Update: Stephen Ballot pointed out on twitter this post should refer to “user input method” rather than “user interface”. Great point. With this clarification I’ll leave the rest of the post as originally published. Thanks for reading.
If you analyze computing eras by circuit type you might get: vacuum tubes, transistors, ICs, microprocessors. By architecture: mainframes, minicomputers, microcomputers. By networking: ethernet, TCP/IP. That’s all great. But there’s a lot to be said for analyzing tech eras by user interface. As in the list below, taken from this post.
Analyzing computing eras by user interface:
- Punch cards – punch cards in, punch cards out
- Command line – type text in, get text out
- Graphical interface – mouse/keyboard/windows added to #2 above
- Touchscreen – Modern smartphones and tablets
Note that the internet is a complete non-event here. The web browser fit into the existing mouse/keyboard interface of the time. And that’s ok. Analysis by user interface has it’s insights and blind spots. In fact the Apple Newton and Palm Pilot, with their stylus and button interface, deserve a place above. I only left them out because they were short lived and superseded by touchscreens. But this brings up a good point. For the stylus/button interface, the Apple Newton arguably failed because it was too early. There’s a popular and common sense view that being first leads to business success. This is wrong. The hardest part of introducing a technology product is timing, timing, timing (see my timing is everything post). It’s just as easy to fail by being too early as too late. Many dot com era failures have modern counterparts that became successful using newer technology: WebVan to FreshConnect, Flooz to Bitcoin. More examples here.
From the user interface perspective being too early causes two kinds of problems:
- Pick the wrong interface. Many early smartphones had physical qwerty keyboards. These worked fine, but the ultimate interface winner turned out to be touchscreens. These were not technically possible until later.
- Pick the right interface but ship before that interface really works. As mentioned, the Apple Newton shipped before stylus/button tech was ready, as it eventually was for the Palm Pilot.
With that introduction, let’s apply the user interface/being too early framework to a couple of markets: console gaming and smartwatches.
There’s been a longstanding argument that touchscreen casual games will eventually disrupt the console gaming market. But this is not certain. From Ben Thompson’s piece arguing disruption may not happen due to the superior experience of console gaming:
I’m beginning to suspect that consoles may be a bit more resilient than many of us in tech may have first believed. And, by extension, I suspect my critique of low-end disruption may have legs: when users are buyers the user experience matters, immensely. And the user experience of a console is, and likely will remain, far ahead of any sort of touch device when it comes to many (but not all) types of games. Moreover, I now suspect that an Apple TV that supports gaming will be less disruptive than I suggested as well; as long as the controller is optional, as I suspect it would be, the immersive experience of a dedicated console will be optional as well.
Let’s see if a focus on user interface sheds any light on this question. From that point of view, games with touchscreen interfaces and games with dedicated physical controllers are simply different markets. They may displace each other in a complementary fashion on the margin, but should co-exist fine. Where console gaming could get into trouble is not through touchscreen games, but if Apple TV and Android TV get an app store and third party game controllers. While these controllers would be optional as Thompson notes, at that point it’s a fair fight. Consoles and Apple TV (for those who choose) will have the same physical interface. For example if Nintendo continues having trouble, I could see them sell a third party controller with a subscription service that works with Apple TV. That is, you buy a third party Nintendo controller for your Apple TV, and then pay $5/month to get the entire Nintendo game library. Pretty sweet.
But what about the premium gaming market? Well, for 2015 we should see Oculus Rift Virtual Reality (VR) take over premium gaming mindshare. And as an aside, it’s worth noting VR is a new computing interface, and hence will create a new market. Nonetheless traditional consoles will get squeezed between casual games on touchscreens and big budget games on VR. Even though on different interfaces. The real killer competition will be third party game controllers for Apple TV/Android TV providing a comparable gaming experience, albeit not on dedicated hardware. But recall that Apple TV runs on the mobile tech stack with an annual refresh cycle, so any lead for console technology will fade rapidly year over year. Will there be another generation of Microsoft Xbox or Sony Playstation released after the current one? I’ll stick with my long standing no, though it’s hard to be certain.
The user interface question for Apple Watch is a bit different. For Apple Watch the question is whether Apple was too early to market with the wrong interface. In Clayton Christensen terms, did Apple get the basics of the user interface correct enough so sustaining innovations (incremental improvements) are all that’s needed? Or will wrist computer interfaces need a rethink once technology improves?
Recall the key points of the Apple Watch interface:
- Crown – twist to zoom or scroll (the crown is the twist knob you’d use to wind an old style non-smart watch)
- Touchscreen – has soft and hard press
- Dedicated communicate button – direct launch to contacts
- Haptic feedback – touches back on the wrist.
- Heartbeat sensor – see my earlier post on why this is likely a biometric identity sensor, and why that’s a big deal
- Siri voice interaction – as always, I’m a big believer in the future of natural language processing (NLP), and believe Siri and Google Now will both improve greatly over the next few years.
The most likely candidate driving change to the interface will come from when the watch gets GPS and cellular connectivity, once that technology can fit into the watch without killing battery life. While that’s a big deal, I don’t see how adding a radio changes the interface itself. Apple has undoubtedly been thinking about this already. As such I think it’s not a big deal that this is missing from the first release. So I’m a tentative yes. The interface seems like it could remain stable even with a computer 100x faster than the current one. Though we won’t really know until it hits the market next year. In fact long term I think the most important interface will be voice interaction. The screen is just too small to type with or read from. I wouldn’t be surprised if a discrete headset paired with the watch were the ultimate direction for this technology.
Some related posts:
- My voice interaction post relating to the final point above: The God Particle Revisited: Augmented Audio Reality in the Age of Wearables.
- Ben Thompson’s post on Apple Watch where he argues the watch can be paired with the phone short term, but then grow out of that phase: WHY NOW FOR APPLE WATCH
I don’t think the console is resilient. Carmack said we’re on the last generation, perhaps because of streaming the whole UI, still not sure. But NUC is better than Xbox/Playstation for same reason all the Unixes died except for one — scale of manufacture. NUC iterates with PC. If as you say, HTPC may be a smartphone system, then that would be at even greater scale of manufacture than PCs (consider units of Wiis versus Xbox).
The command line will never go away – http://www.cryptonomicon.com/beginning.html