Computer use in Gemini 3.5 Flash

Google just announced that Gemini 3.5 Flash can now directly control apps, browsers, phones, and desktops as part of its main model. In plain English: the company wants its AI to do more than chat — it wants it clicking, typing, navigating, and handling office-style busywork for you. Google also says it added safety features, including optional stop signs for risky actions and checks for sneaky prompt attacks. Sounds slick on paper. The commenters? Not exactly standing and applauding.

The loudest reaction was a full-on backlash from people who think “AI using your computer” is a futuristic headache, not a breakthrough. One critic called the whole idea slow, insecure, error-prone, and expensive, basically accusing Google of building a money-burning machine that nobody will actually enjoy using. Others zoomed in on what they see as missing pieces: one user complained there’s still no support for easier app connections in the Gemini app, saying they had to jump to a rival tool just to do an Airbnb search with image filtering. Ouch.

Then came the benchmark drama. One commenter gleefully pointed to Google’s own chart, arguing Gemini still gets beaten by competitors while the graphic tries to look triumphant anyway. And of course, the peanut gallery delivered the meme line of the thread: “Will it skip Ads lol”. That one tiny joke may have captured the public mood best of all — less “wow, the future,” more “cool, but can it survive the internet?”