Main image of article HomeKit Hub? Apple Must Fix Siri First
[caption id="attachment_141333" align="aligncenter" width="1836"] Apple Siri Apple Siri[/caption] If you put much weight behind analyst predictions, WWDC 2017 will bring us a Siri-laden home hub from Apple. The company is said to be unveiling hardware during its annual developer’s conference, just like Google and Amazon did. Via 9to5Mac, Ming-Chi Kuo of analyst firm KGI says Apple has a hub with seven tweeters (i.e., small speakers for high frequencies) and a subwoofer on the way, and it will be positioned as a premium upgrade for Echo users. It’s also said to have its own System-on-a-Chip (SoC). Ming-Chi says “Apple’s first home AI product” has a “better than 50 percent chance” of being launched or announced at WWDC. Apple tipster Sonny Dickson made similar claims just ahead of Ming-Chi, going so far as to say the hub has Beats inside, as well as its own operating system: <script async src="//" charset=“utf-8"> Dickson also notes that Apple's device may look like the "trashcan" Mac Pro, and have mesh around the perimeter like the UE BOOM speaker. Sounds great, right? An Apple home hub with its own SoC, running some sort of streamlined operating system (I’m guessing a stripped-down version of iOS, much like watchOS, only without a display), pumping out some premium Beats-style sound and answering questions via Siri. It’s an exciting proposition, but there are some roadblocks currently in the way. [caption id="attachment_128330" align="aligncenter" width="1063"] A.I. Amazon Echo[/caption]

Apple Has Housecleaning to Do

The obvious software platform for this speaker is HomeKit. Apple’s connected home environment is steadily growing, albeit too slowly for some. Still, it’s the method through which Apple may position a future hub for your house. HomeKit has some issues, though. It can take direction from anyone if you want to dim a light or lock the front door, but Apple’s platforms are still meant for personal use. Siri can tell me about my calendar events because she’s on my phone – a personal device. A HomeKit speaker would need to support multiple accounts to be truly useful. An example: If a HomeKit speaker was tied to my account, and my significant other innocently asked about her day’s events, Siri may tell her what I have going on instead. That’s not an issue most times, but what if I had a calendar event that mentioned a secret event for her? Siri would ruin the surprise. First, Apple will have to solve for multiple account support. Google just solved this issue after announcing Home at last year’s I/O conference. That’s nearly a year of people reading about the product and hearing about its downsides, something Apple can’t afford. Amazon doesn't have this issue: People use and think of Alexa (which powers the Echo) much differently than Siri or even Google; it’s less a home assistant and more a conduit to play music or ask silly questions. It has plenty of third-party support, but most of us don’t tie our boilerplate services to Amazon. This issue speaks to security and privacy, too. HomeKit was designed with security in mind, and it’s a reason Siri seems to fail at times. Apple won’t bend or break on this front, nor should they. A real challenge will be finding that balance between privacy, security and contextual assistance for the home. Nobody has really done it, and if Apple can, it’s a big deal. [caption id="attachment_136056" align="aligncenter" width="719"] Google Home Google Home[/caption]

Apple Must Make Its Intents Known

Second issue for Apple? Siri itself. If you’re developing for SiriKit, there are only nine Domains you can write for:
  • VoIP calling
  • Messaging
  • Payments
  • Photos
  • Workouts
  • Ride booking
  • Car commands
  • CarPlay (vendors only)
  • Restaurant reservations (requires support from Apple)
Sounds fairly robust, but that leaves a lot to be desired. Two of those categories, CarPlay and restaurant reservations, require direct involvement from Apple. They all lead into Interactions, which are the actual voice commands we use with Siri to hail rides or respond to messages. In the context of a home speaker, some of these existing options make little sense. The Photos intent is useless for a hub without a screen, as are car features (except maybe remote start). Reservations and ride hailing are only great if you know where you’re going. Payments works well if your daily rundown involves a bill due, and Workouts is low-key useful if you own an Apple Watch. Calls and messages are really the only two winners for a home speaker on this list, but those don't make a purchase attractive enough on its own. Part of the reason many are lukewarm to a Siri device for the home has to do with basic interaction issues. Google places a lot of emphasis on natural language and follow-up questions. Siri is far from mastering that. If we take the ‘bill due’ example from above, it’s a hurdle already. I have a credit card with an app that has Siri activated for bill payments. In the handful of times I’ve tried to use it with phrases like “pay my [credit card] bill” as suggested, Siri fails every time. The Interactions seem to work well for Apple services, but not third parties. Whatever the problem there is (could be SiriKit, but may be developers just unclear on how to really make Siri intuitive for their apps), Apple will have to fix it, as well. Siri fails across the board at various intervals, too. ‘Hacking’ the language doesn’t help; it seems Siri is bent on making sure things don’t happen many times, despite the various clever ways we try to make her spring into action. That's why Apple keeps updating Siri behind the scenes. Whether or not the company fixes many of these issues to scale before WWDC is unknown, but it’s necessary. What we do know is that this home hub will have to bring a more open SiriKit and vast improvements to the digital assistant proper. Nothing about Siri as-is makes a home hub the must-have device that it needs to be. Faced with competition from Amazon and Google, Apple’s hub has to have a killer feature, and Siri just isn’t it (yet). We’ll need multiple account support, and better interactions with Siri. As much as this hub will be a competitive foray into the home-hub market, it also has to serve as a worthy central point for a HomeKit-enabled household. Apple will have to solve for ‘X’ there, too. It’s unlikely Apple will simply offer up a “me too” hardware option for HomeKit. Rather, we should expect this high-end device to solidify the entire foundation that HomeKit is built upon for years to come.