US start-up Rabbit and tech company Teenage Engineering have teamed up to create the R1, a revolutionary “pocket companion” that aims to challenge the overwhelming presence of smartphones in our daily lives. By harnessing the power of artificial intelligence, this palm-sized, vibrant orange AI assistant is capable of completing tasks on behalf of its user, even if they involve multiple or intricate steps. Unveiled at the recent Consumer Electronics Show (CES) in Las Vegas, the Rabbit R1 offers a wide range of functionalities. From booking flights to streaming music and even editing Photoshop images, this versatile device is designed to cater to various needs.
To interact with their R1, users simply need to press and hold a button located on the right-hand side of the device, just like using a walkie-talkie. They can then issue commands in natural language and enjoy a simplified visual interface that portrays their assistant as a charming pixel art-inspired bunny.
Rabbit’s custom AI, known as the R1, is a form of Large Action Model (LAM) that surpasses the capabilities of the well-known Large Language Models (LLMs) used in chatbots like ChatGPT. While LLMs primarily generate text in response to human input, Rabbit’s AI goes a step further by generating actions on behalf of users. These actions can range from purchasing groceries online to booking taxis or tickets.
The concept of personal AI “agents” has gained significant attention online, but Rabbit asserts that their operating system is the first to incorporate such a LAM. The functionality of the LAM is based on its ability to comprehend users’ intentions and behaviors within specific applications, and subsequently replicate those actions. Unlike traditional approaches that require custom integrations through Application Programming Interfaces (APIs) for each app, Rabbit’s model is universal and can function seamlessly across various mobile and desktop environments.
Upon its launch, the R1 will already be equipped with training to work with the most popular apps. Additionally, Rabbit plans to introduce more functionalities in the future, ensuring continuous improvement and adaptation to evolving user needs. Furthermore, users will have the opportunity to train their personalized agents, referred to as “rabbits,” on more specialized or niche applications.
The device operates independently and does not require a smartphone connection for its functionality. In terms of its design, Rabbit collaborated with Teenage Engineering, a renowned company known for its innovative approach to music gadgets such as synthesizers and speakers. Together, they aimed to create a unique aesthetic with a nostalgic flair. Taking inspiration from the Tamagotchi, a popular Japanese digital toy pet, the company sought to make the device visually striking while maintaining its intuitive functionality. In addition to its 2.88-inch touchscreen display and touch-to-talk button, the physical components include a scroll wheel for navigating the display and a “rabbit eye” camera that rotates for computer vision purposes. This camera allows the device to perform tasks such as “looking” inside a refrigerator and identifying the ingredients present to suggest a recipe.
The operating system of Rabbit, represented by a bunny head avatar on the screen, exhibits various design choices that prioritize security. One such feature is the touch-to-talk button, which eliminates the need for the device to be in an “always listening mode” like traditional smart speakers, a concept deemed outdated by Lyu in his keynote. Additionally, the “eye” of Rabbit possesses a rotating capability that ensures the camera lens remains physically obstructed until the user specifically requests its activation. This serves as an effective measure to prevent unauthorized surveillance.
Rabbit also emphasizes its commitment to providing a high level of encryption, assuring users that they will always maintain awareness and control over the actions performed by the system. Notably, the device refrains from storing user credentials for third-party services, further enhancing security.
All processing tasks are executed within data centers rather than on the device itself. This approach not only contributes to the affordability of the Rabbit device, which retails for US$199 but also minimizes power consumption. However, it is important to acknowledge that this reliance on data centers places a significant demand on resources such as water and electricity. The substantial requirements of these data centers should be taken into consideration.
Lyu emphasizes that the Rabbit R1 device, with its advanced LAM technology, will eventually enable users to accomplish tasks that were previously unattainable on app-based phones. Nevertheless, it is crucial to note that the Rabbit R1 is not intended as a direct replacement for smartphones. Rather, it represents a new generation of AI-powered devices that are just beginning to emerge, distinguishing themselves from the app-based systems that have been in existence for over 15 years.
Since its debut at CES, Rabbit has successfully garnered pre-orders for over 30,000 units of the R1 device.