Rabbit Unveils AI Assistant, Plans Launch of Camera Wearable

WIRED attended the business’s launch event, in which customers received the first R1s and discovered that a companion AI device called an “all-seeing” was on its way. The entrance to Rabbit’s location for the R1 launch event—a device powered by artificial intelligence that was introduced at CES 2024—was lined with vintage technology.

The orange JVC Videosphere was initially released, followed by the Sony Walkman, the Tamagotchi, the transparent Game Boy Color, and the 1998 original Pokédex toy. The Pocket Operator by Teenage Engineering was located at the end of the space, and next to it were a few Rabbit R1 idea prototypes.

Given that it is only a few years old, the Pocket Operator shines out. It is because Teenage Engineering, a Swedish design firm, assisted in the creation of the R1. Furthermore, Jesper Kouthoofd, the person who founded Teenage Engineering, joined Rabbit as a corporate, creative officer (while also holding his position as President of TE), Rabbit Chief Executive Officier Jesse Lyu stated on stage at the opening ceremony.

It’s not simply that this small red retro device is appealing to our nostalgia and drawing inspiration from the latter part of the 20th century in terms of design. Additionally, Rabbit is adamantly asserting that its R1 merits a place in this technology house of merit.

The venue, the renowned TWA Hotel at John F. Kennedy International Airport in New York, which features a reestablished Lockheed Constellation “Connie” L-1649A that has been converted into a cocktail lounge and experiences a lot like a live-action set of The Jetsons, further highlighted the desire for the exhilarating promise of possible future praise and the thrilling technology of the past.

The device you will remember fifty years from now—the R1—is the device that marked the beginning of the AI-powered era. That’s Lyu’s hope, though.

Declaring to a fascinated audience on stage, “Our goal is to develop the most basic computer,” Lyu spent an hour outlining strategies for achieving this objective. As it happens, the R1 is barely the beginning.

When I initially reported on the latest AI gadget from CES, my focus was on its minor aspects, like its orange-red colour, small size (about the size of a stack of Post-its), and small screen. A scroll wheel to navigate the interface is located on the right side of the screen, and just above it, there’s a cam which can be turned to face the front, rear, or even inside the device’s inner case for privacy. The primary button for selecting something on the display is located on the right edge.

You can ask questions about this device in the same way you would using Siri, Alexa, or Google Assistant, just like you would use the Humane AI Pin. Because of its intelligence, you can naturally ask sophisticated questions about it. Perplexity’s extensive language model powers this feature, enabling it to comprehend your inquiry and provide the appropriate response through speakers and a screen. To turn on the microphone, press a side button and speak, just like on a walkie-talkie. No “hot word” is required.

You can direct the lens at an object, and the R1 will be able to comprehend and interpret what you’re looking at, thanks to the camera’s ability to enable vision-powered AI tricks.

In a live demonstration, Lyu held up the R1 camera and directed it toward a spreadsheet that was written on paper. He instructed the gadget to send him a duplicate of the picture and to move two columns around as it took it. Shortly after, Lyu’s computer’s inbox filled with an email containing a digital copy of the spreadsheet and ostensibly the request addressed.

There are plenty of other built-in tricks with the R1. Using the “Rabbit Hole” online portal, you can take notes, see them, and even change them. Once more shown live on stage, it is capable of handling languages in simplicity. The device allows you to record voice messages, and the web gateway will enable you to view the WAV file and receive an artificial intelligence synopsis of the session. Whenever you are required to enter something into the user interface, such as a Wi-Fi password, there is an integrated virtual keyboard.

Lyu also demonstrated the R1’s Teach Mode, which allows you to direct the R1’s camera at an operating system screen while giving it instructions. You may request it to complete the task after it has learned it, saving you the trouble and time. But this function isn’t accessible yet, and Rabbit claims a limited group of users will be the first to beta test it.

However, the R1’s main objective is to replace your apps essentially. Press the button to instruct the R1 to deal with an issue rather than searching for an icon.

During the CES event, it appeared like the R1 would support numerous third-party apps at launch; however, as of right now, the list of available services is limited to Uber, DoorDash, Midjourney, and Spotify.

You can ask the R1 to make phone calls, order food from McDonald’s, create images, play music, or link with these features through the Rabbit Hole web portal. Yes, you are entering your credentials and accessing these services through what appears as a virtual computer hosted by Rabbit. Although it does not use software programming interfaces (APIs) for these services, it was programmed correctly to make use of these websites.

Naturally, Lyu says there will be much more in the near future. We are informed that functions, including a wake-up call, timetable, contacts, GPS, recalling memories, and trip planning, will be available in the summer. Amazon Music and Apple Music interfaces are currently under development, and additional third-party services integration, such as OpenTable, Lyft, and Airbnb, should be visible in the future.

It looks such as a mobile device, so you might be thinking, “Hold on a minute,” and you are probably incorrect.

All these kinds of jobs can be completed by a smartphone more effectively, more quickly, and with deeper interactions—as demonstrated by the cumbersome and constrained Humane Ai Pin. This is the point when you need to start examining Rabbit’s general vision closely.

Speaking first, then calculating is the plan. Applications are not necessary—the computer will comprehend on its own. Though it is still a long way off, Rabbit hinted at a wearable gadget that could recognize objects by pointing at them during the launch event.

Without having to use the words “Nest” or “thermostat,” Lyu proposed that the device might be able to comprehend you indicating to a thermostat made by Nest and requesting to lower the temperature. However, the alleged all-seeing wearable’s image was blurry, so there isn’t much information available.

After bringing up generative user interfaces, which allow users to design their user interface with buttons on a screen positioned at the ideal viewing angle, Lyu said that Rabbit is developing Rabbit OS, an AI-native workstation operating system. Again, only a little is known, but my first thought was of Theo in Her setting up OS1 on his personal computer.

An OS that prioritizes the use of a personal voice assistant. How could something go wrong?

The Rabbit R1 is available for 199 dollars and is already available for purchase, but because of batch shipping, orders placed today will receive a unit in June.

 Lyu is constantly saying this particular device does not require a subscription, compared to the Humane Ai Pin. Still, it’s important to remember that in order for the 4G-enabled R1 to be functional when you are not near Wi-Fi, you’ll need to buy a separate monthly data subscription and put in a mobile SIM card to the device (as long as you intend to attach it to your smartphone).

According to the corporation, sales of 100,000 units were made during the initial quarter of 2024. I unboxed my unit after picking it up at the event. It’s definitely a charming piece of technology, based on my initial impressions, but after using it longer, I will have more to say in a review.