Indeed, a virtual assistant is likely already tucked away in your phone. Honestly, I’d be surprised if you didn’t have at least one smart speaker lying around your house that you could ask to do simple things just by reading Engadget. Rabbit, a new firm, appears to disagree, though, and thinks that Siri and Alexa aren’t the best uses of artificial intelligence.
Showing at CES 2024 in Las Vegas this week, the business imagines a future where communication takes the place of apps and, instead of a distracting device pushing icons on you, you converse with an AI using what is essentially a walkie-talkie.
An undeniably cute little square in an endearingly vibrant shade of orange, the R1 is Rabbit’s first device to be unveiled. Teenage Engineering’s design geniuses have created an aesthetically pleasing virtual assistant device, which is impossible to ignore even if you disagree with the need for such a device.
It has a tiny 2.88-inch touchscreen, an analogue wheel for navigating, two microphones, a speaker, and a “360 degree rotational eye,” which is basically a fancy term for a camera that you can turn to face you or the rear of the phone.
However, the “Push-to-Talk” button is the main method of interacting with the R1. By doing so, you instruct Rabbit OS to begin listening. As you speak to it or assign it a duty, its highly stylized and dematerialized head bobbles slowly before it swiftly begins working.
Require an Uber? Are you looking for a way to go through your fridge’s leftovers? Curious about the identity of the sampler for “That Lady” by The Duets? (“Beastie Boys,” “Basement Jaxx,” and “Kendrick Lamar” are the correct answers, right?). At least in the controlled video demo, the R1 appears to be more than capable of tackling those chores.
The Large Action Model (LAM) is what allows Rabbit OS to take on such jobs. This is the main innovation that the company’s creator and CEO, Jesse Lyu, is promoting. Interfaces, not application programming interfaces (APIs) or apps, are its primary means of interaction.
Simply put, it can be programmed to do about anything a user can do with an interface. It’s similar to a more refined macro in certain ways. In order to showcase its powers, Lyu instructs the R1 via Discord on how to create an image using Midjourney.
After Lyu walks through the process and records it with Rabbit OS, it can do it again when instructed to. As a kind of privacy shutter, the revolving camera automatically points upwards toward the body. When summoned, it will only turn its sensor towards the target. Among its more common capabilities is the ability to recognize real-world objects and people (within reasonable limits, of course).
People will be intrigued by how it works with the AI, though. During the demonstration, Lyu asks the R1 to make a “low in calories” recipe recommendation by analyzing the contents of a full refrigerator.
Naturally, many questions regarding the R1 remain unsolved. Does it have a good battery life? They say it’s “all day,” but how long does it end up being? It also depends on how easy it is to train for the typical user. Nonetheless, we do have some information. We have information that it may be pre-ordered today for $199 and is anticipated to be shipped in March or April.
You might also be interested in these tech posts if you liked this one: