 月发布的测试平台 Search Labs 主要用于测试各种利用生成式 AI 来改善搜索功能的点子，但谷歌此前曾表示，部分功能最后不会在引擎中正式添加。
虽然部分功能不会在引擎中正式添加，但这一升级将有助于谷歌更好地利用其强大的 AI 技术来改善搜索体验。
新闻翻译：
Google has recently upgraded one of its new features in Search Labs platform, which can generate images based on user input. This means that users can now test the feature directly in the relevant search page.
For example, when a user enters “Please draw a picture of a rabbit wearing a chef’s hat and making breakfast” and conducts a search, the search engine generates four images that match the description. Users can select any one of them and see the corresponding keywords, as well as edit these keywords to modify the image.
It is worth noting that the images generated by AI are marked with watermarks to indicate that they are generated by AI to prevent users from being misled.
Google’s test platform Search Labs was released in May this year and is used to test various ways to improve search functionality using generative AI. However, Google has previously stated that not all features will be added to the engine.
Despite the fact that not all features will be added to the engine, this upgrade will help Google better utilize its powerful AI technology to improve search experience.
【来源】https://www.ithome.com/0/724/831.htm
Views: 1