After blogging about the rabbit r1 (the business insists on lowercase spelling) since CES in January, I got it. I noticed it within 30 seconds of turning it on, despite receiving it a few months early. Is that okay? This odd, affordable AI device is an experiment. I think we should support that, not dodge it.
The r1 has clear issues: it lacks app connectors and “could just be an app.”
The first issue is now present. You can only link Uber, DoorDash, Spotify, and Midjourney. Besides the too-small quantity, they are useless. I seldom use vehicles (and use Lyft), order meals (DoorDash is a poor business), utilize Midjourney (and wouldn’t use a speech interface), or use Spotify. While your experience may differ, four is not a significant number.
People who are concerned about whether it could just be an app and whether it runs on Android or uses established APIs may have missed the point that we already have too many apps, and the goal is to offload a lot of common tasks and services to a simpler, less distracting device.
This isn’t for me. I have one, and I write for a prominent tech newspaper, so take this seriously.
The reality is that I like the rabbit R1 concept and am willing to wait for it to develop. Rabbit is building a version 1.0 (mostly 0.1) of the all-purpose AI assistant Google, Apple, and Amazon have been faking for a decade. Siri, Google Assistant, and Alexa are all natural language command lines for APIs. They don’t know what to do, so they’re supporting a fast horse and trying to catch up. As the industry’s 900-pound gorillas flailed, Rabbit claimed they wanted to ship quickly.
The challenge is to separate a company’s vision from its product. Rabbit’s equipment is far behind CEO Jesse Lyu’s demonstrations and videos. We have valid reasons for it, but the r1 delivers the bare bones.
I can’t recommend buying one today. It accomplishes virtually nothing for me. Nevertheless, 100,000 people purchased one, and I don’t think they fell for the deception. Rabbit has made it clear that it will launch a minimum viable product as soon as possible (despite delays, which have been rapid) and add features later.
There are a few popular applications and a decent conversational AI (that you’d typically have to pay for) that can look things up or recognize photographs. There are about three settings.
It works—for a narrow sense of “works.” Is it worth $200? What if Rabbit introduced WhatsApp video calls? Will Lyft, Tidal, audio transcription, Airbnb, GPS, and Snake make it worth $200? How about training on any app next year? (If the company’s acclaimed Large Action Model proves effective.) It totally depends on what you believe is worth paying for.
While $200 isn’t nothing, it’s hardly a large-cost item in consumer electronics, particularly with $1,000+ iPhones. Every day, people spend $200 for RAM, a clever measuring tape, and excellent mechanical keyboards. If I could get two Feker 75 aluminums for $200 right now, I would never regret it! If you have one, send it to me! I’ll never buy a MacBook Pro at full price. Again, it is our choice. You might also wait for a security audit, as they’ll have approved sessions for several of your accounts.
Personally, I like seeing a potential future. I can ask, “What kinds of hawks and eagles live around here?” During a stroll, I can’t open the Sibley app and filter by area because my phone is in my bag and the R1 is in my pocket. I may then say, “Add the prairie falcon to the list of birds I’ve seen in Simplenote.” I could then say, “Call a car to the Golden Gardens parking lot to take me home and use the cheap option,” and that’s exactly what happens. I ask them to record and identify the bonfire song. Just ask? Not done in Seattle. And so forth.
I could do all that on my phone, but I grow weary of carrying it, switching applications, and receiving alerts for unimportant things.
A concentrated device appeals to me. Its compact size and safety orange color are nice; however, its lousy camera has a difficult swivel mechanism for no reason (they create double-ended camera stacks for this purpose).
Companies used to manufacture odd things. Remember Google’s odd Nexus Q music? Remember the crazy keyboards, trackballs, materials, and launchers of old smartphones? Tech is now dull. Everyone uses the same gadget for everything, and it’s practically identical.
What song is this? Pull out your phone, unlock it, swipe, touch, and tap.
“We should try to find a cabin out that way for Memorial Day weekend.” Phone, swipe-type scrolling.
“Who were the two postal service guys again?” Phone, type Scroll Tap.
Every day, we engage in the same few activities. It’s helpful, but dull. Still the same after years! Phones are where computers were in 2007, and smartphones show us another way. Rabbit wants to accomplish the same thing with r1 less.
I admire the R1’s fantastic, futuristic, and comically limited existence. Tech should occasionally be odd and enjoyable. Dependability and efficiency are overestimated. Additionally, the homebrew and hacking communities will go crazy over this. I can’t wait to play Tempest or peruse through social media apps or readers. Why not? Technology is our creation. The R1 is bending in a way that I like.
