Sources
Make it yours
Inspired by this story? Turn the idea into your own AI art in seconds — free to start, no card required.
Start creating freeDiscuss this with
Pick a companion and get their take on this story

Theo turns AI news into things you can actually try in tonight's session.
Inspired by this story? Turn the idea into your own AI art in seconds — free to start, no card required.
Start creating freePick a companion and get their take on this story
Google has updated its Gemini app for macOS with a new "computer use" feature, allowing the AI to interact directly with local files, desktop applications, and web services to automate multi-step tasks. This update positions Gemini as a more integrated AI agent, capable of executing complex workflows that previously required manual intervention, a significant shift for creators managing numerous digital assets and application-specific tasks.
This means that instead of manually dragging files, opening specific apps, and copying data, you could, for example, prompt Gemini to "gather all images from my 'inspiration' folder, resize them to 1024x1024, and upload them to my Charmloop gallery." For AI-art creators, this could translate into quicker asset management and less time spent on repetitive administrative tasks, freeing up more time for actual creation. Imagine generating a batch of images, then having Gemini automatically rename them based on your prompt, sort them into project folders, and even initiate an upscale process with a desktop tool.
The core of the "computer use" feature lies in its ability to bridge the gap between AI commands and desktop actions. Previously, AI assistants were largely confined to their own interfaces or web-based interactions. Now, Gemini can open applications, manipulate documents, and navigate file systems on your Mac. Google suggests this could be useful for tasks like speeding up workflows, filling in forms, and making presentations, but the implications for creative professionals are broader. A generative artist could use it to manage their extensive libraries of reference images, automate the tagging of newly generated art, or even script complex sequences involving multiple image editing tools.
For instance, if you're regularly generating character concept art, you might have Gemini:
.png files.This level of automation moves beyond simple text generation to actual command execution within your operating system, making the AI a proactive assistant rather than just a conversational partner. The ability to chain these steps together through natural language prompts could significantly cut down on the manual overhead of managing large creative projects. You can learn more about managing your prompts and generations on our /guides page.
Google emphasizes that the "computer use" feature operates with strict user permissions. Gemini will not access your applications or data without explicit consent, allowing you to control precisely which parts of your system the AI can interact with. This is crucial for creators who work with sensitive or proprietary content. Before Gemini takes any action, it will prompt you for approval, showing you the steps it intends to take, which is a necessary safeguard against unintended consequences or data exposure. This granular control means you can grant access for specific tasks without broadly opening up your entire system.
"Users will have full control over what Gemini can see and do on their Mac."
— The Verge
This update follows Google's broader strategy of integrating AI more deeply into everyday computing, much like Microsoft's push with Copilot. For those who frequently generate images or manage complex digital assets, this could be a powerful tool for optimizing their daily routines. It shifts the focus from purely generating content to actively managing and processing it within your existing digital environment. As AI models continue to evolve, expect to see more of these direct operating system integrations, further blurring the lines between AI assistant and personal computing agent. This move pushes closer to the vision of a truly intelligent workflow assistant, one that understands your creative intent and executes the technical steps to achieve it. You can explore a wide range of models and styles available for generating images in our /catalog.