Two AI-powered gadgets have been making waves this spring: the Humane AI Pin and Rabbit R1, both of which promised AI automation and seamless conversations with an always-present, always-helpful AI assistant.
they failed.prominent technology critic Marques Brownlee The Humane AI Pin was called “the worst product we’ve ever reviewed,” while the Rabbit R1 received a slightly milder verdict of “hardly reviewable.”
“I don’t think it will take more than six months to see the actual results.” [AI] The same goes for applications running on PCs and mobile devices. ” —Dwith Chenna, AMD
Dr. John Pagonis, a lead UX researcher at Zanshin Labs, observed that new consumer devices must prove they are better than existing ones, a challenge both AI gadgets fail at. . “What is the problem [these devices] serve? What needs do they cover? It’s not clear. ”
So that’s a wrap, right?
Not completely. Humane and Rabbit failed, but the solution to the problem that held these new entrants back is just around the corner and could change consumer technology forever.
ChatGPT, are you there? Hello? Hello…?
Today’s best AI large-scale language models (LLMs) face a common enemy. That’s a delay. You expect it to respond when you tap or speak, but the best LLMs are located in data centers, so there can be some lag. This is the crux of Humane and Rabbit’s troubles. Reviewers complained that the gadget was slow to respond and was useless when internet access was spotty or unavailable.
But there is a solution. It is to equip the device with LLM. Regarding this possibility, IEEE spectrum A lot has happened since December. meta llama 3, microsoft phi 3and Apple’s OpenLEM— all announced in April 2024 — brought significant improvements to the quality of small AI models. Chipmakers like Apple, Intel, AMD, and Qualcomm are also working on this problem, improving the performance of AI coprocessors in laptops, tablets, and smartphones.
“All these apps, all the different notifications. It’s too much. It’s exhausting. There’s a lot of research that says this shouldn’t be the way we interact with technology.” —Patricia Reiners, UX Designer
Dwith Chenna, an AMD engineer specializing in AI inference, said these improvements will enable LLM without the cloud. The Humane AI Pin and Rabbit R1 were released too early to take advantage of these advances.
Apple’s M4, which debuted in the new iPad Pro, has “Apple’s fastest Neural Engine ever,” the company says.apple
“There’s a lot of focus on trying to narrow it down.” [large language] so that the model can run on devices such as PCs and mobile phones,” says Chenna. “I don’t think it will take more than six months to see the actual results.” [AI] The same goes for applications running on PCs and mobile devices. ”
Bringing LLM to consumer technology also addresses another important issue: privacy.
Rabbit’s R1 uses Large Action Model (LAM) to automate apps and services. However, some reviewers expressed discomfort with this idea.. LAM requires personal information, such as a login and password, in order to work on your behalf.rabbit’s promise To handle information safelyHowever, the AI models used are hosted in the cloud, so data is necessarily sent off-device.
“I think one of the main concerns is privacy and security. Not everyone is comfortable sharing their information in the cloud,” Chenna says. Pagonis agreed, noting that major technology companies are already working to address this issue. “At Google I/O, he mentioned running Gemini Nano on the device to improve privacy, and that’s a strategy I’m sure Apple will follow.”
A sour AI gadget whose design stumbles
A smaller, faster LLM running on the device could solve the latency issue, but it’s not an instant redemption for the Humane and Rabbit gadgets: both made significant design mistakes that hindered the ease of use they promised.
“I think this technology is really fascinating and… super innovative,” he said. Patricia Reinersa freelance UX designer, The Future of UX Podcast“But it has to work. It shouldn’t have any basic usability issues.”
Reiners explained that Humane and Rabbit’s predicament can be avoided by taking a step back from our AI ambitions and thinking about how people use technology in the real world. The Humane AI Pin can overheat with heavy use and relies on a projector to display information. This becomes a problem when using the pin outdoors. On the other hand, Rabbit disables the R1’s touchscreen in some menus but not in others, confusing users.
Humane AI Pin uses a projector to display information. It is also said that the projected image is difficult to see outdoors.humanitarian
“I think this is very important for anyone reading your article,” Reiners said. “Test your device or product from the first idea early on. Start with a prototype.”
Mr. Reiners and Mr. Pagonis had different opinions about the fate of Humane and Rabbit. Reiners doubts Humane’s design issues are solvable, but said the Rabbit R1’s “middling” functionality could be resolved with an update. Pagonis was more skeptical.[Humane and Rabbit] “In my view, companies are failing in the fundamental effort of product discovery: understanding the usefulness of a product and making it easy,” he said.
Is it possible for AI to kill apps?
But they agreed on one thing: The failure of early AI gadgets left the door wide open for Apple, Google and Microsoft.
“Apple and Google, they’re not sleeping,” Reiners said. “That’s why they were so quick to ship the Rabbit R1 and Humane AI Pin. They know the big companies are working on it.”
Pagonis went a step further, predicting a sweep of big tech companies: “Who will win? The companies that own your data, like Google; the companies that, like Apple, it dictates the user experience; and the companies that, like Apple and Google and, of course, Microsoft, have a relationship with you.”
That may sound disappointing. This suggests that future AI-enabled devices will maintain the status quo and look, feel, and function exactly like the smartphones we’re used to.
But Reiners doesn’t think this is the end of the story. AI may not be able to reinvent the look and feel of tomorrow’s consumer technology, but it could help reinvent the software we use on our computers, tablets, and smartphones.
“When you open your phone, there’s a lot going on. All these apps, all different notifications. It’s just too much. It’s very tiring,” Reiners said. “There’s a lot of research that says this shouldn’t be the way we interact with technology.”
Reiners believes companies like Apple and Google will try to make good on the promise of Humane and Rabbit with simplified AI operating systems that predict what users need and automate common tasks. She points out that smartphones would be easier to use if they presented users with fewer choices after unlocking the device. Mobile phones could also replace apps with automation controlled by on-device AI agents.
“Users don’t need an app,” Reiners said. “Users have goals, they have things they want to do, and they want to accomplish them. So as designers and people who work in the technology industry, we need to rethink the way we interact with users.”
From an article on your site
Related articles on the web