The Rabbit R1 has been collecting dust on my shelf since I grew disappointed with the device not meeting my initial expectations. It's been two months since I used it. Jesse Lyu, the CEO behind Rabbit R1, overpromised and underdelivered with the Rabbit R1.
Admittedly, some features like Vision, which shows off how the Rabbit R1 can identify objects by “looking" at them before explaining what it "sees," aren’t half bad. I also loved that it could identify foreign languages, like on a menu for example, and spit it back out in English. (These features are already available on my phone via apps like ChatGPT Plus, however.)
Plus, some of the marquee features, like ordering Uber and DoorDash via Rabbit R1, or invoking the AI to launch a Spotify song, did not work well.
And not to mention that Rabbit R1 has been suffering from a critical security flaw, according to the The Verge, that further threatened its viability in the market.
To be fair, this isn't just a Rabbit R1 issue. This whole "we'll promise stuff now, but deliver after you give us money" tactic has been running rampant across the tech world. But the likes of the Rabbit R1 and the Humane Ai Pin, which is reportedly suffering from more returns than it can can sell, shows how this strategy can backfire.
But Rabbit hasn't given up yet. In a press email to Mashable, Rabbit said that it has issued extensive updates to the R1 device, goading me to give it a try. The team warned that these new features, dubbed "beta rabbit mode," are still rough around the edges, but that I should experiment with them. So I did.
According to Rabbit, beta mode enhances R1's capabilities with more thoughtful, comprehensive responses to complex questions that require deep reasoning and multi-step research. It can also be used with Vision. In other words, based on what Rabbit R1 is looking at via the camera, it should be able to deliver advanced responses to questions.
Beta rabbit supposedly makes Rabbit R1 more conversational and intelligent.
To access beta rabbit mode, all I needed to do was say, "beta rabbit." Just as I was about to give this beta rabbit mode a bash, an update suddenly occurred, preventing me from using the device for about three minutes.
And it must be connected to a USB-C charging cable to successfully complete the update. It took about 30 long minutes from start to finish due to Wi-Fi connectivity issues (that weren't on my end).
Rabbit gave me the following commands for testing beta rabbit mode:
"Beta rabbit, can you suggest three books similar to The Power of Now? Include page length, year of release, and ratings. And save that as a note titled 'Reading list.' Also, include pictures of the authors." To show contextual capability, Rabbit suggested following up with, "Beta Rabbit, can you also get me summaries for those three books?"
"Beta Rabbit, can you give me a two-day itinerary for San Francisco? I want to go to one historical site, a museum, eat Japanese food, and try a cocktail bar. Can you also give me photos of each location?
Beta Rabbit, can you find me the best deals for the Dyson V8 vacuum? I want to get the cheapest price and the purchase links.
Point at the nutrition chart of a Japanese snack, or any snack in a foreign language, and ask, "Beta Rabbit, what are the nutritional facts for this snack? Is it healthy?"
I tweaked these a bit so that it suited my personal life. For example, for number one, I asked for film recommendations in lieu of book suggestions. (Plus, I don't want to follow a Rabbit-curated list; I want to challenge it.)
For number two, I swapped San Francisco for Greece because I'm headed there in two months. For three, I'm in the market for the Instant Vortex Plus Air Fryer, so I plugged that in. And lastly, I don't have a Japanese snack, but I do have a cup of Ramen Noodles, so i'll ask "beta rabbit" how healthy it is.
Initially, when I tested beta rabbit mode on Aug. 8, it worked well.
I asked, "Beta Rabbit, can you suggest three films similar to The Wolf of Wall Street? Include movie length, year of release, and ratings. Also, include pictures of the directors." It suggested The Big Short, American Hustle, and The Social Network, alongside the requested information. Since my two-month break from the Rabbit R1, one thing that caught my eye is that the device finally supports pictures in its response, delivering photos of each director as requested.
I also followed Rabbit's suggestion to ask for book recommendations. I said, " Beta Rabbit, can you suggest three books similar to Killers of the Flower Moon by David Grann. Include page length, year of release, and ratings. Include pictures of the authors, too." Again, it surfaced exactly what I asked for, including The Devil in the White City by Erik Larson, with pictures in tow.
However, a week later, I retested these queries and discovered that the photo-filled responses have been replaced with a rough, rudimentary, raw reply, with some backend coding exposed. I no longer saw photos.
