‘The best design is invisible‘ is the interaction design phrase of the moment. The images above are from my ever-growing collection of quotes about how design and technology will ‘disappear‘, become ‘invisible‘, or how the ‘best interface is no interface‘.
The Verge has recently given both Oliver Reichenstein and Golden Krishna a platform to talk about this. It has spawned manifestos, films, talks, books, #NoUI hashtags and some debates about what it might mean. I’ll call this cluster of ideas ‘invisible design’.
I agree with some of what’s driving this movement. Design’s obsession with touchscreens really is a problem. I’ve spent the last eight years writing against glowing rectangles, studying our fascination with screens, and watching how that fascination has become a cultural phenomenon. In response I’ve been researching and inventing interfaces for pulling interaction out from under the glass.
I disagree with a lot of the rest, for reasons I’ll set out below.
1. Invisible design propagates the myth of immateriality
There’s already plenty of talk celebrating how invisible and seamless technology is, or should be. We’re overloaded with childish myths like ‘the cloud’, a soft, fuzzy metaphor for huge infrastructure projects of undersea cables and power-hungry data farms. These myths can be harmful, and are often just plain wrong. Networks go down, hard disks fail, sensors fail to sense, processors overheat, batteries die.
Computing systems are suffused through and through with the constraints of their materiality. – Jean-François Blanchette
Invisible design tells us technology will ‘disappear’ or ‘just get out of the way’. It glosses over the qualities of interfaces that make them difficult or delightful.
When we deliberately hide the seams of an interface, smoothing over the edges and transitions that every technical system has, we lose something. Designers lose a sense of what they’re working with. Users lose the ability to understand what’s happening. Without that understanding we fall back on folk-theories that make systems harder to use, and we lose the ground from which to criticise them.
As systems increasingly record our personal activity and data, invisibility is exactly the wrong model.
By removing our knowledge of the glue that holds the systems that make up the infrastructure together, it becomes much more difficult, if not impossible, to begin to understand how we are constructed as subjects, what types of systems are brought into place (legal, technical, social, etc.) and where the possibilities for transformation exist. – Matt Ratto (2007)
In other words, invisibility takes agency away from both users and designers.
2. Invisible design falls into the natural/intuitive trap
The movement tells us to ’embrace natural processes’ and praises the ‘incredibly intuitive’ Mercedes car interface. This language is a trap. (We should ban natural and intuitive, by the way.) It tells us nothing about how complex products actually become simple or familiar.
Invisible design leads us toward the horrors of Reality Clippy. Does my refrigerator light really go off? Why was my car unlocked this morning? How did my phone go silent all of a sudden? Without clear, readable systems for handling all this ‘smartness’, we’re going to get very lost and very frustrated. The tricky business of push notifications and the Facebook privacy train wreck are just the tip of the iceberg.
Take the Nest thermostat, which ‘learns’ your habits to control the home temperature. A good example of invisible learning. But the Nest has a highly visible interface that reassures you, tells you when it’s learning, and a large dial for adjusting temperature. Beautiful, legible microinteractions. A Nest without these visual and direct-manipulation interfaces would be useless, uncanny and frustrating. Nest wants UI.
Much of the discussion around invisible design points to sensors and tangible interfaces as the answer. But these systems aren’t any simpler or more familiar. They have their own material qualities, edges and ‘grain’ that still need to be understood and learnt. Their literal invisibility can cause confusion, even fear, and often makes things more unpredictable and more likely to fail.
In our work with interface technologies like RFID and computer vision, we’ve learnt that it takes a lot of work to understand the technologies as design materials. So it’s not useful to claim UI is ‘disappearing’ into sensing, algorithms and tangible interfaces, when we don’t fully understand them as UI yet.
3. Invisible design ignores interface culture
Interfaces are the dominant cultural form of our time. So much of contemporary life happens through interfaces, inside UI. They’re part of cultural expression and participation. Skeuomorphism is evidence that interfaces are more than chrome around content, more than tools for solving problems. To declare them ‘invisible’ is to deny them a cultural form, a medium. Would we say ‘the best TV is no TV’, ‘the best typography is no typography’, ‘the best buildings are no architecture’?
A lot of what we do at BERG is cultural invention, as much as it is problem-solving:
We’re not interested in this idea of the invisible technology in a modernist sense. Tech won’t be visible but only if it’s embedded into the culture that it exists within. By foregrounding the culture, you background the technology. It’s the difference between grinding your way through menus on an old Nokia, trying to do something very simple, and inhabiting the bright bouncy bubbly universe of iOS. The technology is there, of course, but it’s effectively invisible as the culture is foregrounded.” – Jack Schulze (in Domus 965 / January 2013)
We should be able to celebrate the fantastic explosion of diversity in UI while also developing a healthy critique of things like touch screens. But by calling for UI to disappear altogether in the name of efficiency, we stay stuck in the same utilitarian mindset that produces inert technological visions like this one. Interfaces are part of the cultural landscape. We should treat them that way.
4. Invisible design ignores design and technology history
The movement ignores at least thirty years of thinking in design and technology. A few examples.
Much of the recent invisible design discussion repeats the thinking in Jared Spool’s ‘Great Designs Should Be Experienced and Not Seen‘ and Donald Norman’s ‘Invisible Computer‘. A better reference point would be Norman’s earlier book, The Design of Everyday Things, where he talks about the ‘problems caused by inadequate attention to visibility’ and about supporting our mental models of systems. We need a lot more thinking about our mental models of algorithms in particular.
