The Rabbit r1 shipped half-baked, but that’s kind of the point

I finally received the rabbit r1 (the company insists on this lowercase styling) I’ve been writing about since its debut at CES in January. And I was able to tell within about 30 seconds of turning it on that it was shipped a couple months too soon — but honestly…that’s fine? This AI gadget is […]
© 2024 TechCrunch. All rights reserved. For personal use only.

​

I finally received the rabbit r1 (the company insists on this lowercase styling) I’ve been writing about since its debut at CES in January. And I was able to tell within about 30 seconds of turning it on that it was shipped a couple months too soon — but honestly…that’s fine? This AI gadget is weird, relatively cheap, and obviously an experiment. To me, that’s something we should be rallying behind, not dunking on.

The actual issues with the r1 are obvious: it doesn’t have enough app integrations, and it “could just be an app.”

As to the first problem, well, it’s completely true at present. There are only four things to connect to: Uber, DoorDash, Spotify, and Midjourney. Leaving aside the clearly too-small number, these aren’t useful for me. I don’t take many cars (and I often use Lyft); I don’t order much food (DoorDash is a bad company); I don’t use Midjourney (and if I did, I wouldn’t use a voice interface); and I don’t use Spotify (Winamp and Plex, if you can believe it). Obviously your mileage might vary, but four isn’t a lot.

As to whether it could just be an app, and for people hung up on the idea that it runs on Android or uses some established APIs — maybe you missed the whole pitch, which is that we already have way too many apps and the point is to offload a lot of common tasks and services to a simpler, less distraction-inducing device.

Clearly I’m not the target audience for this thing. But I’m still the guy holding one and writing for a big tech publication, so let’s take this seriously.

The simple truth is I like the idea of the rabbit r1, and I’m OK with waiting until that idea has some time to mature. Rabbit is trying to build version 1.0 (though it’s more like 0.1 at this point) of the all-purpose AI assistant that Google, Apple, and Amazon have been faking for the last decade. Siri, Google Assistant, Alexa… they’re all just natural language command lines for a collection of APIs. None of them really know what to do, so they’re just backing one of the fast horses and hoping to catch up at some point. Rabbit has said that their intention was to move fast and ship something while the 900-pound gorillas of the industry are flailing.

The problem comes in separating a company’s ambition from the product. Certainly rabbit’s device is nowhere near the state that CEO Jesse Lyu showed off in various demos and videos. We have perfectly good explanations for that, but it doesn’t change the fact that the r1 is shipping in a totally barebones state.

I can’t in good conscience advise anyone to buy one now. I mean, for me, it does almost nothing. But that hasn’t stopped 100,000 people from buying one, and I don’t think they’ve been deceived in any way. Rabbit has been pretty open about the fact that it’s going to market with a minimum viable product as fast as possible (which, despite delays, has still been pretty fast),  and that it will add the features it has talked about later.

In the meantime, you have a few popular apps to use and a competent conversational AI (one you’d normally have to pay for) that can look things up for you or identify stuff in pictures. There are, like, three settings.

So it works — for a limited definition of “works.” Is that worth $200 to you? What if rabbit added video calls via WhatsApp? Will it will be worth that $200 when it adds Lyft, Tidal, audio transcription, Airbnb, navigation, and Snake? What about next year, when you can train it on whatever app you want? (Assuming the company’s vaunted Large Action Model works.) I’m not being facetious; it really is just a question of what you think is worth paying for.

$200 isn’t nothing, but when it comes to consumer electronics — especially in these days of $1,000+ iPhones — it’s not exactly a big ticket item, either. People pay $200 for RAM, for a smart measuring tape, and for nice mechanical keyboards every day. If you told me I could get an Feker 75 Aluminum for $200 right now, I’d order two and never regret it! (If you have one, email me!) Meanwhile you’ll never catch me paying full price for a MacBook Pro. Again, it’s up to each of us to decide. (Though you might wait for a security audit too, considering they’ll have authorized sessions for a lot of your accounts.)

Personally, I think it’s a fun peep at a possible future. My phone is in my bag but the r1 is in my pocket, and I can pull it out on a walk and ask “what kinds of hawks and eagles live around here?” rather than opening up the Sibley app and filtering by region. Then I can say, “add prairie falcon to the list of birds I’ve seen in Simplenote.” Then I can say “call a car to the parking lot of Golden Gardens to take me home, and use the cheap option,” and that happens. Then I ask it to record and identify the song playing by someone’s bonfire. (Just ask? In Seattle it isn’t done.) And so on.

Sure, I could do all that on my phone, but I get kind of tired of holding that thing and swapping between apps and getting notifications for stuff that isn’t actually important right now.

I like the idea of a more focused device. I like that it’s smallish and safety orange and it has a really bad camera with a complicated swivel mechanism for basically no reason (they make double-ended camera stacks for this exact reason).

Companies used to make all kinds of weird stuff. Remember Google’s weird Nexus Q music thing? Remember how wild smartphones used to be, with unique keyboards, trackballs, cool materials, and weirdo launchers? Tech is so boring now. People do everything on the same device, and everyone’s device is almost exactly the same as everyone else’s.

“What song is this?” Out comes the phone, unlock, swipe swipe tap tap.

“We should see if we can find a cabin out that way for Memorial Day weekend?” Phone, swipe swipe type type scroll scroll.

“Who were the two guys in the Postal Service again?” Phone, tap type scroll tap.

Every day, every thing, same handful of actions. It’s useful, but it’s boring. And it’s been the same for years! Phones are where laptops were in 2007, and smartphones came along to let us know there’s another way to do it. Rabbit is hoping to do the same thing to a lesser extent with the r1.

I like that the r1 exists and that it is simultaneously both amazingly futuristic and hilariously limited. Tech should be fun and weird sometimes. Efficiency and reliability are overrated. Plus, let me tell you, the homebrew and hacking community are going to go to town on this thing. I can’t wait til I’m playing Tempest on it or, honestly, scrolling down a social media app or reader. Why not? Technology is what we make of it. The r1 is leaning into that, and I for one think that’s cool.