With the right balance of enthusiasm and prudence there's no reason we can't be members of that community.
I hope to meet you all there, on the other side of Singularity. I'm looking forward to it. I hope you are, too.
Hypermedia and the Singularity
-A Child Dying of Adrenoleukodystrophy
-Sanskrit literary style
-Buttons to begin an Article on Hypermedia and the Singularity
-Flight in Information Space
-Definition of Hypermedia
-Road Map
-Definition of Singularity
TO SEE ONE OF THESE SECTIONS, JUST POINT AND CLICK
(oops—this is a paper document, not a computer document).
Road Map
This article is about the relationship of the technology of hypermedia to the approaching time of technological Singularity. There are a lot of ways we could start this discussion; up above, in italics, you see a list of the starting places that I considered before writing. The article, as it now stands, has the following layout:
1) the list of buttons (the section in italics at the beginning of the article),
2) the road map (that's where we are now),
3) a major section to define hypermedia, with sidesteps to consider:
a. Sanskrit literary style
b. flight in information space
c. hypermedia art, and
d. issues of hyperstyle
This lengthy discussion of hypermedia is followed by:
4) a shorter definition of the Singularity, and
5) a discussion of how the Singularity and hypermedia are interrelated. This discussion of interrelationships wraps up with an example of how hypermedia will accelerate our approach to Singularity: the story of the child with adrenoleukodystrophy.
Finally, the article ends with:
6) a discussion of the next steps in hypermedia development, who is taking those steps, and where it will lead.
Definition Of Hypermedia
Hypermedia is much easier to use than to define. In one sense, you have already seen a definition of hypermedia in the early layout of this article, though in practice it's difficult to grasp without a computer-based example.
Hypermedia is the child of hypertext. Ted Nelson coined the term hypertext in the sixties and defined it simply as "nonlinear writing." 1
Linear writing has been mankind's standard for millennia. One alphabetic character follows another, one word follows the next, building sequential sentences, paragraphs, and chapters. The writer designs his document for a reader who is trapped: the writer assumes that the reader only has the ability to go forward one step, or backward one step, but nowhere else.
Of course, we have had limited forms of nonlinear writing mixed in: the table of contents and the index are modern (though primitive) nonlinear writing tools—though slow, they do help the reader skip to the sections of the document of most interest to him.
Sanskrit Literary Style
Nonlinear writing goes back at least as far as Sanskrit. With one of the stylistic approaches used in Sanskrit, the document's opening passage was a series of one-line descriptions of what would follow. The next section contained a paragraph for each one-line description; the next section devoted a chapter to each paragraph description. This design encouraged the reader to skim only as far as he needed to go, reaching into the deep, extensive discussion only as a last resort. The article that you are reading now, with the list of items as its first paragraph, is organized in a way similar to those ancient Sanskrit documents.
Newspapers also encourage nonlinear reading—the headers for the different articles appear in a bolder, larger style, which the human eye can automatically pick out (the powerful perceptual computers behind our eyes that do this automatic selection are a major reason why people find Macintosh-like, icon-oriented software easier to use—using imagery, we can grasp many features without recourse to our conscious reading abilities). After the eye has picked out an interesting article, the first journalistic paragraph summarizes the whole article—only the most in-depth reader must go beyond that first paragraph.
But this is exactly where newspapers fail—the in-depth reader must page back and forth from the front to the far back just to read a whole article.
And while the newspaper frustrates the in-depth reader, the textbook with its index frustrates the skimming reader, who has no real way of perusing just the summaries. The nonlinear extensions to linear books fail because paper is inherently a linear medium. Enter the computer.
Modern desktop and laptop computers have grown powerful enough so that they can give us a truly nonlinear medium for document presentation. No longer must we ask the question "should I set my article up like a newspaper, for skimmers, or should I set it up like a book, for detailed readers?" Set it up for both.
With either Guide2 or HyperCard™ 3 (the two widely-accessible hypermedia tools at the time of this writing), the table of contents can be a series of one-line entries that are treated as buttons: when the reader points at the entry and clicks, the computer brings forth the detailed backup information in the twinkling of an eye. This detailed information can in turn contain other buttons.
Buttons do more than link a brief description with its detailed explanation. Buttons can link multiple, partially related items. A paragraph describing a disease might have a link to a separate paragraph about the cure, which might be linked to a list of related medicines, each of which is linked to a manufacturer, each of which is linked to a list of products, each of which is linked to a list of diseases for which that product is the cure. In this sense hypermedia offers instantaneous references, akin to the suggested reading lists tagged on to encyclopedia articles.
