Humane built a $699 device to replace the smartphone before it proved anyone wanted to put their phone down.

The pitch — name stripped

What actually happened

Humane was one of the most hyped hardware startups in years. Founded by two senior ex-Apple designers, Imran Chaudhri and Bethany Bongiorno, it raised more than $230 million from names like Sam Altman, Marc Benioff, Microsoft, and Qualcomm Ventures, reached a reported valuation near a billion dollars, and spent years in stealth teasing a device that would free people from their phones. The pitch was a screenless, AI-powered pin you wear on your chest, with a tiny laser that projects text onto your palm.

The AI Pin shipped in April 2024 at $699 plus a $24-a-month subscription, and the reviews were brutal. The most-watched tech reviewer on YouTube called it 'the worst product I've ever reviewed.' Buyers and critics hit the same wall: it overheated, the battery drained fast, responses were slow, it confidently got things wrong, and the palm laser was nearly impossible to read in daylight. It asked you to do, worse and slower, the things the phone in your pocket already did instantly.

The market answered fast. By the summer of 2024, Humane's returns were outpacing its sales — more devices were coming back than going out. The company told owners to stop using the charging case over a battery fire risk, and dropped the price by $200. None of it moved the underlying problem: there was no painful, widely felt reason to give up the phone screen, so there was no reason to pay a premium and a subscription to do so.

In February 2025, less than a year after launch, HP acquired most of Humane's assets for $116 million — well below the $230 million-plus it had raised and a fraction of its peak valuation. The AI Pins themselves were bricked: on February 28, 2025, Humane shut off its servers and the devices people had paid $699 for simply stopped working. The hardware that was meant to outlast the phone outlived its own company by days.

Ambient, always-on AI is a real direction, and Humane was early to it. But being early to a future is not the same as proving people want to buy the present version of it. The company built a beautiful answer to a question almost nobody was asking loudly enough to re-buy their entire computing life over.

The lesson

Kasspian's read on the stripped pitch went to the premise, not the engineering: the riskiest assumption was never 'can we build ambient AI hardware' — clearly they could — it was 'does anyone actually want to replace their phone badly enough to pay for a worse one.' Hardware is dangerous here because the building feels like progress. Years of beautiful engineering can hide the fact that the core want was never tested.

'Replace the phone' is one of the hardest premises in tech, because the incumbent is free, you already own it, you love it, and it improves every year. To beat that, a new device has to be dramatically better at something people care about a lot — not marginally different and calmer. Humane was worse at most things and quieter at one, and quiet was not a $700 problem for enough people.

The cheap test was demand, not the device. Before the custom OS, the laser projector, and the years of secrecy, the question to answer was simple: will real people put down real money to give up their phone screen? Prove the want before you build the wonder. A flawed premise does not get less flawed with more capital — it just gets more expensive to disprove.

Common questions