Cause of death: no reason to exist
Every few years, tech convinces itself the smartphone is finished. And every few years, the smartphone attends the funeral of whatever was built to replace it. Today''s autopsy: the dedicated AI gadget wave of 2024–2025 — the Humane AI Pin, the Rabbit R1, and the funded belief that talking to a small orange box beats tapping a screen.
The body of evidence
The numbers read like a cautionary tale written in invoices. Humane raised about $230 million to build a laser-projecting lapel pin. It shipped fewer than 10,000. HP bought the remains for $116 million — half what investors put in — and the Pin wasn''t just discontinued, it was bricked. Since February 28, 2025, every unit on earth has been warm jewelry.
The R1 is the volume version of the same story. Rabbit''s $199 orange box moved 100,000 units on launch hype. Then the returns started, in bulk, as buyers learned the "large action model" mostly couldn''t do what the demo showed. The company now reportedly struggles to make payroll.
Users wrote the epitaph: in a 2026 survey ranking, these sit among the worst-reviewed devices ever made.
The misdiagnosis that killed them
The failure pattern barely varies. Viral demo. Breathless coverage. Long waitlist. Both teams read all that attention as validation — but it was curiosity, which is a different currency entirely. When the hardware finally arrived, it was a slower, dumber version of the phone already in your pocket. No screen worth using. No apps. Worse battery.
That''s the real finding here. Excitement about AI never became a reason for AI to live in separate hardware. People wanted the assistant; the gadget was a tax on it. Novelty products die the moment a market matures enough to ask one question: what is this for?
The sequel already in production
And here''s the twist that keeps this corpse interesting — the industry didn''t take the hint. OpenAI is reportedly building its first device, a screenless speaker that can move. It also just shipped a $230 dedicated keyboard for Codex, mid-lawsuit, no less.
Maybe the biggest AI company alive studied the Pin and learned what not to build. Or maybe wave two is about to prove the problem was never execution — it was the premise. Either way, the lesson from the first wave holds: watch the return rate, not the waitlist.
