Rabbit, an AI startup, has unveiled the Rabbit r1, a mobile device aimed at simplifying the smartphone user experience. The r1 is powered by the Rabbit OS, an operating system that leverages a Large Action Model (LAM). This model is designed to learn how users interact with computers, specifically the apps they use. The r1 is capable of understanding complex user intentions, operating user interfaces, and executing actions through voice commands. Trained on data from users interacting with apps, the r1 can perform tasks like ordering food, booking rides, playing music, and more.
Interacting with the Rabbit r1 is achieved through a push-to-talk button, and it understands natural language requests. Notably, the device doesn’t have traditional apps and relies on internet connectivity to perform tasks. Developed in collaboration with Teenage Engineering, the Rabbit r1 features a 2.88-inch touch screen, a rotatable camera for photos and videos, a scroll wheel for navigation, and a USB-C port for charging. It runs on a 1,000mAh battery, with battery life depending on usage. The device requires a SIM card for cellular data but doesn’t have a subscription associated with it.
Users can teach the Rabbit r1 to perform tasks, including multi-step actions like researching locations, booking hotels, securing restaurant recommendations, and more. Once a task is learned, the r1 can repeat it independently. The device is claimed to interact with apps like Photoshop, executing multi-step actions within the software. To complete tasks involving specific services, users need to provide the r1 with their logins for those services. The Rabbit r1 is available for pre-order on the Rabbit website at $200 and is expected to ship in March or April 2024.