The Home Hub is Apple's most honest admission about Siri's problems yet

Apple has had the hardware for its long-rumored smart home hub ready to go for months. What it hasn't had is a version of Siri worth putting in front of it. That gap, reported consistently since early this year and reaffirmed in Bloomberg's August 9 Power On newsletter, is the most honest thing Apple has said about its AI assistant in years, even though nobody at Apple actually said it out loud. The silence is the admission.

Mark Gurman reports that a September preview of the device, widely referred to as HomePad, is plausible. It likely won't go on sale until later this year or in 2027, though. Two related products, an updated Apple TV 4K and a new HomePod mini, are expected to actually ship this fall. Both are built specifically to run the rebuilt Siri architecture arriving with iOS 27. All three devices share one dependency. None of them work as intended until the new Siri is ready, and the new Siri has been the single most delayed piece of software at Apple in recent memory.

A hub built to solve a software problem with hardware

The HomePad concept, as described across multiple reports this year, is a roughly 7-inch square display that can sit on a table or mount to a wall, running a new operating system built on tvOS foundations. It's meant to serve as the visual and conversational center of Apple's smart home push, the device you'd talk to, glance at, and use to manage everything else in the house.

That's a reasonable hardware bet on its own terms. Amazon has owned the smart display category since the Echo Show launched in 2017. Google's Nest Hub has steadily closed the gap with better proactive suggestions. Apple entering that category at all makes sense, especially on a device built specifically around a screen and a voice interface, given how far behind it's fallen.

What doesn't quite hold together is the underlying assumption. A dedicated screen doesn't fix a conversational AI problem. It just gives that problem a permanent, visible home in your kitchen. If Siri still can't reliably parse a multi-step request, or hold context across a follow-up question, putting a screen in front of that failure doesn't resolve it. It just means you'll see the failure happen in a more prominent location than you did when it was buried in an iPhone lock screen widget.

The delay history tells the real story

This device has been delayed enough times that tracking its slippage has become its own minor beat among Apple reporters. It was originally expected in spring 2025, then pushed to March 2026, then pushed again to fall 2026. At every stage, the reason given has been identical. The hardware is finished. The software isn't.

That's an unusual position for Apple to be caught in publicly, and it's repeated itself for going on two years now. This is a company that has built its entire design reputation on shipping things only once they're ready. The idea has always been that a late Apple product beats an on-time compromise. Applied to hardware, that principle has mostly served the company well. Applied here, it's produced a strange situation. Genuinely finished, tested hardware has sat waiting on a software team that can't hit its own deadlines, for an assistant that was supposed to ship as part of Apple Intelligence over a year ago.

The interaction model is more important than the delay

Set the timeline aside for a moment and look at what Apple is actually proposing to build once Siri catches up. Reports describe an interface that resembles a blend of watchOS and iOS, with circular icons and widgets, plus facial recognition that can identify who's standing in front of the display and surface that person's calendar or music library accordingly. That's a genuinely different interaction model than tapping through menus on a phone. It's closer to a shared household object that behaves differently depending on who's looking at it.

Getting that right is a harder design problem than it sounds. A phone is single-user by default. Everyone understands whose device it is and what it should show them. A kitchen display is inherently shared, glanced at by multiple people throughout the day, often from across a room rather than six inches from your face. Text needs to be readable at a distance. Interactions need to work without requiring someone to walk up and touch the screen every time. Voice has to carry more of the load than it does on any other Apple product, and that loops straight back to the Siri problem. You can't build a genuinely ambient, glanceable home hub around an assistant that still needs you to phrase things exactly right to get a useful answer.

What Apple TV and HomePod mini reveal about the strategy

The two products actually shipping this fall, the updated Apple TV 4K and HomePod mini, are reportedly getting new processors capable of running Siri's AI features. Otherwise there's little to no visible change. That's a telling detail on its own. Apple isn't redesigning its existing home hardware around the new Siri. It's quietly upgrading the chips inside products that already work fine, so that whenever the software is finally ready, the hardware doesn't become the bottleneck.

That's a sensible, almost invisible strategy for the low-risk half of the smart home lineup. It also underscores how differently Apple is treating HomePad, the one product in this wave actually built around Siri rather than merely updated to support it. Apple TV and HomePod mini can ship quietly and improve later. HomePad can't. Its entire reason for existing depends on an assistant that, as of this writing, still hasn't shipped in the form Apple originally promised.

What a September preview will tell us

If Gurman's prediction holds and Ternus previews HomePad in September without a firm ship date, pay attention to what gets demonstrated on stage versus what gets described in a press release. A live demo of natural conversation, context retention across a multi-step request, and accurate proactive suggestions would suggest Apple genuinely believes the software is close. A preview built around industrial design instead, the wall mount, the speaker base, the display quality, while keeping Siri's actual capabilities mostly off-screen, would suggest the hardware team is ready and the software team still isn't. Apple has shown its hand on this exact pattern before. Watch the demo, not the spec sheet.

You'll know within about ninety seconds which version of that story you're watching.