I'm Not Sure What It Is, But I Want It

Is it a smartphone? No, that is not quite it. A digital assistant? Sort of, but that is not exactly right either. Maybe a Tamagotchi? The r1 pocket companion from AI startup rabbit inc. (which seems to strongly dislike capital letters) is hard to define, but I definitely want to get my hands on one. The r1 is in a category of its own, as a digital assistant of sorts with a very unusual natural language operating system that is meant to take the friction and frustration out of voice-based interactions.

The 115 gram r1 is about the size of a stack of Post-it notes, and comes equipped with a 2.88 inch touchscreen display, a camera, microphone, button, and scroll wheel. The hardware specifications are decent with a 2.3 GHz MediaTek Helio P35 processor, 4 GB of memory, and 128 GB of storage, but do not come close to rivaling the latest smartphones. Instead, much of the processing takes place in the cloud (Wi-Fi and cellular connectivity options exist). That can certainly raise some red flags for many consumers concerned with privacy, but unlike most of the smart devices that we have grown accustomed to, the r1 is not always listening in on us. Rather, the user must be intentional, pressing a button to initiate a voice-based interaction. Further, the camera rotates to a position where the lens is physically blocked until the user requests an action that requires its use.

So far, the r1 looks like a low-cost smartphone, albeit with a very slick design. But the real innovation is in the Large Action Model on which the operating system is based. Its purpose is to ease human-machine interaction by observing how humans interact with a variety of user interfaces, then mimicking that process. With that sort of understanding, rabbit OS can translate natural language requests into actions that can be taken in a wide range of apps, with varying interfaces.

These actions are carried out by personalized AI agents called “rabbits” that run in a cloud-based environment. Of course most apps do not have an API to carry out all of the tasks that are done within them, so the apps run in this environment where the rabbits interact with them via their normal interfaces behind the scenes to carry out the user’s wishes, then return the result to the r1. The developers call this setup an app-free experience, but in reality, the user still uses all of the same apps that they currently do, but the details are hidden behind rabbit OS. In this way, a user of the r1 could, for example, research and book travel plans, place online shopping orders, and more, all through the same voice-based interface.

A web portal called “rabbit hole” enables users to give rabbits access to their existing apps and enable a variety of features. Many popular apps are already available for use with rabbit OS, and there is also an experimental feature available that allows users to train their own rabbits to use other apps that are not yet supported by rabbit OS.

The rabbit r1 sounds intriguing. It could potentially transform the way that we interact with our devices, or it could fade away and become a footnote in the history of human-computer interaction. We will have to wait for the official launch in late March of this year to find out which way it goes. The r1 is now available for pre-sale at $199 if you want to be among the first group of people to give it a whirl.