The elements of a hypermedia document do not all have to be text—they can also be pictures, sounds, and full-color videos and animations. For many documents, the table of contents should not be a list of chapter names, it should be a picture. A car repair manual, for example, might have a picture of the car as its first item. The mechanic would point to the part of the car he needed to know about, which would give him a closeup view and a short textual description of what it is and how it should work, with a few extra pictures (through other buttons) of typical forms of wear that would call for replacement. The mechanic would zoom in again and again until he found the specific part to replace—and he would then press the button that runs a short video sequence showing how to remove the old part and install the new one (for a simple example of zoom, see Figure 1). With these links interconnecting pictures, text, and video sequences, we have true hypermedia (for the syntactically finicky, the word "hypermedia" is a singular group noun, unlike the word "media," which is the plural of medium).
The links inside a hypermedia database allow the hypermedia reader (a hyperreader?) to leap through a document as quickly as today's linear reader can turn a page.
Indeed, the whole concept of a document—a standalone volume of text and pictures—becomes less meaningful with hypermedia. As more documents are added to the hypermedia database, with rich crosslinks to other documents, the reader finds himself browsing, not through documents, but through an information space. And information space, like normal space, is designed for flight.
Flight In Information Space
I remember witnessing hypermedia for the first time at the Microsoft CD-ROM conference in February of 1987. Owl International was announcing Guide, the first commercial hypermedia system. The presenter pointed at a line in the table of contents, expanded that section to show a list of subsections, and quickly hit four buttons. In an eyeblink, four new windows popped open on the computer screen, each showing a different section of the document, three with pictures. He continued to click, bringing new windows to the fore with new information. I experienced a momentary sense of disorientation. That sensation quickly developed into a sense of breathless movement, of flight, the feeling I usually reserve for watching the stars flash past on Star Trek.
/> With the advent of hypermedia, the quiet but explosive revolution of the "paperless office" draws close at last. The paperless office received much acclaim years ago, but the vision faded as computerization actually expanded the creation of paper. This disillusionment with the paperless office came about through a tragic misunderstanding. The vision of the paperless office, however dim, was correct.
When a new technology is introduced, people's natural first reaction is to use the technology to do the same old tasks more quickly. Thus people first used computers to create paper—and they succeeded beyond their wildest nightmares. Computers have dramatically increased paper production—but they have not, by any current measure, increased productivity. This is about to change, and will change with ever greater speed for the next decade as we build tools that make computerized data more effective than paper counterparts.
Hypermedia is a key ingredient for creating that effectiveness (though other ingredients are still necessary to match the merits of paper, namely durability, portability, and resolution; these, too, will be solved, but that's another article). Once data is put into a hypermedia-based information space, it is easier to retrieve and easier to read than it would be on paper. People will not want to print hypermedia documents: the translations will lose so much value, writers will instead give readers access on the computer. As all the worker's data begins to show up in the same format, interlinked with all the other data so that he can toggle back and forth at the touch of a finger, productivity will start to rise.
Alas, in our modern society, improved productivity no longer guarantees improved performance. Bureaucrats can always increase the demand for paperwork (or computerwork) to negate any increase in productivity. Technological solutions to more classical problems have not faced such an unbounded obstacle. Not even the government will move a city ten times as far away from you just because your new car goes ten times as fast. Bureaucrats can, however, require ten times as much paperwork once you get there. Consider the effect of our recent tax simplification: it made bestsellers out of thick books about taxes. It seems clear that bureaucrats will require even more paperwork.
Hypermedia Art
Besides giving a big boost to the paperless office, hypermedia will give rise to a new form of art to stand beside painting, cinematography, and literature. Indeed, hypermedia may become the culmination of these separate lines of artistic expression, as it weaves the now-disparate genres into a stunning tapestry. In a hypermedia novel, the reader has much more control over which pieces he reads—he might choose to follow a single character through the course of the story one day, and a different character the next. Winds of War by Herman Wouk might be much more readable in this fashion. David's Sling, the world's first hypermedia novel, allows the reader to follow not only charactersrs but also subject topics through the period of time chronicled by the story (see Figure 2). 4
Following multiple character threads in this manner opens up even more unusual possibilities. The reader may revel in reading a single scene from several different points of view—Roger Zelazny's Amber series offers some interesting hypermedia possibilities in this regard. And the reader would benefit from hypermedia when engaging Heinlein's future histones Just because he could finally find things.
In SF novels in particular, hypermedia gives the writer a Way of sharing his mountains of background material with the interested reader, without imposing on the tight construction of the plot. David's Sling has a separate section of Blueprints for those who desire more technical information about the Sling Hunters (see Figure 3). Indeed, in David's Sling a reader skilled in the arcane mysticisms of modern management can read the entire story with the Program Evalution and Review Technique (PERT), in a series of PERT charts (see Figure 4). This is almost certainly the first time that management science has been intentionally used for artistic expression (though from what I hace seen, management science is often used unintentionally to create works of fantasy).
