谷歌重磅发布Gemini 2.5 Computer Use - AI代理首次实现真正的计算机控制

就在今年早些时候，谷歌曾承诺要通过Gemini API为开发者带来计算机使用能力。如今，这个承诺终于兑现了。谷歌DeepMind正式发布了Gemini 2.5 Computer Use模型，这是一个基于Gemini 2.5 Pro视觉理解和推理能力构建的专用模型，能够驱动AI代理与用户界面进行真正的交互。
腾讯网

Gemini 2.5 Pro能力再进化：谷歌推出专用计算机操作模型，让AI智能体丝 ...

谷歌刚刚发布了一项更新：正式推出Gemini 2.5计算机使用模型（Computer Use model）这是一款基于Gemini 2.5 Pro视觉理解与推理能力构建的专用模型，旨在赋予AI智能体（agent）与图形用户界面（GUI）直接交互的能力 ...
腾讯网

清华团队推出轻量级端侧GUI Agent，用代码生成技术提升任务效率

从感知式 AI（理解图像、文字和声音）到生成式 AI（创造文本、图像和声音），再到能够感知、推理、计划和行动的智能体（即 AI Agent），我们正见证着 AI 能力的下一代进化。 Claude Computer Use、OpenAI Operator、Manus 等这些能够操控电脑、手机等终端设备的大语言 ...
新浪网

性能逼近闭源最强，通义实验室开源Mobile-Agent-v3刷新10项GUI基准SOTA

覆盖桌面、移动和 Web，7B 模型超越同类开源选手，32B 模型挑战 GPT-4o 与 Claude 3.7，通义实验室全新 Mobile-Agent-v3 现已开源。 一眼看到实力：关键成绩速览。 GUI 智能体，就像你的跨平台虚拟操作员，能看懂屏幕、点鼠标、敲键盘、滑手机，在办公、测试、RPA 等 ...
Ars Technica

Y Combinator’s Xerox Alto: restoring the legendary 1970s GUI computer

[url=http://arstechnica.com/civis/viewtopic.php?p=31444501#p31444501:2qxdzs2l said: tlhIngan[/url]":2qxdzs2l]Xeros was given the shares as a in-kind payment for the ...
Computer Weekly

NordVPN goes full GUI on Linux

The latest trends and issues around the use of open source software in the enterprise. Arguably best-known VPN company around right now NordVPN has updated its Linux application. Updates feature the ...
The Business Journals

Goodbye GUI? Ambient orb a computer 'mood ring'

Ambient Devices in Cambridge has won a big distribution deal for two products that executives say could change the way computers and people get along. Later this year, Nashua, N.H.-based Brookstone ...
The Verge

40 years of icons: the evolution of the modern computer interface

Fifty years ago, the word “computer” had a very different meaning. Prior to World War II, the word referred not to machines, but to people (mostly women in order to save costs) hired as human ...