Rabbit’s r1, the AI assistant gadget, has released a new “beta rabbit” mode that aims to improve its conversational AI capabilities. This update allows users to give more complex or multi-step instructions to the device and enables it to ask follow-up questions when it requires clarification. For example, users can now ask the beta rabbit to suggest three books similar to “the power of now”, including page length, year of release, and ratings, and save that information as a note titled “reading list”. The beta rabbit can also provide summaries of the recommended books upon request.
While these new features may seem impressive, it’s important to consider their practicality in real-life scenarios. As occasional users of chatbots ourselves, we’ve found that tasks like setting travel itineraries or comparing product prices are often more convenient to accomplish on larger devices. Additionally, relying on the beta rabbit for book recommendations may not be the most trustworthy option, as the sourcing of these recommendations is unclear.
The update also includes improvements to the alarm and timer functions, which are certainly welcome additions. However, there are still instances where the beta rabbit’s capabilities fall short. For example, setting a timer for baking chocolate chip cookies without specifying crucial details like temperature, quantity, and cookie type could lead to culinary disaster. While the AI assistant can’t possibly know these specifics, it would be reasonable to ask for guidance on the appropriate baking time for a specific scenario.
Despite these limitations, users are eagerly anticipating the release of the highly anticipated “large action model” (LAM) feature. The company initially promised that this model, trained on phone and web app interfaces, would allow the r1 to autonomously navigate and accomplish user-chosen tasks. However, to date, the company has only demonstrated this capability in demos, and it remains unclear whether the LAM feature is truly distinct from regular action scripting or API usage.
Nevertheless, there is still optimism surrounding the potential usefulness of the r1 gadget. Although the author of this article hasn’t had many opportunities to utilize the device since receiving it for review, they have not yet dismissed it completely. They have reached out to rabbit for updates on the LAM feature and will provide further information if they receive a response.
