Ordering an Uber or dinner might never be the same — but not because it’s faster. Gemini’s new task automation feature on Pixel 10 Pro and Galaxy S26 Ultra takes apps out of your hands, though it moves at a snail’s pace.
A Glimpse of AI Taking Control
Testing Gemini’s task automation feels like peeking into the future, even if that future is a bit slow and awkward right now. For the first time, this AI assistant can actually operate apps on your phone, handling tasks like ordering food or booking rides. But you won’t want to hand over your phone when you’re in a rush. Gemini’s pace lags far behind any human’s quick taps and swipes.
Still, the key here isn’t speed. This system is designed to work quietly in the background while you get on with other tasks. And it keeps going even when your phone’s screen is off, giving you a chance to check your bag for your passport without worrying about booking your ride simultaneously.
Slow and Steady, But Not Always Surefooted
Watching Gemini stumble through a food order is oddly captivating. For example, when asked to order a chicken teriyaki combo, Gemini correctly split the order into two half-portions to match the menu’s options. That shows a surprising level of understanding and adaptability.
But then it struggled to find a simple side of greens listed prominently on the screen, taking multiple wrong turns before finally getting it right. The whole process took about nine minutes — far from the instant gratification most users expect from their phones.
The interface offers a window to watch Gemini’s progress, with text updates like “Selecting a second portion of Chicken Teriyaki for the combo” appearing at the bottom of the screen. For those curious about AI in action, it’s a rare chance to see how the system reasons and adjusts on the fly.
Why You Still Need to Double-Check
Gemini doesn’t quite trust itself to hit the final confirmation button. That’s a good thing. So far, it won’t place an order without your final approval, which prevents accidental purchases and keeps control in your hands.
It’s a cautious approach that adds friction but makes sense given the current state of the technology.
Over several tests, Gemini hasn’t gone rogue or made decisions without user consent. And its accuracy is surprisingly high when it comes to interpreting menu options and ride selections. But the slow speed and occasional missteps mean you’re still better off tapping the screen yourself if you’re in a hurry.
Looking Ahead: The Road to Seamless Automation
Gemini’s task automation is still in beta and only supports a handful of food delivery and rideshare apps. That’s a tiny slice of what smartphones can do today. But it marks a major step toward AI assistants that don’t just answer questions but take real actions on your behalf.
Imagine a future where your phone handles complex errands while you focus on other things — booking rides, ordering meals, even managing appointments without lifting a finger. Gemini is far from that today. But it’s proof that AI can learn to deal with apps on its own, even if it’s clunky and slow for now.
The challenge will be speeding up the process and expanding the range of supported tasks without sacrificing accuracy. Plus, developers will need to ensure users remain in control so the AI doesn’t make unwanted decisions. For now, Gemini’s cautious design reflects that balance.
So while Gemini won’t replace your quick taps anytime soon, it’s a major step forward in AI-driven task automation — a slow, clunky, but genuinely impressive glimpse of what’s coming.
