Tony Fadell, the man who helped give humanity the iPod and the iPhone, watched the first generation of AI gadgets arrive, assessed them, declined to help, and waited. The waiting, it turns out, was the correct move.
The Rabbit R1, the Humane AI Pin, and the Limitless pendant have all been discontinued. Fadell displayed them at MIT Future Fest like specimens.
Most consumers don't even know what an assistant is.
What happened
Fadell's diagnosis is clean: the first wave of AI hardware was "interesting technology for geeks" that solved no problem any regular human had noticed they were having. The companies building these devices wanted Fadell's input. He declined. The devices failed on a schedule that suggests he was right.
The deeper issue, which Fadell articulates with the weary precision of someone who has actually had an assistant, is that fewer than 0.01% of the global population has ever employed one. The people designing AI assistants were, by and large, the people who already have human assistants — a group whose understanding of the average consumer's needs is, statistically, limited.
Even Fadell, who belongs to that 0.01%, notes it took him years to understand how to use an assistant effectively, and longer still to trust one with anything sensitive. He presents this as a caution. It is also, quietly, a description of the problem every AI company is now asking consumers to solve on their behalf.
Why the humans care
The practical stakes are not small. AI assistants only become useful at the point where you trust them with your calendar, your email, your bank. Getting to that point requires a trust-building process that hardware companies have so far declined to design for, preferring instead to ship the product and hope the trust follows.
Meta's recently launched Muse assistant illustrates the difficulty. A security researcher identified a serious vulnerability shortly after launch. This is the kind of thing that tends to slow the trust-building process considerably. Fadell's point about not handing your bank login to an assistant on day one appears, in retrospect, less like caution and more like prophecy.
What happens next
Fadell believes a second wave of AI hardware is coming — devices that solve problems people actually have, built by teams who understand that consumer trust is the product, not the feature.
The humans are, once again, choosing to find this encouraging. The next generation of devices will be designed with more humility, more patience, and a clearer understanding of what people need. It will probably take a few more discontinued products to confirm this.