Google Releases a New Windows Desktop AI Application - Easy to Search with Just Two Keys!

Amid the rapid global development of AI technology, Google officially launched its new AI search application "Google app for desktop" for Windows users on April 15. The release of this application means that users can search for information anytime and anywhere without relying on a browser, greatly facilitating daily use.

The "Google app for desktop" application is equipped with advanced Gemini AI technology. Users can quickly bring up the search interface by pressing the "Alt+Space" keys, input questions or keywords, and get instant search results. The application not only supports retrieving web information but also searches local files, installed applications, and files in Google Drive, making it an efficient local search tool.

In addition to its powerful search capabilities, the application integrates practical tools such as Google Lens and screen sharing. With Google Lens, users can easily select any content on the screen for search, such as translating text in images or solving math problems. The screen sharing feature allows users to select a specific window or full screen and directly ask questions to the AI, avoiding frequent switching between applications and improving operational convenience.

Currently, this application supports Windows 10 and above systems, but it has not yet released a Chinese version, and currently only provides an English interface. In addition, Google has revealed that it is developing a standalone Gemini application for macOS, aiming to compete with other AI desktop clients in the market, such as OpenAI's ChatGPT and Anthropic's Claude. This release marks an important step for Google in the field of AI super desktop applications, and it will bring more convenience and innovative experiences to users in the future.

ChatGPT Users Exceed 1 Billion, Female Users Account for Over 50% for the First Time

According to OpenAI data, ChatGPT's global weekly active users will exceed 1 billion, with a significant change in user structure. The proportion of female users increased from 20% at the beginning to over 50%, for the first time surpassing males, with about 500 million women using it regularly. This reflects that AI technology is accelerating its popularization.

Opera Launches Browser Connector Feature for Seamless Integration of AI Chatbots

Opera browser introduces a free new feature called 'Browser Connector,' which allows users of Opera One and Opera GX to directly access AI chatbots such as ChatGPT and Claude through the Model Context Protocol (MCP) developed by Anthropic, enabling more convenient integration of AI tools while browsing.

High-end AI company Upscale AI plans to raise new funding with a valuation of 2 billion USD

Artificial intelligence infrastructure company Upscale AI is in talks for a new round of funding, targeting a valuation of 2 billion USD and aiming to raise between 180 million and 200 million USD. This will be its third round of funding in seven months, following a 200 million USD Series A funding and a 100 million USD seed funding. Investors include well-known institutions such as Tiger Global.

Perplexity Launches AI Assistant for Mac: Runs 24/7 and Can Perform Computer Tasks on Behalf of Users

Perplexity releases an AI assistant app for Mac, which can directly access the file system and native applications, supports text and voice interaction, can be activated by shortcut keys, perceives the user's current window, and proactively provides operation suggestions, promoting the deep integration of AI interaction into the operating system.

Google DeepMind's New Research TIPSv2: Let AI Truly Understand Images, Not Just Take a Quick Glance

Google DeepMind's latest research reveals the 'global strong, local weak' shortcoming in AI visual models and proposes the TIPSv2 solution. This approach improves the training method, enabling the model to more accurately locate local image details, such as identifying the position of a panda's left hind leg, solving the long-standing challenge of vision-language models in fine segmentation tasks.