Today’s technology and media pose a distraction. We are overwhelmed by information and alerts. We are weighed down by big data and we struggle against information overload. We wake up to technology and fall asleep next to it. In 2010, Hans Vestberg, CEO of Swedish Telecom Ericsson, predicted 50 billion wirelessly connected devices by 2020. In the technology and business world, this idea genuinely excites people. They ask: How many chips can we produce as part of those devices? How many can we own? How many things can we connect together? During the mid-1990s, Xerox PARC researcher Mark Weiser predicted that the scarcest resource in the 21st century will not be technology, it will be our attention. We live in that era now.
One of the original promises of technology was to augment our lives and expand our capacity, time and creativity. The technology was supposed to increase our available time and rescue us from the mundane human tasks of modernity. Weiser urged us to have “smarter humans” not smarter devices. How can we build a future where devices worked alongside us, instead of against us? Where our technology can amplify the best of humans while giving us more free time? Many “smart” products fail. Many do not deliver services that are genuinely needed, and often, the privacy, security and connectivity problems outweigh the real benefits. Many smart products are also written in new and complex languages that lose support or funding, rendering these products non-functioning in only a short period of time. One of the problems with smart technologies is that they are not tested in suboptimal conditions. Does the app work when there’s little battery life left on the phone? Can you still turn on and off the lights in your house if the wifi is down? The first sophisticated computing technology to come into our homes was the desktop computer. It had a wired connection, a power cord that delivered electricity directly from the wall, relatively stable software, a set screen resolution, and relatively few options for customization. The desktop computer was able to command all of our attention safely, but mobile devices and products are something that shares our attention sometimes. Yet we still design “desktop-like” devices as though we can safely assume the user’s full attention. Smart products, be they phones or watches, are usually advertised being used in optimum situations, for seemingly perfect lives. The setting is a luxury condo in San Francisco that is perfectly quiet. The user is a twenty-something developer with a perfect American accent. When he speaks, the voice assistant in his home understands him perfectly every time. Contrast this with the real world, where a non-native English speaker struggles to get her voice-activated assistant to recognize a command while walking down a crowded city street. Most suboptimal scenarios are not thought through before market delivery, creating mismatched expectations. In 1995, Weiser and his colleague John Seely Brown coined the terms “calm computing” and “calm technology” as a reaction to increasing complexities of information technologies. They felt the promise of computing systems was to “simplify complexities, not introduce new ones”. We can no longer design applications as we did for the desktop. The capacity for our attention is no longer there. How can we put information into the background so that we can have smarter, more present humans, not smarter devices?
Principles of Calm Technology
If we want to design products that enrich our lives, we should design them to work with us, not against us. Technology should require the smallest possible amount of attention to creating the desired effect. Technology should create calm. Imagine you are trying to watch a movie on your smart TV, and every time you start to watch something, it tells you to download the new software update. In the desktop era, you could pay complete attention and run an update. But in the era of ubiquitous computing, often these pieces of information come when it is not our intent to focus on that particular task. Placing these notifications in the right place changes the user experience. What if before you turned off your smart TV, it asks if you would like to install the update and turn off afterwards? All you have to do is change the timing of the notification, and everyone will allow the update to install after they go to bed. No need to delay watching a movie in order to take care of an update. Technology should empower us. The LUMOBack Smart Posture Sensor is a device that goes around your waist and it sends you a slight buzz when you are slouching. It captures your attention only when relevant. It does not send you text messages or push notifications. It gives you a direct physical understanding over time of whether you are slouching or not, with a simple message delivered only to you. In 2011, I had a diabetic employee who was excited to have an insulin pump installed in his body. The pump did not allow him to change the alert from a loud "BEEP". He couldn't hear the beep at concerts, but others could hear him beeping at weddings and funerals. One of the alert choices should have been a haptic, personal alert, as the insulin pump information was intended only for him, and the alert style should have matched. Technology should move easily from the periphery of our attention to the centre, and back, informing without overburdening the user.
Calm Technologies are all around us. They are successful when they just “work”. A great example is an electricity. It is ubiquitous, yet it stays in the background and commands very little of our attention. We can interact with it through a light switch, refrigerator or outlet. The promise of electricity was that it would be available one hundred percent of the time. We only notice electricity when it fails, and it rarely does in most of the developed world. Compare this to the frequent failures of wifi or cell networks. Mobile applications demand more bandwidth than web applications, but updates in network capacity lag behind. Granted, the cellular and wireless grids, which we have been developing since the late 70s (1G in Japan) and 90s (wireless Internet), are relative newcomers when compared to the electrical grid, which we began developing in the late 1880s.A kettle uses your peripheral attention to give you information that your water is ready. The signal is unambiguous, it has the same signal every time, and you can usually change notification if you need it to be quiet. I have a tea kettle that does not make a lot of sounds because it was designed for coffee shops where a whistle would be intrusive. A different environment often implies different priorities for designing a signal.
