There's been a lot of chatter the past few months about an updated Apple TV being around the corner. With Apple's annual Fall iPod event around the corner (where they introduced the original Apple TV back in '06) and Google making a serious play for the television market themselves, the time seems right for Apple to start taking their little hobby seriously.
However, there's one thing that still bothers me about this product. With each of Apple's previous major consumer electronics devices, the iPod, the iPhone/iPod Touch, and the iPad, one of the reasons it's been a major hit is because Apple managed to solve one major element that was wrong with that category of device extraordinarily elegantly. With the iPod, the issue was navigating long lists of music, which they solved with the click wheel. With the iPhone and iPod Touch, it was having a full querty keyboard without taking away from screen real estate. And with the iPad, it was answering the question of why to buy one when it sits in between a laptop and a desktop computer.1 The Apple TV, then, must have a similar major problem to overcome.
Direct Control, Not Remote Control
The Apple TV shares a problem2 with all other devices in it's category such as the Roku box and the upcoming Google TV: the remote. The current Apple remote, while refreshing in its simplicity compared to standard remotes nowadays, still shares the same major navigational problem: the 5-way directional arrows. They're a horribly slow, clodgy, inefficient way to navigate long lists of media (there's a reason Apple came up with the click-wheel). Others around the tubes have begun hypothesizing about what Apple's solution to this might be. The other day, John Gruber linked to an interesting proposal by Dan Wineman that suggested a wireless remote that acted like a Magic Trackpad. The solution seems to make sense, especially since Apple pretty much has this available with today's Apple TV via their Remote App (App Store Link). However I have the same issue with it that Gruber does. The reason touch screens work on iOS touch-based devices is because of the direct manipulation. When you do the fling gesture to scroll a page, it feels like you're actually throwing the contents of the screen, not like you're doing a fling gesture that then starts a scrolling animation on the screen. The action and result are directly connected. With a solution like the Magic Trackpad, you remove that and create an second-rate experience that is more akin to how current desktop computer mice and trackpads work: a step backwards not forwards.
I see two real solutions that Apple could reveal for controlling your television, a conservative one and a risky one. The conservative approach tackles the directional pad problem head on with a solution that Apple already has on-hand: the click wheel. It's probably no accident that the last Apple remote design adds a non-functioning click-wheel-like design to the remote. However, this only solves navigating long lists. If we're truly to have an Internet-connected software platform hooked up to our television, we're also going to need to solve the problem with typing while at your television. Sure they could very well go with a solution like Boxee has with their remote, but this also seems inelegant (plus Apple has shown their hatred for small plastic keyboards in the past).
The riskier, more elegant solution brings us back to the question of how do we bring direct manipulation and interaction, like touch-based iOS devices have, to the television screen. It won't be via touch since the screen is usually too far away and touching something that is vertical is pretty uncomfortable. So if we can't touch it, how do we interact with it? The same way we interact with each other, our voices.
Commanding Your Television
Navigating user interfaces via voice commands is up there with flying cars as things 20 years ago we thought we'd use every day by now but that never materialized. Voice control has made inroads in certain places, like on cell phones for hands-free dialing, but it never really made it into the mainstream for controlling electronic devices. Your voice, however, seems like the most natural thing of all to use to control a television set.
Microsoft, of all places, is ahead of the curve on this stuff with their Xbox Kinect (formerly Project Natal). Their implementation is way more complex than is needed for just a set top box, but then again the main focus for them will be gaming. However, I've seen some very clever demos where it uses facial recognition to automatically log you in, you can wave your hands in the air to navigate the UI, and more importantly you can control the media playback with simple voice commands.
Apple's implementation would be more voice driven, I believe. Potential commands can be pretty simple: "pause", "play" and "movies." But it doesn't take much imagination to make them more complex (and useful), such as "subscribe to all new episodes of Lost that I haven't watched yet" or "find me a good movie to watch rated 4 stars or higher." If you think computers aren't good enough at understanding natural language search yet, then you haven't played with a little program called Siri Siri's main task is to perform exactly the kind of complex natural language searches I suggested above, and it is scary accurate. Not only is the speech-to-text almost always right on the money, but it can string together multiple API searches based on your request. Saying "recommend me a good Italian restaurant near here" will pull up Yelp restaurants of a certain rating and limit the search to within a few miles of your current location.
And low and behold, who bought Siri not that long ago? Why Apple, Inc of course.
Interesting that with the iPad Apple focused on solving not a hardware or UX problem like with its previous two products, but a sales and marketing problem. The major hardware/UX issues with the iPad, like how do you type on your lap comfortably and how do you hold it without it slipping, remain largely unsolved (the Apple Case does not count as a solution, at least not as a good one). I don't think it was the wrong problem to tackle, if I had to choose one I would have done the same, I just find it interesting. ↩
I'm politely glossing over another major issue that these devices have, for convenience of my argument. Jobs, in his wonderful interview during D8, went into detail about the struggle Apple has with a.) the content providers who don't seem to want to deal, and b.) with the cable providers who own the boxes that 98% of consumers hook up to their televisions. Each of these boxes have their own distinct UI and there's no national cable provider that would be able to standardize it, making it hard for Apple to strong arm their design expertise the way they did with AT&T. ↩