Regardless of how you feel about AI gadgets and their usefulness, there’s no doubt that products like the Humane Ai Pin and Rabbit R1 have reignited interest in new hardware.
Both standalone AI gadgets have received poor reviews, to say the least. To make matters worse, both Humane and Rabbit have been plagued by startup problems, nearly two months after the AI Pin was panned by many tech pundits last week. Bloomberg Humane is reportedly seeking potential buyers. The tech startup, co-founded by former Apple veterans Imran Chaudhry and Bethany Bongiorno, is reportedly seeking $750 million to $1 billion, but may not find a buyer or be forced to settle for less. Rabbit has faced criticism for founder Jesse Liu’s past involvement in NFT projects that have caused problems down the line.
What was meant to be a breakthrough year for AI hardware has so far been a frustrating one, with skeptics suggesting the entire product category may just be a passing fad. When AI and chatbots are on a collision course with smartphones, who could blame anyone for thinking that? Who needs standalone AI gadgets when AI is integrated into iOS and Android? But are AI gadgets destined to be swallowed up by AI-powered iPhones, Samsung Galaxy phones, and Google Pixels?
AI is never complete
Consider the critics’ main argument: AI gadgets will never be perfect. AI devices like the Ai Pin and R1 have an agile architecture that allows them to be updated remotely and more frequently than iOS or Android software updates, which usually bring a host of new features you can expect. Both Humane and Rabbit are already taking advantage of this, releasing frequent updates that improve latency, accuracy, and intelligence.
For example, I’ve been testing the latest (beta) 1.1.4 software version of CosmOS for my Ai Pin, which virtually eliminates thermal shutdowns, reduces latency by about 30%, and extends battery life by 3 hours. And a recent backend update switched the default AI model to GPT-4o just hours after OpenAI announced it, and suddenly my Ai Pin felt like it was 50% smarter.
My point is that even if an AI gadget is released in a crappy state from the get-go, changes to the AI model and rapid optimizations can turn it into a completely different (and hopefully more capable) product literally overnight. Because most of the device’s AI processing happens almost entirely in the cloud, there’s no need to wait for hardware improvements — upgraded CPUs or neural processors, more RAM, for example — to “fix” shortcomings.
Niche hardware is fine
Long before the Ai Pin shipped, Humane made a fatal mistake when Chaudhuri said “smartphones are dead” and suggested that the Ai Pin would replace the shiny slabs of glass in our pockets.
In that light, it’s no wonder most people don’t get Ai Pin’s value proposition. The clothing-worn device requires a separate phone number for calling and texting; the CosmOS software runs on Android, so it doesn’t support iMessage; and you need a separate $24/month subscription for Ai Pin to work. And you’ll have to pay for Tidal (if you don’t already have a subscription) to listen to music, as it doesn’t support Spotify, Apple Music, or other music streaming services.
As a replacement for a phone, Ai Pin isn’t good enough. Really not good enough. It doesn’t have many apps that people want or need, it can’t do email, and you can’t use it to shop. Humane says some of these features will be added in future software updates, but as it stands, Ai Pin alone won’t be enough for most people to ditch their black rectangular device.
In contrast, Rabbit’s Lyu was clear from the start that the R1 wasn’t meant to replace your phone, but more like a companion to it, so while it’s not as feature-packed as the Ai Pin, it’s also a lot cheaper ($200 vs. $700).
But why should AI gadgets either replace phones or put them out of business? I think they have potential as alternative devices for niche uses, if companies keep pumping out updates at a rapid rate. AI gadgets are great for people tired of doomscrolling. Apple and Android can’t solve the “attention economy” problem that both the Ai Pin and R1 are trying to “solve” because the business model of the phone platform is based on time spent in apps (usually time watching ads). But you’re telling me to just buy a dumb phone (aka flip phone)? Sorry, I don’t care. New Yorker Despite the claims, no one is actually flocking to dumb phones. They have an old-fashioned look, and many of them have addictive social apps you’re trying to escape (Instagram, Facebook, WhatsApp, etc.), but no AI. Maybe not having AI is the point, but having AI in a non-phone form factor is more progressive than going back to old product ideas.
AI Pin may have arrived too early, too soon, before the AI was ready to make the company’s ambient computing dreams a reality.
I’m no market researcher, so it’s possible (and very likely) that there won’t be many people who want a device that attaches to their clothes like the Ai Pin or a smartphone with fewer features like the R1. But it’s also an opportunity to focus AI gadgets on a niche audience. Instead of making a mainstream device that most people won’t embrace, make a niche device that serves a specific audience and their needs. Not every new device has to be right for everyone.
Before his retirement, veteran consumer technology journalist Walt Mossberg envisioned a future in which technology (i.e. computers) would “disappear” and become all around us, rather than a single device at the center of our lives. we In situations where humans still need to be involved throughout the day, AI unlocked by microphones, cameras and other technological means allows computing to work on our behalf. Think of it as the next level of automated Alexa skills or Google Assistant, where computers understand exactly what you need, when you need it. That’s what Humane is trying to do. Ai Pin may have come too early, before AI was ready to make the company’s ambient computing dreams a reality.
AI hardware requires a lot of resources
A sale of Humane wouldn’t necessarily mean the end of Aipin’s vision — it might just mean there will be more resources to reuse for future generations.
Photo: Raymond Wong
Humane has not commented. Bloomberg However, if the reports are true, the sale itself does not necessarily mean the end of the company or Aipin, according to the report.
Perhaps hardware startups like Humane, Rabbit, Limitless, Tap, Tone, Friend, and countless others are doomed to be cannibalized by AI integrated directly into phones. Or perhaps Humane could gain access to more resources by selling itself. Oculus is a perfect example of this. The VR company emerged with success on Kickstarter, then sold itself to Meta (then called Facebook) for $1.6 billion. While you could argue that Oculus lost its way (and lost most of its original talent), you could also see the deal as a huge win. The small VR startup had access to nearly limitless resources, could rapidly iterate on its headset, and could build its own company, Beat Games (Beat Saber), Sanzaru Games (Asgard’s Wrath and its sequels), Lady at Dawn (Lone Echo), BigBoxVR (Population: 1). If Oculus had developed it on its own, the Quest 3 might not have been released so soon.
Back to the original question: Are AI gadgets doomed? I wish I could answer it with a definitive yes or no. The reality is, we just don’t know. We’re still in the early stages of the AI ”revolution.” AI hardware is still uncharted territory, and anything’s possible until we know if AI integrated into phones actually makes a difference. But ask me again by the end of this year, and my answer might change.