The right amount of technology is the minimum needed to solve the problem. A smart fridge has been developed that use smart locks to prevent consumers from eating sweets. Why lock someone out of a fridge when sweets are often stored in a relatively ‘dumb’ cupboard? Other fridge technologies remind us to pick up milk and eggs when we go to the store. We do not need a piece of technology to tell us to do that. You do not need your fridge to text you a message to tell you that the bananas have gone bad and you need new bananas. The whole point of a banana is that you can look at it and see that it has gone bad. A banana peel already provides as fantastic notification system based on colour. It is unambiguous and never fails. In this case, technology is not needed to solve the problem.We do need a piece of technology to notice that we have been driving for a certain number of miles and we are due for an oil change, or that the sink underneath our cupboard is about to leak.Oil change lights are calm. They do not send us a text message. It just appears. When we are in the car, we remember it. It does not intrude on the rest of our lives. Homes full of smart technologies are more dystopian than utopian. Heaven forbid you actually inherit a house full of smart technologies because then you have to figure out how to use all of them. Or if your partner managed them and you get a divorce. Technology should amplify the best of technology and the best of humanity. A person’s primary task should not be computing, but being human.The best automation systems work when we work in a symbiotic relationship with technology. Google is successful because bots filter results for us, allowing us to make the final decision.
Todd Huffman’s company 3Scan employs a tissue scanning robot to do the work of humans at fifteen hundred times the speed of human hands. This frees up time for biologists to do important cancer research, and the scanned results are sent to the voting system that lets doctors train machine learning techniques for better cancer prediction.
Newspapers can use algorithms to compile information about sports events, and technological aids can be helpful for research, but important news stories should always be written by humans. You cannot automate Dostoevsky. This content comes from lived experiences as a human being. When a voice interface sounds too much like a human, our expectations of it increase. The robotic vacuum cleaner Roomba does not have a voice interface, it has an unambiguous tonal language. The machine beeps excitedly when it is done a cleaning. When it is stuck it emits a sad tone. This communication method is universal, does not need to be translated into different languages, and is less annoying than a spoken interface. The challenge of voice interfaces is that if you have a slight accent or you mumble at all, your voice recognition system may not understand you. This forces you into using short, clear command phrases that are easier for the system to understand, but this turns us more into robots than humans as we modify our language to cater to the technology. Humans should not need to act like computers in order to communicate with technology.
And technology should work even when it fails. Technology should default to a usable state rather than break down completely. Many satellite radio services stop mid-song when the car drives through a tunnel. Text messages fail to send when users are on underground transport. We need to design for situations where conditions are not optimal.
We can borrow from the concept of “graceful degradation” used in web design. When properly constructed, a website might render beautifully on a fast computer, but also degrade gracefully to a lower level of user experience in older browsers. There are many ways to implement graceful degradation into products. Some hand-held radios have cranks as a backup if batteries run out, or a power cord to plug them into the wall. Escalators degrade gracefully into stairs when out of order. In video streaming, visual quality can degrade, but audio is kept at high resolution. We can handle low-resolution video, but low-resolution audio can take us out of the moment.
A lot of technology is created for optimum situations, not older products, bad connectivity, poor battery life, or distraction. These crucial edge cases must be considered meticulously during the product design and engineering process. Design for long-term function.
The Future of Attention
Mark Weiser died in 1999 at the age of 46, and he is no longer around to see the world he helped predict. He would not be surprised by the world he envisioned. PARC researchers were already discussing privacy, virtual reality and chat assistants in the late 80s. If we design our technology right, we get more time to be ourselves. We get more time to be human. We get more time with family, friends, and loved ones. We can look at sunsets and fall in love. No matter how much AI we build, a computer is never going to have the embodiment or context required to be able to appreciate a sunset. The Greeks had two words for time. One was “Chronos”—quantified, industrial time. The other was “Kairos”, or lived, qualitative time—the moments and memories that comprise who we are.
I believe that technology should not require all of our attention, only some of it, and only when necessary. It is possible to compress information into our peripheral senses and free ourselves in the process.
For more information, r
ead more on my book Calm Technology: Principles and Patterns for Non-Intrusive Design from O’Reilly Books.