User Interface and Toilets Several years ago for reasons too tedious to mention, I found myself getting my hair cut in a salon in New York popular with Japanese expats. All was going well, with only the usual amount of cultural misunderstanding, until I tried to go to the bathroom. Toilets in Japan are wonders of ingenuity with myriad buttons powering such an array of jets and sprays as to make the fountain at the Bellagio pale by comparison. All I wanted was the standard, western flush and that button was proving difficult to locate. The button icons didn’t match any toilet function I knew. Eventually, after a brief, but memorable hydrodynamic interlude, I assumed the button with the most wear and tear was probably the correct one. It was. So I was pleased to learn that a trade group of Japanese toilet manufacturers has agreed on a new, standardized set of icons. Apparently, the Japanese are preparing for the influx of western visitors for the 2020 Olympics in the most practical of ways – by making it easier to go potty. Why does this matter? User interfaces have been on my mind lately. Currently, the trend in UI is starkly bifurcated. On the one hand, we have user interfaces that are so frictionless that they actually disappear (like Alexa.) On the other hand, there are user interfaces with a totalizing complexity as if their function was to discourage use by the technically unqualified. With such user interfaces, you are either steeped in the culture and idioms of the technology or you have no hope of mastering it. I think this bifurcation reflects a frustration on the part of the technology industry with translating complex functions into simple interfaces. Engineers don’t like to explain things in laymen’s terms. So they offer the user either no detail or far too much. Our user interfaces are confusing because the culture of technology is still in its infancy. We don’t have a shared language of interface design to inform our approach and the engineers driving the technological revolution don’t value design enough. Eventually, like Japanese toilet manufacturers, we will have to come to some agreements for the good of our users. In a nutshell: Don’t get too creative with your user interfaces – allow precedent to be your guide. Read More Voice and User Interface CES was a coronation of sorts for Alexa. Since Amazon decided to open-source Alexa, voice control has proliferated to a huge range of different devices. In a recent article, Liz Stinson of Wired talks about some of the challenges of using this technology when multiple devices within the home are voice activated. Amazon is trying to control some of the resulting confusion by including some standards in their developers kit. But the ultimate solution may be to have one central hub in the home for voice controls that distributes tasks to the appropriate connected device. That way both your blender and your fridge aren’t trying to order laundry detergent at the same time. That being said, there are bigger challenges ahead for a voice-activated user interface. Why does this matter? The screen offers a huge amount of information and differentiation for a user interface. At times, this fact has been abused to create user interfaces that are too complex. But voice commands offer none of that immediate information and feedback. Siri proved tedious because it was easier and faster to just use the screen for most functions. The more we trust Alexa with, the more complexity we have to add either to the voice command (as input) or to the underlying logic. The ambiguities of spoken language make the latter impractical. The system shouldn’t be relied upon to guess the user’s intent. Which leaves us with progressive complexity of input. Alexa boosters would have us believe that most digital interactions can be reduced to something as simple as: “Alexa, buy more detergent.” That is not my experience. My equivalent would be “Alexa, buy more detergent, but not the kind we bought last time. I like the one with no scent that comes in a white bottle. I think it has blue lettering. Or maybe it’s green. And not that organic stuff. It didn’t work at all. And don't order from the same company as last time. They said it would take a day, but it took a week. Try that other one. something-something.com. I think their logo is a fish.” Tell me that doesn’t work better on a screen. In a nutshell: Voice controls are great, but they will augment, not replace screens. Read More User Interface and Cars The always insightful Ben Evans wrote a recent blog post on the evolution of the user interfaces of cars. In his memorable analogy, cars today are like a Nokia feature phone from 2007 – “they’ve added all the smart stuff, badly.” On a modern car dashboard, functionality is nested – buttons within buttons and the org chart of the car maker is readily visible in the way functions are divided. Companies like Motorola were eventually saved from their incompetence in designing an operating system by iOS and Android. Evans suggests that the car’s operating system will also eventually be designed by a software maker. Why does this matter? When it comes to software, don’t build what you can borrow or buy. The car companies have always had a bias towards controlling every aspect of the design. This comes from an institutional prejudice that one is either a “car guy” or not. However, car interfaces are not materially different from office interfaces or kitchen interfaces (or toilet interfaces). All have the same group of users – the general public. Every industry that begins to develop digital interfaces starts from a place of insisting they need to build in house from scratch. And all of them grudgingly admit with time that developers of software are better at developing software. In a nutshell: Don’t build what you can borrow or buy. Read More