The personal assistant that many saw as the next step in the evolution of AI met with a botched release.
The Rabbit R1 caught me by surprise. Here I was, looking at new releases of LLMs every other day and their comparative performance and boom! An Apple Keynote-like presentation explaining the new “app-killer” AI assistant was right in front of my eyes. You won’t be using apps and your smartphone anymore. Heard that before with Humane’s Ai Pin.
How the R1 Failed
It was exciting stuff, not going to lie. A small handheld device powered by an LAM (large action model, LLM’s sibling that’s built not to generate text but to accomplish tasks), connected to the internet, with a camera, and able to use multiple APIs for getting cabs, ordering food, or booking hotels? Sign me up.
But they kind of overpromised. The launch didn’t happen when scheduled and when it did, the experience was … terrible, to say the least. You can do a single search on YouTube and find videos from LTT, MKBHD, The Verge, and more – all voicing the same opinion. The opinion? It doesn’t work.
Yep, the product is nonfunctional.
It cannot book cabs. It cannot identify objects correctly. In a way, it’s worse than opening the ChatGPT, Gemini, or any other LLM-based chatbot app on your phone manually. For all its sleek, minimalist approach to simplifying life and the promise to become a physical AI assistant or agent, it fell short of delivering any real value.
Shady Background
What’s more, the company behind this product used to be called by a different name and was previously invested in NFTs (2021). That’s never a good sign. To add to it, the founder said that the NFT thing (called GAMA) tied to an Unreal Engine 5 game was just a side project. The project will later on have you control a power grid with your NFTs and the founder said that they will launch comic books, MMORPGs, a store in a prime location, a TV show, and brain logic control (whatever that means).
Too much.
The gaming-NFT-power project in 2021 raised $6 million. People who looked into this found it to be more than a side project. Jesse Lyu, the CEO of that NFT project and now Rabbit, had big plans about all that, and it included Metaverse, crypto, Bitcoin 2.0 built on clean energy, and a bunch of other Web3 buzzwords.
But enough about the older project. The R1 from Rabbit, a completely different project, was also overhyped. The earlier buzzword-soup from the founder didn’t include AI much (as 2021 was a different time), but did say that the NPCs in the game will be powered by AI algorithms and that you’ll interact with them using your brain logic.
Current Situation
Anyway, fast-forward to today and they’re trying to fix it. Every review of the device is negative. I do follow a bunch of tech channels on YouTube. To quote their videos, here’s the general consensus (and these guys are professionals when it comes to reviewing technology products):
- Rabbit R1: Barely Reviewable (MKBHD) – video title.
- “It’s not ready yet.” – Linus on LTT.
- “Like the Humane Ai pin, this thing is just so underbaked. The R1 feels like it’s missing so many features … that make it what it’s supposed to be.” – Dave on Dave2D.
- This sucks (The Verge) – video thumbnail. David Pierce in his review said, “It just absolutely 100% does not work well enough to be worth your time or your money.”
- “Something seems wrong,” said Dan Barbera of MacRumors while complaining about the terrible battery life.
All of this is from roughly 3 weeks ago.
The interface might also just be an Android application, as per a story by Android Authority.