Published on [Permalink]
Reading time: 4 minutes

Another AI gadget, another solution looking for a problem

There is now a button. A literal one. Two ex-Apple engineers who worked on the Vision Pro have built a brushed-aluminium puck that looks like an iPod Shuffle, costs US$179, ships in December, and exists for one reason: you press it, and a chatbot answers.

It is called Button, it is in Y Combinator’s Winter 2026 batch, and Wired’s Boone Ashworth got the demo.

The founders are Chris Nolet and Ryan Burgoyne. The pitch has two pillars. Privacy: it only listens when you push it, unlike the always-on Friend pendant or the Bee. Speed: replies inside a second, unlike the Humane AI Pin, which took so long to answer it died of embarrassment.

I want to be fair to it. The privacy stance is genuinely good. Nolet’s anecdote about discovering someone had been recording their entire conversation on a wearable is the kind of moment that should make anyone uneasy, and a hardware-level “off until I press it” promise is the right answer to that.

But I keep coming back to the same question.

I already have a button. It’s on my wrist.

I can raise my Apple Watch, say “Hey Siri, best sandwich shop near me,” and get an answer. I can tap the side button and skip the wake word entirely. The microphone is right there. The speaker is right there. The thing is already strapped to me. And Apple’s Google Gemini-powered Siri is getting closer - but so is Christmas!

So what does a separate US$179 puck buy me?

Three things, in theory. A faster response. A model that isn’t Siri. And a device that does not double as a notification firehose, calendar, heart-rate monitor and message machine.

The first one is real today and gone in twelve months. Apple will close the latency gap. They have to.

The second is more interesting. Button is presumably wired to a frontier model (it isn’t clear which one, and that matters), so for now it will out-think Siri on almost any query - for now. It remains to be seen what Google infused Siri is capable of.

The third is the only argument that survives contact with reality, and it is a philosophical one rather than a technical one. The pitch is that voice AI deserves a device that does nothing else. No screen to glance at. No buzz on your wrist. Press, ask, listen, done. An appliance, not a platform.

I am not convinced that is worth a second device, a second charger and a second thing to lose.

What it is actually competing with

Not the iPhone. Nolet is careful to say so. The real competitive set is narrower and stranger:

The Apple Watch with improved Siri. Already on your wrist, already paid for.

Meta Ray-Bans. Voice AI that also takes photos and plays music, in a form factor people already wear.

OpenAI’s Jony Ive device, reportedly also iPod-Shuffle-shaped, also screenless, aiming for 2027. Same form factor. Bigger budget. Direct model access.

Button has a year. Maybe eighteen months. They have to ship, they have to be loved, and they have to be acquired or pivoted before either Apple or OpenAI eats the category from above and below.

The honest verdict

I want hardware experimentation. I want people building weird single-purpose objects. The Humane Pin failing did not prove the category is dead, it proved that one execution was bad. Button’s privacy-first, press-to-talk design is the right correction.

But I cannot, sitting here with an Apple Watch on my wrist, articulate the job this device does that my watch will not do by Christmas. The founders are betting that a dedicated object beats a multi-purpose one for voice AI. That is a real bet. It is also the same bet the iPod made against the multi-purpose Pocket PC, and the iPod won that one.

The difference is that in 2001, Apple was the scrappy challenger. In 2026, Apple is the incumbent and a button on your shirt is fighting a button on your wrist that is already there.

I’ll be watching. I won’t be preordering.

Sources:

✍️ Reply by email