As of Aug. 22, however, the Rabbit R1 is back in working order after a final retest. (In Rabbit's defense, it's a beta mode.)
I plan on visiting Greece later this year, so I asked the Rabbit R1 to curate a travel plan for me: "Beta Rabbit, can you give me a two-day itinerary for Athens? I want to go to the nearest beach, a historical site, eat greek food from a highly rated restaurant, and go to the best rooftop bar. Also, can you give me photos of each location?"'
To my surprise, the Rabbit R1 actually put together a useful Athens-based itinerary. Now, I'll have to actually check out Rabbit R1's suggested places to determine if the AI device can double as a beneficial travel assistant. But based on what Rabbit R1 has produced, I could see myself using it for moments when I just don't have the time to do the research to put together a fleshed out itinerary.
The only downside is that there doesn't seem to be a way to retrieve a history of responses on the actual Rabbit R1 device, so you'd have to be careful not to accidentally press a button that makes the two-day itinerary disappear from the screen.
As mentioned, I'm in the market for a new air fryer (i.e., the Instant Vortex Plus Air Fryer).
As such, I asked the Rabbit R1 to surface the best deals for it. The Rabbit R1 pointed to a $131 Amazon "deal." However, I did a little research myself and discovered at it was not the cheapest online offering at the time. The best discount for the air fryer was via Williams Sonoma, which had the same fryer for about $99.
(Interestingly, two weeks later after testing, that $131 Amazon deal has dropped down to $99 as of this writing.)
I pulled out a cup of Ramen Noodles. Using the Rabbit R1's "Vision" feature, I pointed the camera at the back of the Ramen Noodles, which sports the nutrition facts label. I asked, "Beta Rabbit, what are the nutritional facts for this cup of Ramen? Is it healthy?"
When I mentioned "cup of Ramen," the Rabbit R1 gaslit me and told me that I was not holding a cup of Ramen. I re-asked the question, but this time, I simply said, "Beta Rabbit, what are the nutritional facts for this food? Is it healthy?"
In response, it told me that the Ramen Noodles are too high in salt and saturated fats, which may impact cardiovascular health. Rabbit R1 continued to say that the product "lacks essential vitamins and minerals."
Rabbit R1 concluded that Ramen Noodles are not healthy and suggested that I avoid eating them.
Surprisingly, at launch, users couldn't set timers nor alarms with the Rabbit R1. Now, you can use the cute little orange gadget to wake you up or remind you that your cookies are almost done.
Unfortunately, the Rabbit R1's chance to captivate the masses has come and gone. I believe Rabbit would have had a glimmer of hope if it were transparent about the device's half-baked feature set from the start. To be fair, there were some utilities that did operate smoothly day one. For example, the translation and Vision features worked better than expected, as well as the sound recording capabilities and the relatively fast, Perplexity-based AI assistant (though its contextual abilities are limited). On the other hand, the team promised a slew of other perks, including UberEats and DoorDash ordering, seamless integration with Spotify, and more. But these features struggled to operate properly.
Plus, people kept wondering the one question Rabbit likely hates hearing, "Why can't this just be an app? Why do I have to spend $200 on a device that already does things my daily driver phone can do?" (This investigation led to a number of explosive reports, including the revelation that the term "RabbitOS" is a bit disingenuous because it's really just an app running on a cheap Android handset.)
And in Rabbit's defense, would people really give a monkey's behind about the R1 if it rolled out to the Google Play Store as a $20-a-month app?
The Rabbit R1, despite offering the same features that are available on your phone, could have gotten away with being a success just from being a bright-orange, nostalgia-inducing device with funky, fidgety controls that taps into one's inner child. But again, the Rabbit R1 is simply too disgraced to experience a revival. However, the Rabbit R1 was such a hyped piece of tech in recent pop culture, I wouldn't be surprised if the Rabbit R2, if Lyu dared to give it a second shot, landed in the spotlight again. Its popularity would likely stem from people wondering if the R2 will be another unfinished gadget.
I suspect that consumers are at their wits end with overpromise-and-underdeliver tech. At the very least, consumers want something as close to perfection as possible (as opposed to the deliberate and exploitative launch of half-baked product releases to secure a return on investment). One can only hope that the wave of negative reviews that the Rabbit R1 and Humane Ai pin have suffered lights a fire under the tech industry's butt to quit the sloppy practices.
