The Apple Watch is not the first wrist-worn computing device — the Timex Datalink from 1993 could be programmed using the flickering of light from a personal-computer screen — but it seems to be the first designed to interface with the body, not just eyes and fingers. In his review of the device’s first generation, the New York Times‘s Farhad Manjoo notes that it was quite a transition to use it, but in making it, he found himself less absorbed by the device than extended by it.
Wearable computing has always had at its heart the difference between productivity and augmentation, between making specific tasks easier and giving people new abilities. It is a subtle distinction, until it isn’t. A cordless drill lets a worker mount a sheet of drywall much faster; a crane changes how buildings are built and how they can be repaired. Spell-checking fixes your spelling; dictation software spells for you. A constant video log of your life shows you a reality that memory doesn’t.
This distinction may be one factor in Apple’s excellent run in the last fifteen years, a realization I came to while reading an Economist article about Satya Nadella‘s changes at Microsoft. We might have had an iPad years sooner, if Microsoft had let it not be a Windows device. Bill Gates’s insistence on using the Windows interface as a precondition to all devices finally limited his vision.
Apple’s view of device and software as one, complementary and necessarily integrated, freed it to discard old software ideas when building new devices. iOS, the operating system for the iPhone, worked so intuitively that back in 2007 it seemed a little childish, poking icons instead of moving mouse cursors around or keying Alt-F4. What Apple saw was that a touchable screen had made the desktop obsolete, and there was no point in carrying it over. For a watch, the screen is too small to be the whole of the interface, and Apple again designed accordingly.
The wrist-worn computer won’t replace the smartphone, any more than a phone is a place to write a novel or update a 600-cell spreadsheet. The trick is to decide from the outset what a device is for, not to shoehorn in what other devices already do.