No matter how you feel about AI gadgets and their usefulness, there’s no doubt that products like the Humane Ai Pin and Rabbit R1 have revitalized interest in new hardware.
Reviews for both standalone AI gadgets have not been great. Making matters worse, both Humane and Rabbit have been mired in their own concerning startup issues. Almost two months after the Ai Pin was panned by many tech reviewers, last week Bloomberg reported Humane is looking for a potential buyer. The tech startup, co-founded by ex-Apple veterans, Imran Chaudhri and Bethany Bongiorno, is reportedly seeking $750 million to $1 billion, though it’s possible it might not find a buyer or have to settle for less. Rabbit has faced backlash for founder Jesse Lyu’s past involvement with an NFT project that’s come back to haunt it.
What was supposed to be a breakthrough year for AI hardware has so far turned out to be frustrating, with skeptics suggesting the whole product category might be a fad. And who could blame anyone for thinking that when AI and chatbots are on a collision course with smartphones? Who needs a standalone AI gadget when AI will be integrated right into iOS and Android? But are AI gadgets doomed, destined to inevitably be swallowed up by the AI-infused iPhones, Samsung Galaxy phones, and Google Pixels?
AI Will Never Be Finished
AI gadgets will never be finished. AI devices like the Ai Pin and R1 have a simple architecture that can be updated remotely and more frequently than the typical annual iOS or Android software update that usually brings a slate of new features to look forward to. Both Humane and Rabbit are already taking advantage of this, shipping frequent updates that improve latency, accuracy, and intelligence.
For example, I have been testing out the latest (in beta) 1.1.4 software version of CosmOS for the Ai Pin, and it virtually eliminated thermal shutdowns, cut down latency by roughly 30 percent, and added three hours of battery life. Another recent backend update switched the default AI model to GPT-4o, hours after OpenAI revealed it, and suddenly, the Ai Pin felt 50 percent smarter.
My point is: Even if an AI gadget launches sloppily out of the gate, if the AI model is changed and optimizations are made swiftly, it could be a completely different (and hopefully more competent) product literally overnight. Because most of a device’s AI processing is done almost entirely in the cloud, you don’t need to wait for improved hardware — for example, with an upgraded CPU or neural processor or more RAM — to “fix” what was underbaked.
Niche Hardware Is Okay
Long before the Ai Pin shipped, Humane made a fatal mistake: Chaudhri said “the smartphone is dead,” implying that the Ai Pin would replace the glowing glass slabs in our pockets.
With that lens, it’s not surprising that most people can’t see the Ai Pin’s value proposition. The clothing-worn device requires a separate phone number for calling and texting; it doesn’t work with iMessage because its CosmOS software runs on Android. And, a separate $24/month subscription is required for the Ai Pin to even function. Oh, and you have to pay for Tidal (if you aren’t already a subscriber) to listen to music because it doesn’t support Spotify, Apple Music, or any other music streaming service.
As a phone replacement, the Ai Pin falls short — way short. It doesn’t have many apps that people want or need; it doesn’t have email; you can’t use it to buy things. Humane says some of these features are coming in future software updates, but as it currently stands, the Ai Pin simply isn’t enough for most people to ditch their black rectangles.
In contrast, Rabbit’s Lyu was vocal from the start that the R1 isn’t out to replace phones; the device is more of a companion. That makes it not as feature-packed as the Ai Pin, but also a lot cheaper ($200 versus $700).
But why does an AI gadget need to replace the phone or go bust? I could see them having potential as an alternative device for niche use cases if the companies keep shipping updates at a rapid rate. AI gadgets are great for people who are tired of doomscrolling. Apple or Android won’t solve the “attention economy” problem that both the Ai Pin and R1 are attempting to “fix” because phone platform business models are built on time spent in apps (usually looking at ads). But just get a dumb phone (aka flip phone), you say? I’m sorry, I don’t care what The New Yorker claims, nobody is actually flocking to dumb phones. Dumb phones look ancient; many have the addictive social apps you’re trying to escape from (Instagram, Facebook, WhatsApp, etc.); they don’t have AI. Maybe not having AI is the point, but having AI, but in a different form factor than a phone, is arguably more progressive than going back to an old product idea.
The Ai Pin might have come too soon before the AI is ready to deliver on the company’s ambient computing dreams.
I’m not a market researcher, so it’s possible (and very likely) that the number of people who are asking for a clothing-worn device like the Ai Pin or a watered-down phone like the R1 isn’t very large. But that’s also an opportunity to focus an AI gadget on niche users. Instead of making a mainstream device that most people will not bend themselves to, just make a niche device that caters to specific users and their needs. Not every new device needs to be for everyone.
Before retiring, veteran consumer technology journalist Walt Mossberg envisioned a future where technology (read: computers) would “disappear” and become “ambient.” Instead of a singular device at the center of our lives that we would have to engage with all day, AI, unlocked by microphones, cameras, and other technology means, would allow computing to work on our behalf. Think of it as the next level of an automated Alexa skill or Google Assistant, where computers understand what you need exactly when you need it. That’s what Humane is trying to do. The Ai Pin might have come too soon before the AI is ready to deliver on the company’s ambient computing dreams.
AI Hardware Needs Lots of Resources
Humane hasn’t commented on Bloomberg’s report. But if the report is true, selling itself doesn’t necessarily spell the end of the company or the Ai Pin.
Perhaps Humane, Rabbit, Limitless, Tap, Tone, Friend, and countless other hardware startups are doomed because AI integrated directly into phones will eat their lunch. Or, perhaps Humane selling itself could mean access to more resources. Oculus is a perfect example of this. The VR company landed on the scene as a successful Kickstarter… and then it sold itself to Meta (then called Facebook) for $1.6 billion. You could argue that Oculus has lost its way (and most of its original talent), but you could also look at the deal as a big win. The small VR startup got access to near-infinite resources, which enabled it to rapidly iterate on headsets and secure crucial software developers such as Beat Games (Beat Saber), Sanzaru Games (Asgard’s Wrath and its sequel), Ready At Dawn (Lone Echo), and BigBoxVR (Population: One). We may not have gotten the Quest 3 as soon as we did had Oculus gone it alone.
Back to the original question: Are AI gadgets doomed? I wish I could give you a straight answer of yes or no. The reality is, we just don’t know. We’re still in the early innings of this whole AI “revolution.” AI hardware is still the Wild West and until we see if AI integrated into phones will actually move the needle, anything goes. Ask me again by the end of this year, though, and my answer might be different.
