Following OpenAI and Anthropic's release of Mac-specific AI applications, Google seems poised to seize control of Apple users' desktops. According to...Bloomberg NewsGoogle is currently testing a standalone macOS version of the Gemini app externally. It will not only bring a complete AI assistant experience, but also is expected to introduce a new feature called "Desktop Intelligence," which will allow Gemini to directly "see" and read the context information of the user's screen and open applications, enabling a more seamless and personalized AI collaboration experience.
Breaking Webpage Limitations: The All-New "Desktop Intelligence" Makes AI Think Like You
For a long time, computer users have primarily accessed Gemini's services through web browsers. The launch of the standalone application will allow Gemini to officially enter the underlying realm of operating systems, directly competing with ChatGPT and Claude, which already have Mac versions.
The biggest highlight of this Mac version of Gemini is undoubtedly its screen-aware feature, known as "Desktop Intelligence." According to a description of the original code cited by Bloomberg News: "When you enable desktop intelligence for an application, you are allowing Gemini to see what you see (such as screen context) and extract content directly from those applications to improve and personalize your experience as you use Gemini."
This means users no longer need to go through the tedious process of copying and pasting text or uploading screenshots. When reading a long PDF or writing code, Gemini can directly sense the current work window and provide accurate summaries or modification suggestions based on the content displayed.
Beyond simply "understanding," could it potentially enable "operation on behalf of clients" in the future?
The ability to read screen content is already available in Claude and ChatGPT's macOS applications, while Gemini has long had a similar screen-sensing function on mobile devices (Android platform).
However, the industry is more concerned about whether the Mac version of Gemini has the ability to "take action." While it's currently uncertain whether this version can function like Anthropic...The widely popular "Claude Cowork"orThe newly launched "Dispatch" featureIn that case, the mouse and keyboard would be directly taken over to automatically perform tasks; however, considering that Google has already provided a limited agent operation experience on smartphones, it is foreseeable that such agent functions will be extended to desktop operating systems in the future.
Entering the external testing phase, forming a deep alliance with Apple Intelligence
The report indicates that the Gemini application is currently open to external testing by non-Google employees, which is usually an important signal that the product is about to be officially released.
Interestingly, regardless of the final market response to this standalone application, Gemini's technological DNA will inevitably be deeply embedded in future Mac computers. Back in January of this year, Google and Apple had already...AnnounceGoogle's Gemini model will be used as...driveOne of the core driving forces behind the future version of Apple Intelligence.
It is rumored that Apple is completely overhauling Siri, making it more like a chatbot with deep conversational capabilities, and the key driver behind this is very likely Gemini.
