Rabbit R1 Large Action Model Underwhelmes
- John Warner
- May 4, 2024
- 2 min read
Updated: May 5, 2024

I took a flyer on buying one of the new Rabbit R1 devices, which was supposed to have a superpower energized by a Large Action Model (LAM). I looked forward to taking an LAM through its paces.
Large Language Models (LLM) are discussed in some detail in The Intelligence Tsunami (https://www.amazon.com/dp/B0CXH5DDRP). LLMs powering services like ChatGPT and Google Gemini excel at understanding and manipulating language. They can generate different creative text formats, translate languages, and answer your questions in an informative way.

The promise of LAMs is they will be able to understand human intentions and translate them into concrete actions. LAMs will break down complex tasks into smaller steps, interact with external systems like APIs or smart devices, and even control physical objects. That's the theory, anyway. The Rabbit R1 is underwhelming. The initial version they shipped only connects with four services: Spotify, Uber, DoorDash, and MidJourney. I don't use any of those services, so this isn't helpful to me. I thought the Rabbit R1 was going to allow me to log into services I do use and then learn how use those apps itself. Nope. It doesn't.
For example, here's a video promising that the Rabbit R1's LAM can choose the best options for a trip to London and then book the transportation and hotel for you. I entered the exact prompt used into the video into my Rabbit R1. It gave me a list of instructions for how to do it myself. It did not return the best flights and hotel, much less book the flight and hotel for me as I was expecting. When I specifically asked it to book my flight, it told me it couldn't do that.
The most interesting feature of the Rabbit R1 is that the interface is voice based. You press a button and verbally ask it a question, and it provides an audio response. If it really did allow you to access the apps you commonly use with your voice from one device, the Rabbit R1 would be very useful.
Unfortunately, it's not close. I agree with other reviewers who said it feels like the device was shipped before the product was complete. Rabbit, Inc. way overpromised and underdelivered. That's disappointing because I hoped the device would be a cool example of a LAM in action.
Let's talk.
Let's inspire your team and your organization to excel.
John Warner
864-561-6609




Mình thường xem mấy dự đoán như một kiểu giải trí trước khi ngủ, vừa thư giãn vừa so xem linh cảm của mình có giống người khác không. Hồi đầu toàn nghe bạn bè kể lại cho nhanh, nhưng về sau mình bắt đầu tự ghi chú những con số hay xuất hiện trong mơ hoặc mấy chi tiết cứ lặp đi lặp lại trong ngày, rồi đem ra đối chiếu. Có lần đọc một bài du doan xsmb, họ nói về lô gan với chu kỳ nên mình thấy cũng hợp lý ở mức tham khảo, thế là theo dõi thêm chứ không tin ngay lập tức. Mình thích kiểu viết có giải thích đàng hoàng, không làm…