Rabbit R1 Hype: A Case Study in AI Mislabeling

Irene Okpanachi
This rabbit is so undercooked, you can hear it singing AI-EIO.
Rabbit R1 Hype: A Case Study in AI Mislabeling 3
Image: Rabbit

Editorial Note: Talk Android may contain affiliate links on some articles. If you make a purchase through these links, we will earn a commission at no extra cost to you. Learn more.

Among the cool gadgets to turn up this year, the Rabbit R1 was high up among my most anticipated. Of course, I wasn't going to buy it. But knowing something that awesome existed in the same space as me was enough. Is that hypocritical? Yes. But did it save me money? Yes, yes, yes.

What was promised was a next-level AI assistant with a deep understanding of natural language. Perhaps, that was the only real problem with the device — market hype. Beyond the aesthetics, all you really have is a plastic box that you point at things to tell you what you could’ve easily found with Google Lens for free. 

Rabbit R1 is ironically a heavily manual device

Rabbit R1 Hype: A Case Study in AI Mislabeling 4
Image: Rabbit

Rabbit does have a voice control feature. But you don't wake the assistant AI with your voice, which is the first red flag. Instead, you'll hold down a side button and release it when you're done talking, kinda like a walkie-talkie… over

Then you'll shake the device to reach the menu and use a slider button to navigate the interface or do everything. There's a touch screen, believe it or not, but you can use it in Terminal mode where a keyboard pops up for typing. Let's not forget the creepy camera tucked at the side for identifying the world around you.

Now, I didn't expect to see a revolutionary AI like they promised. I mean, come on. It's a $200 plastic box. But I was even more shocked to discover it was really just a repackaged version of Google Assistant or Siri. What initially tickled everyone's imagination was the fluidity of conversations. We wanted to tell Rabbit, “order my usual from Starbucks and have it delivered to my office”. Then watch in awe as the AI whizzes through your order history and map locations to fulfill the request.

The device can't even process such advanced commands, nor can it follow up on conversations. Worse, it gets most answers wrong and supports integration with only four services at the time; Uber, DoorDash, Spotify, and Midjourney. What's more interesting is how developers at XDA were able to port the Rabbit R1 app onto a Pixel device, and run LineageOS on the R1. Android Authority also installed the Rabbit R1's launcher APK on a Pixel 6a and set it up as if it were a Rabbit R1.

Rabbit R1 launcher on Google Pixel 6a phone
Image: Android Authority

In case you're not all caught up with Android's history, LineageOS is a free and open-source operating system that's based on the Android mobile platform. So basically, the Rabbit R1 is a glorified replica of your Android smartphone, even though its creators have denied it.

Rabbit R1 is the best phone for kids & seniors

There are many promised features missing on the Rabbit R1. For one, the battery isn't optimized for a full day's use. You'll get six hours tops. What's the point of an “all-day” pocket companion if you can't use it throughout the day?

Rabbit R1 Hype: A Case Study in AI Mislabeling 5
Image Credit: Rabbit

It's almost criminal how they got away with this much unrealistic marketing. You just know that this device is no mistake. Rabbit Inc. knew its shortcomings from the beginning. It would've been much better if they'd marketed its actual features, and redirected them to a specific demographic. Maybe seniors and kids?

There are plenty of arguments online about whether it's even qualified to be called an AI assistant. Tech companies have started this annoying trend recently where they take existing smart features and slap the AI tag on them. Then we're like “ooh, aah. Take my money!” and it's become a vicious cycle of brainwashing. 

Arguments for what is AI and what is not

Rabbit R1 Hype: A Case Study in AI Mislabeling 6
Image: Rabbit

Michael I. Jordan, a leading AI and machine learning researcher, once explained that most large models do not even possess true intelligence. They're only good at low-level pattern recognition rather than high-level cognitive tasks. He stressed that the term “AI” is in fact misused, and quite many people have it all wrong, including experts. True AI, which would have human-level reasoning and thought, is still far from possible.

I agree with his point because I've used Gemini and ChatGPT repeatedly, and haven't felt like I was interacting with an AI. Both programs don't think for themselves. Instead, they take the prompts you give them and generate responses based on patterns they learned from compiled human data.

Think of it this way. Would you say Google Maps is an AI because it calculated the fastest route to your house? That's not AI. It's just a computer processing traffic data and maps with pre-programmed algorithms. In the end, most of them are exactly that. Advanced computers.

Total
0
Shares
Leave a Reply

Your email address will not be published. Required fields are marked *

Previous Post
Oppo Says It Will Bring AI Features To All Of Its Phones 7

Oppo Says It Will Bring AI Features To All Of Its Phones

Next Post
Infinix' Entire Note 40 Series Gets Racing Edition Models 8

Infinix’ Entire Note 40 Series Gets Racing Edition Models