Tony Fadell Says 3 First-Wave AI Gadgets Failed as Trust and On-Device Design Become Essential
Updated
Updated · TechCrunch · Oct 7
Tony Fadell Says 3 First-Wave AI Gadgets Failed as Trust and On-Device Design Become Essential
2 articles · Updated · TechCrunch · Oct 7
Summary
At MIT Future Fest, Tony Fadell pointed to three discontinued devices—the Rabbit R1, Humane Ai Pin and Limitless pendant—as proof that first-wave AI hardware solved no real consumer pain point.
Less than 0.01% of people have ever used a human assistant, he said, so most consumers neither understand the value of an AI agent nor trust it with sensitive tasks like banking or scheduling.
Trust is the bigger hurdle because AI assistants need deep access to personal data; Fadell cited Meta's Muse launch, where a researcher found a serious flaw and employees reportedly rushed to fix security issues before release.
On-device computing will be required for a successful AI agent, Fadell argued, because keeping data local improves privacy and avoids dependence on data centers while still using increasingly powerful battery-run hardware.
Apple is best positioned on hardware, chips and privacy goodwill, he said, though it lacks a top in-house AI model; rivals like Meta and OpenAI are chasing gadgets partly because they lack Apple's installed base of devices and sensors.