Withthe sale of its creator to HP, the Humane AI Pin is dead. And with the death of the AI Pin, it’s a lot harder to imagine there’s a real case to be made for dedicated AI gadgets. Viewed one way, the writing has been on the wall since the beginning for Humane, but the company at least tried to stick around,pivoting its hardware product into a software offeringthat can act as the AI-powered operating system for your next car, speaker, or yes, smartphone, tablet, or laptop.
We’ll have to wait and see how far the team formerly known as Humane gets at HP – Palm’s hardware and software didn’t live all that long after they were acquired – but in the meantime, AI’s potential as the next big thing in computing is in a much more perilous place. For generative AI to stick, it’s going to need to rely on the gadgets people already own. The smartphone was the obvious answer before, and now it seems like the only one still left.
Is Alexa’s generative AI upgrade doomed to fail? I hope not
There’s no love lost for Amazon, but someone needs to make AI assistants better.
At some point, the AI Pin made sense
Former Apple talent, slick design, and lots of funding should have added up to something
The AI Pin was never aggressively sold as a smartphone alternative, but when you looked at its truncated list of features at launch, it certainly seemed like it could be one. Messaging, calling, and the basic web searches (via AI) that make up daily phone use were all present inHumane’s original announcement video. Switching from a screen to a projector that you can’t see very well outside seemed like a leap, but the pedigree of talent at Humane (the company’s founders worked on things like the iPhone and iPad), the attractive, compact design of the AI Pin, and the significant funding the company pulled in all seemed like things that could smooth over any rough edges.
That the AI Pin wasn’t successful doesn’t mean another company can’t be. The broad idea of a device that interacts with different generative AI assistants and agents (the industry buzzword for AI that acts for you) on your behalf lives on in theRabbit R1and Apple and Google’s plans for their smartphones. There was just little about the AI Pin at launch that worked without caveats.
That the AI Pin wasn’t successful doesn’t mean another company can’t be.
Humane pushed updates to fix things, including a major software release thatadded things like timers in 2024, but it wasn’t enough.The New York Times reportedthat the company was hoping to be purchased by HP for $1 billion not long after the AI Pin was released. HP ultimately bought the underlying “AI capabilities” of the Pin for $116 million, with no plans to support existing hardware.
Smartphones have similar pieces in a more familiar shape
There’s not much that makes current AI hardware unique, other than size
Getting people to use something new in addition to their smartphone is hard, but that was ultimately the pitch of the AI Pin, and that’s currently how the creators of the Rabbit R1 and other AI gadgets expect their hardware to be used. The criticism levied at these devices when they came out – that they could have been an app on your phone, for example – might have ignored the benefits of standalone devices that could come in new shapes, but they were clearly more right than wrong.
A smartphone already has a lot of the same components an AI Pin or Rabbit R1 has, just arranged in a more familiar shape. It also has some real advantages in the sense that it likely stores information that’s personally valuable to whoever owns it, and most people have already built their lives around it. That’s likely one of the main reasons both Apple and Google are leaning hard on the AI features in their respective phones. If it’s going to work, it would make sense it would work here.
There’s a strong desire to make that happen, too. Given the billions of dollars that have been spent on training AI models, coming up with features to use them, and actually running them, it’s likely a little concerning all of that spending hasn’t delivered the technological transformation that’s continually promised. If generative AI in enterprise products likeGoogle Docs or Sheetsisn’t moving the needle, it seems like it has to be phones and more casual, consumer-friendly uses that do. Something like asking Gemini Live for advice, the premise ofGoogle’s Superbowl ad. A softer, less fact-driven use-case that fits a paradigm that’s already familiar with phones (talking on speakerphone), but still wows the average person because you’re “talking to a computer.”
The accuracy of the things AI chatbots tell you is still a problem, however.
The pressure to sell AI to normal people is on
Only smartphones have the scale to make it happen
If generative AI doesn’t become an essential part of people’s lives on smartphones, it’s not going to happen at the scale companies want. you may view a lot of recent decisions companies have made through that lens. Like why Apple Intelligence is turned on by default when you update iOS or why theiPhone 16eexists at all. Apple,like Google, is trying to make generative AI happen.
We might not see many more AI gadgets after 2025. MaybeJony Ive and OpenAI can cook up something to the contrary, but it seems much more likely that phones are going to be the final battleground for generative AI. For a taste of what that looks like right now, you’ll want to read Pocket-lint’s coverage ofGeminiandApple Intelligence.
