[[{“value”:”
Artificial intelligence is the hip new craze sweeping the globe, with corporations everywhere attempting to shove machine learning tech into every nook and cranny they can find.
Unfortunately, such an overenthusiastic approach can sometimes result in tech being applied regardless of whether it’s actually a good fit. With everyone eager to boast that their shiny new product uses AI, it’s become a classic case of companies being so preoccupied with whether or not they can use the tech, that they don’t stop to wonder if they should.
As such, not every new AI product will succeed, some of those failures having already played out in a very public manner. Here are three AI products that crashed and burned in 2024.
Rabbit r1 AI voice assistant
Credit: Joe Maldonado / Mashable
The Rabbit r1 debuted to a frenzy of interest in January, selling 20,000 units within just two days. This $200 AI voice assistant seemed to promise the world in a handheld orange box, using a Large Action Model to navigate apps on your phone in order to complete tasks given to it via voice commands. The Rabbit r1 can also take photos, describe what it’s seeing on its camera, recommend recipes based on a photo of ingredients, and summarise written text.
According to Rabbit CEO Jesse Lyu, the idea was apparently to offer a more focused, less intrusive experience than a phone. Rather than open your Uber app to order a ride, you’d just tell Rabbit r1 to do it for you.
To some, the idea of what is essentially a less functional phone that has no apps, can’t be used to make calls or browse the web, and requires another phone to get anything done doesn’t sound particularly appealing. However, many were enticed by the Rabbit r1’s claim that it could utilise AI to offer a seamless, easy way to interact with their tech — one that doesn’t require scrolling through apps..
Credit: Mashable
Unfortunately, excitement soon gave way to disappointment when it became clear the Rabbit r1 was buggy, underwhelming, and undercooked, offering more dreams and potential than solid, usable product. The Rabbit r1’s early momentum quickly stalled, its rapid descent into irrelevance further hastened by the discovery of a massive security flaw that could expose all responses every device had ever generated.
Even if the Rabbit r1 somehow manages a magical girl transformation into the functional, feature-packed gadget it aspires to be, its moment in the sun has now long passed. Overcoming its poor reputation seems like an insurmountable task at this point.
Meta’s AI Personas
Credit: Meta
Meta‘s AI Personas were basically chatbots that used celebrity likenesses. That’s it. That’s the entire product. CEO Mark Zuckerberg unveiled AI Personas last September, proudly launching the beta for the celebrity lookalike chatbots. By August the chatbots were dead, barely surviving just one year.
AI Personas were not AI versions of celebrities, though some of the chatbots’ personalities did have faint connections to the person whose face it was wearing. For example, footballer Tom Brady’s likeness was used for Bru, a “wisecracking sports debater who pulls no punches.” But by and large, AI Personas were entirely different characters who just so happened to look like public figures such as MrBeast, Kendall Jenner, and Naomi Osaka.
Though Meta touted AI Personas as fun interactions with unique personalities, the result was just awkward, creepy, and unsettling. The tech giant reportedly paid up to $5 million for the right to use celebrities’ likenesses, but it seemed you could get a close to the same effect just looking at a photo of your favourite actor while using ChatGPT.
Considering this, it’s unsurprising that Meta pulled the plug on AI Personas after just one year. The tech giant didn’t outright give a reason for ending the project, merely telling Mashable at the time that it “took a lot of learnings from building [AI Personas] and Meta AI to understand how people can use AIs to connect and create in unique ways.” But it seems safe to say that AI Personas failed to gain the traction Meta had hoped for.
Humane’s Ai pin
Credit: Humane
Last November Humane opened preorders for its Ai pin, a $699 wearable AI gadget that responds to voice commands and operates without a smartphone. Like the Rabbit r1, Humane aimed to change the way people operate and interact with their tech, and “integrate AI into the fabric of daily life.”
To that end, Humane’s Ai pin went all in on futuristic but uncommon design choices. Rather than a screen, the ostensible “smartphone killer” has a tiny projector for your palm, and though it has a touchpad, the vast majority of operation is performed via voice commands. Designed to attach to your clothes like a brooch, Humane’s Ai pin can take photos and short videos, answer questions, and send messages — basically everything a phone can already do, just without a screen.
Yet despite Humane’s lofty sci-fi ambitions, issues with the Ai pin began to appear before the devices were even released. One of Humane’s own promotional videos showed the Ai pin answering two questions incorrectly, an embarrassing mistake which prompted the company to quickly update the clip.
Credit: Angel Garcia / Bloomberg via Getty Images
Then the Ai pin began shipping in April, and the reviews started coming in. A consensus soon emerged that while Humane’s Ai pin had some cool ideas in theory, in practice it’s slow and frustrating to use. Music playback was broken (and only supported Tidal anyway), the palm projector was terrible and unreadable in bright light, and the device heated up to an uncomfortable level, which is not ideal on any gadget much less a wearable. The Humane Ai pin also has a mandatory $24 monthly subscription, turning into a useless lump of aluminium and Gorilla Glass without it.
Within mere weeks, Humane was reportedly searching for a buyer, hoping to sell for between $750 million and $1 billion. It’s an ambitious goal considering the scathing reviews of its only product. Humane’s troubles didn’t end with the Ai pin’s launch either. The company subsequently sent an email to customers telling them not to use the Ai pin’s charging case, as a “quality issue” with the battery made it a potential “fire safety risk.”
By August, Humane’s Ai pins were reportedly being returned at a pace that outmatched its sales, with only around 8,000 still left in the wild. Humane slashed the Ai pin’s price to $499 last month, but at this point, grabbing one for even a tenth of that still doesn’t seem worth it.
“}]] Mashable Read More
The Rabbit r1 AI voice assistant, Meta’s AI Personas, and Humane’s Ai pin are three AI products which failed in 2024.
