The AntiAssumption Interface
Unfortunately, we’ve now reached the limits of the current GUI paradigm. Displaying commands in menus, toolbars, and dialog boxes works with a limited number of elements. But Microsoft Word 2003 has 1,500 commands, and users typically have no clue where to find most of them.
Another WYSIWYG downside is that it forces too much manual labor on users and requires a stretch of imagination to envision results in advance.The next version of Microsoft Office (code–named “Office 12”) will be based on a new interaction paradigm called the results–oriented user interface. As the demos show, the most obvious departure from the past is that menus and toolbars are all but wiped out. The focus is now on letting users specify the results they want, rather than focusing on the primitive operations required to reach their goals.
—Jakob Nielsen’s Alertbox, October 10, 2005: R.I.P. WYSIWYG
The AntiMac Interface by Don Gentner and Jakob Nielsen holds an interesting premise — even outside the specifics of the paper (which I’ve strayed from substantially here). In response to the lament of GUI experts that interface design seems fossilized, Gentner and Nielsen point to a mental exercise scientists use of thinking how things would be were key assumptions false. This has resulted in a number of innovations now seen as commonplace, including entire branches of study in the sciences. In applying this concept to human–computer interaction and interface design, the paper may mark a turning point in software design.
People do not resist change per se. They resist conventional change. They resist change they don't understand or change that is imposed on them. They resist change that threatens or interferes with their priorities.
— Creative Solution Finding, Gerald Nadler and Shozo Hibino with John Farrell (Prima Publishing 1995)To properly explain you need an example of an invisible assumption, the problem with the existing assumption, and how things could be different. Take paper…
Paper with left–to–right text is a technology. An extremely old one we don’t consciously think about any more, but a technology nonetheless.
Basically this concept, this paradigm, is what you’re looking at now — a scrolling text field. Or, in other words, you are using 21st century hardware running a paradigm suited to the information needs of Medieval society. Electrifying it and pushing it past its operational limits doesn’t change that essence.
The Problem
The current UI is still tied indirectly to the PC’s original root metaphor, a typewriter. …In short, we need to recognize that the role of the user has shifted from passive consumer to active director, someone continually choosing from among multiple feeds to construct the desired experience, and reconstruct the UI to serve that new end.
—Geoffrey Moore
A Medieval paradigm would be an improvement, as taking this old technology and moving it essentially unchanged onto the computer gives a performance hit — users read a little slower on computer than paper, and due to various reasons it’s subliminally annoying to the user. Which is one reason paper is still, in a number of ways, better than the “information appliance” seeking to replace it.
It doesn’t have to be that way.
The mechanics of how humans read has undergone many tests. One element that came out of these tests was a new way to display text on computer. Instead of taking each line of text and forcing the reader to waste time following them, they broke the text file down. They took each word, blew it up to, say, 90 points and rapidly flashed each word of the file on the screen, one word at a time.
When you do this rapidly, you find the text grows more comprehensible and finally downright slow. This process of speeding up, getting accustomed to the new speed and then increasing it goes through several cycles. The process happens nearly instantly and subconsciously; within the first minute my own reading speed tripled. And I have demonstrated similar machine assisted reading software to classrooms full of people with comparable results.
The result is average users read at speeds of 1500 words per minute. And since I’ve demonstrated this to dozens of people face–to–face I already know some of what you're thinking. …And yes, you have the exact same retention rate of what you read. No better, no worse than regular reading.
A Solution For Data Smog

Zap Reader, a browser based interface
Information exchange between computer and user is the economically effective bandwidth computer bandwidth depends on; increasing an organization’s ability to understand and react. Simply increasing the bandwidth between computers, or cranking up the speed of traditional interfaces causes user burnout and resistance to change. In Creative Solution Finding: The Triumph of Full–Spectrum Creativity over Conventional Thinking, Gerald Nadler and Shozo Hibino write, “People do not resist change per se. They resist conventional change. They resist change they don't understand or change that is imposed on them. They resist change that threatens or interferes with their priorities.”
It may simply be that people, being sensible, resist change which is so conventional as to offer little real benefit over the status quo. It is up to the software developer to offer new benefits which give users more effective results, not just new features.
These MARS (Machine Assisted Reading Software) programs are a prime example of an anti–assumption breaking the rules. The assumption anything nonstandard as having an exponentially longer learning curve — yet through dozens of observations I’ve seen instant results. Why? Because users aren’t learning about the computer so much as their own latent potential — which is the true goal of user–centered software design.
Resources
- Crossing The Chasm author Geoffrey Moore on Renovating the User Interface.
- BUDDYBUZZ allows you to read 700 to 800 words per minute on your cell phone display. ZAP Reader (beta) is an online version of the interface design talked about here.
- The Antimac Interface An article about the value of questioning assumptions. Jakob Nielsen’s Alertbox, October 10, 2005: R.I.P. WYSIWYG introduces Microsoft Office 12 and the results–oriented user interface. Related Design Crux article on moving from features to benefit–directed design.
- A backgrounder explaining data smog unfortunately uses data and information in an interchangeable manner. David Shenk’s book describes symptoms, but my premise is output productivity exceeds input productivity as a major cause of data smog. It is simply easier to deliver many reports, emails, and web pages then it is to sort through or make sense of them.
- Examples of MARS software: This software or maybe that software. There are several types, none of which is exceptional. Good as proof–of–concept, but not so great as an application. Needed: One dozen similar improvements integrated under a coherent focus into a full fledged application interface. Although it is always possible I am overly picky, as one company built a long–running informercial around this kind of software. Don’t let that discourage you from coming up with reasons why the ideas you find here can’t work in your special case.
- Creative Solution Finding: The Triumph of Full- Spectrum Creativity over Conventional Thinking, Gerald Nadler and Shozo Hibino