The competition to build the iPhone of artificial intelligence is heating up. On Tuesday, technology startup Rabbit unveiled its contender: the Rabbit r1, a small, orange, walkie-talkie style device equipped with AI agents that can carry out tasks on behalf of the user. In a keynote address at the Consumer Electronics Show in Las Vegas, Rabbit’s founder Jesse Lyu showcased the device’s capabilities, including planning vacations, booking trips, ordering food, and generating images. The r1 joins a growing market of AI-first devices that prioritize natural language interactions without relying on screens or app-based operating systems. Priced at $199, it offers a more affordable alternative to competitors like the Humane Ai Pin and Meta and Rayban smart-glasses. With prominent tech investors like OpenAI’s CEO Sam Altman and Softbank’s Masayoshi Son backing these advancements, the race to design the ideal hardware form factor for AI is underway.
Rabbit’s r1 stands out for its utilization of a large action model, a new type of AI system. Unlike large language models, which excel at understanding user intentions but struggle to execute real-world actions, the r1 is trained on graphical user interfaces. This training allows the device to navigate websites and apps designed for humans and successfully complete tasks on their behalf. To utilize the r1’s capabilities, users must sign into their various accounts via Rabbit’s web portal. Rabbit’s AI agents, called rabbits, run on external servers to execute actions through the designated accounts. The company assures users that each individual has a dedicated and isolated environment on its secure servers and that it does not retain user passwords.
While Rabbit claims to work with top industry partners in natural language intelligence, their privacy policy remains unclear about the disclosure of these partners’ identities. Additionally, the policy retains the right to send user data to third parties for data processing, raising concerns about privacy and data security.
Critics question the necessity of new AI-first devices, arguing that consumers simply need more intelligent phones rather than additional hardware. OpenAI’s Altman has also expressed his intention to enhance OpenAI’s software to eliminate the need for dedicated AI devices. However, the trend towards AI agents with increased agential capabilities has raised safety concerns. Powerful agential AIs could pose risks if they pursue dangerous goals, leading to potential harm to humanity. As society becomes increasingly reliant on complex networks of interacting AI agents, issues like inescapable feedback loops and goal drift may further exacerbate these risks.
In an era where personalized computing and AI advancements are at the forefront, the quest for the iPhone of AI continues. Startups like Rabbit, Humane, and others aim to redefine user interactions with AI through innovative hardware and AI-first devices. As the race intensifies, entrepreneurs, tech investors, and industry leaders strive to strike the right balance between empowering AI agents and ensuring safety and security for users and society as a whole.