Adam Greenfield has investigated the social and ethical issues around the development of ubiquitous computing systems, and is particularly concerned by their disappearance:
“Ubiquitous systems must contain provisions for immediate and transparent querying of their ownership, use, and capabilities. Everyware must, in other words, be self-disclosing. Whether such disclosures are made graphically, or otherwise, they ensure that you are empowered to make informed decisions as to the level of exposure you wish to entertain.” – Adam Greenfield (2006)
Other designers have talked about the qualities they want from ubiquitous computing interfaces, such as polite, pertinent and pretty:
“The vast quantities of information that personal informatics generate need not only to be clear and understandable to create legibility and literacy in this new world, but I’d argue in this first wave also seductive, in order to encourage play, trial and adoption” Matt Jones & Tom Coates (2008)
Matthew Chalmers has, more than anyone else, uncovered the history of seamlessness. Seamlessness is ‘the deliberate “making invisible” of the variety of technical systems, artifacts, individuals and organizations that make up an information infrastructure. This work actively disguises the moments of transition and boundary crossing between these various parts in order to present a solid and seemingly coherent interface to users.’ (Ratto 2007). Although Mark Weiser is often cited as an advocate of seamless systems, Chalmers found otherwise:
Weiser describes seamlessness as a misleading or misguided concept. In his invited talks to UIST94 and USENIX95 he suggested that making things seamless amounts to making everything the same, reducing components, tools and systems to their ‘lowest common denominator’. He advocated seamful systems (with “beautiful seams”) as a goal. Around Xerox PARC, where many researchers worked on document tools, Weiser used an example of seamful integration of a paint tool and a text editor (Weiser, personal communication). He complained that seamless integration of such tools often meant that the user was forced to use only one of them. One tool would be chosen as primary and the others reduced and simplified to conform to it, or they would be crudely patched together with ugly seams. Seamfully integrated tools would maintain the unique characteristics of each tool, through transformations that retained their individual characteristics. This would let the user brush some characters with the paint tool in some artful way, then use the text editor to ‘search and replace’ some of the brushstroked characters, and then paint over the result with colour washes. Interaction would be seamless as the features of each tool were “literally visible, effectively invisible”. Seamful integration is hard, but the quality of interaction can be improved if we let each tool ‘be itself’. – Matthew Chalmers (2003)
Matt Ratto investigates the darker side of this drive toward invisibility, arguing that seamlessness encourages:
“a particular kind of passivity and lack of engagement between people and their actions and between people and their social and material environment” and that we must “critique the clean, orderly, and homogenous future that is at the heart of these modernist visions” – Matt Ratto (2007)
Anne Galloway suggests that it is in the seams where the design work can be done:
“Although seamlessness may remain a powerful and effective metaphor to guide particular projects, when it comes to actually getting the work done—and the challenges of having to do it with people who can be very different from each other—then I suggest it is in everyone’s best interests to recognise the importance of seams and scars in marking places where interventions can be made, or where potential can be found and acted upon.” – Anne Galloway (2007)
In interaction design we need to look at the long history of Durrell Bishop‘s work. He is one of the strongest advocates for self-evident design, whether physical or virtual, through his teaching and design practice. Durrell’s ‘Platform 12’ in the RCA Design Products course attempts to see design as:
“a celebration of a model for how things work, where once again we can treat function as beauty, instead of merely treating design as form and image.”
Durrell’s work on the Marble Answering Machine (1992) is a brilliant piece of self-evident design, and remains a touchstone for all interaction design work.
Designers also need to read the first four chapters of ‘Where the action is‘ by Paul Dourish, a clear-eyed account of how human abilities and computer interfaces have developed together over the last fifty or sixty years. Dourish shows interfaces becoming more social and more tangible. They’re becoming more present, not less.
And finally, from a design perspective, there is a long tradition of making complex products legible and understandable. Industrial designer Konstantin Grcic talks about the relationship between the technologies and the use of an object:
“A machine is beautiful when it’s legible, when its form describes how it works. It isn’t simply a matter of covering the technical components with an outer skin, but finding the correct balance between the architecture of the machine… and an expressive approach that is born out of the idea of interaction with those using the object.” – Konstantin Grcic (2007)
And perhaps more famously, Dieter Rams has always spoken of honesty and understanding in his product design practice. Making a product understandable is one of his Ten Principles of “Good Design”.
This drive for understanding needs to go further than physical form (as it has done at Apple) and start to inform the design of systems and UI.
Towards legible, evident interaction
We must abandon invisibility as a goal for interfaces. It’s misleading, unhelpful, dishonest. It leaves room for unusable, harmful and frustrating interfaces, and for systems that gradually erode users’ and designers’ agency. Invisibility might sound attractive at first, but it sidesteps the real, thorny, difficult work of designing and using complex systems.
We might be better off taking our language from typography. Talking about legibility and readability, while keeping room for the way typography can also call attention to itself in beautiful, spectacular ways. Our goal should be to ‘place as much control as possible in the hands of the end-user by making interfaces evident‘.
The interfaces we design may become normalised through use, effectively invisible over time. But that will only happen if we design them to be legible, readable, understandable, and to put culture in front of technology. It’s how you build trust and confidence in an interface in the first place, enough that it can comfortably recede into the background.