前端UI的工作,被谷歌AI一夜干没了。
就在最新能力展示中,Gemini 3.0 Pro居然自己“捏”出了一个macOS。
只靠几行提示词,谷歌的AI就在浏览器里复刻出了苹果引以为傲的UI。
Prompt: Design and create a web os like mac os full functional features from text editor , to dile manager to paint to video editor and all important mac os pre bundled software Use whatever libraries to get this done but make sure I can paste it all into a single HTML file and open it in Chrome.make it interesting and highly detail , shows details that no one expected go full creative and full beauty in one code block.
苹果式动画、窗口最小化、工具栏、浏览器……一应俱全,并且所有功能都能正常运行。
甚至还有个彩蛋——
在终端输入“matrix”,就会触发《黑客帝国》同款炫酷特效。
重点在于成功率!
因为在A/B测试中不可能重复生成,以上你看到的所有都完完全全是一次生成的结果,并且源代码已经在CodePen上公开。
网友惊呼,如果正式版真能达到这样的水准,Gemini无疑将成为史上最强的编程型模型。
万物皆可被UI
短短一天之内,苹果、微软、Linux这三足鼎立的PC操作系统江湖就被Gemini 3.0 Pro全包了。
有网友尝试让Gemini 3.0 Pro生成一个网页版Windows,结果一次成功。
而且,它不仅在终端里内置了Python,还能玩游戏、运行代码,功能完整得离谱。
Prompt: Design and create a web os like windows os full functional features from text editor , terminal with python and code editor and a game that can be played to dile manager to paint to video editor and all important windows os pre bundled software Use whatever libraries to get this done but make sure I can paste it all into a single HTML file and open it in Chrome.make it interesting and highly detail , shows details that no one expected go full creative and full beauty in one code block.
没有对比就没有伤害——
有网友展示了相同提示词下,Claude 4.5 Sonnet的「杰作」,连应用图标都点不开。
Linux自然也跑不掉。
结果依然不错,还可以上维基百科、调用计算器、更换壁纸。
Prompt: Create a fully functional Linux desktop environment (Ubuntu/GNOME style) as a complete web operating system in a single HTML file with embedded CSS and JavaScript. All applications must be fully functional Use whatever libraries to get this done but make sure I can paste it all into a single HTML file and open it in Chrome.make it interesting and highly detail , shows details that no one expected go full creative and full beauty in one code block.
看到这些,硅基生命派表示非常满意:
谁说LLM只会鹦鹉学舌的?!
毕竟macOS和windows在网上连行源代码都没有。
但也有网友指出,这只是一个模拟环境,距离真正的OS还差得远。
确实令人印象深刻,但我们必须区分“仿真”和“实现”的区别。这些其实只是“粘合代码”,底层依赖的都是现有技术。就像程序员不能因为调用了库就把功劳算在自己头上一样,AI也不能。
One More Thing
虽然谷歌尚未公布Gemini 3.0 Pro的正式发布时间,但从以往的泄露节奏和发布规律来看,有业内人士推测它可能会在未来几个月内亮相。
一个明显的信号是:网上来自各路influencer的演示视频越来越多了。
不知怎么的,我们在Gemini 3正式发布前看到的demo,比当初2.5 Pro还要多。
上一轮NanoBanana的营销成功,显然让谷歌尝到了“预热”的甜头,这次或许又想再赌一把。
不过,期望值拉得越高,落差往往也越大——就像当初的GPT-5那样。
感兴趣的朋友们到CodePen操作起来啦~
macOS:https://codepen.io/ChetasLua/pen/LEGzZaQ
Windows:https://codepen.io/ChetasLua/pen/yyezLjN
Linux:https://codepen.io/ChetasLua/pen/LEGzZaQ
参考链接:
[1]https://x.com/chetaslua/status/1977936585522847768
[2]https://www.ainewshub.org/post/google-gemini-3-pro-rumors-release-date-features-and-what-to-expect-in-late-2025?utm_source=chatgpt.com
本文来自微信公众号“量子位”,作者:Jay,36氪经授权发布。