Future hypermedia art will require ever more innovative intertwining of graphics with text and animation. The development of truly great hypermedia documents—whether they be pure art, pure information, or a weaving of the two—will require an array of skills that includes illustration, writing, cartoon creation, and cinematography. The most urgently needed skill will be a new one that might best be called link architecture: the design of sets of links that offers the reader intuitive flight paths. Whereas modern writers only have to worry about the transition from one paragraph to the next paragraph, the hypermedia author will consider dozens, perhaps even hundreds, of such transitions. Part of skillful design will be to keep the number of transitions, the number of branches, small, while still guaranteeing fast and understandable traversal of the whole document.
Fortunately, these designers will have hypermedia style guidelines on-line to assist them.
Hyperstyle
Some discussion of hyperstyle here may help illuminate the meaning of hypermedia.
Hypermedia designers, just like ordinary writers, would do well to start with the Elements Of Style by Strunk and White. As Strunk would say, "Omit needless words. Omit needless words! OMIT NEEDLESS WORDS!"5
Items in hypermedia should generally omit even more words. In linear literature, one must occasionally summarize terms and ideas from other chapters to guarantee that the reader has the proper context for the current discussion. In hypermedia, however, the author would simply plunk down a button linked to the explanation, leaving the reader the choice of plunging ahead or getting a refresher.
Bold writing, another major thrust of Elements of Style, also becomes easier. In linear writing, how can one address complex problems that demand complex analysis? Throughout history, speakers who have strived for accuracy have labored at a disadvantage: Men of reason find that true statements require so many qualifiers that their sentences turn to quicksand—and their audiences turn to stone. Meanwhile the short-sighted sloganeer can stir the soul with his quick, simple, wrong answers.
Given hypermedia, reasonable men can make bold, powerful statements, slugging it out slogan for slogan with fanatics—with one telltale difference. Numerous buttons will surround the reasonable man's bold statement, allowing skeptics to branch to more careful supporting arguments.6 Writers will thus discourse in a way similar to the way we teach physics to children. All of our early education, about little electron and proton spheres whirling around, is a pack of lies—but the lies are eminently satisfactory unless you need that deeper understanding, easily achieved in later courses.
This hypermedia approach to bold writing, incidentally, could create a new legal problem if widely used. We might want to consider laws prohibiting the paper printing of hypertext without the author's explicit permission: such printing, which would rip the bold statement from its supporting links, could supply the unscrupulous opponent with the ultimate tool for quoting out of context.
Critics repeatedly cite two fears of hypermedia. First, they fear that you won't be able to find things, i.e., that the information will be "hidden" someplace. A related fear is that you won't be able to find yourself, i.e., that you will become lost in the maze of buttons. The sense of disorientation I felt upon first witnessing hypermedia, that sense of "Star Trek Warp Speed," is only pleasant for people on a lighthearted romp; it is catastrophic for people trying to do their work.
The worries about hidden information are probably overblown. There are two causes for this fear. First, one can make legitimate complaints about hidden information in most of today's hypermedia documents. All of today's hypermedia authors are necessarily novices; they are still experimenting with what works and what doesn't. Hiding information behind concealed buttons is cute, fun, and alluring for the beginning writer; but those writers that respect their readers will quickly outgrow the urge to play games. Hidden buttons may have a place in hypermysteries, but nowhere else.
The second cause of hidden information phobia is that people compare hypermedia to a hypothetical per
fect system, rather than to an everyday paper system. Even with hypermedia assistance you'll occasionally lose items. You'll know that a critical datum resides somewhere in that information space, but you'll know you'll never find it.
But you'll find it more often in information space than in paper space. In my office, with two huge filing cabinets bursting with linear paper, information loss occurs every day (my file cabinets are organized a lot like Heinlein's future history). If I could reduce the loss rate to once a month, it would transform my life (of course, people who organize their file cabinets more carefully have fewer problems in paper space. But people who arrange their buttons more carefully will also have fewer problems in information space; they will literally retrieve items in the blink of an eye).
The other oft-cited fear, of losing yourself, is considerably more serious. Disorientation is virtually universal for people encountering a hypermedia document for the first time. A key to successful hypermedia construction will be the creation of the road map that shows the reader how the pieces are interconnected. The road map will be the visual presentation of the link architecture mentioned earlier. A sample road map can be seen in Figure 5.
Novices in hypermedia design often predict that the road mapping problem will be solved automatically, that future hypermedia presentation systems will magically build the maps themselves. After all, the computer knows where all the links are, why not let the computer build a composite picture of them all? Of course, the computer can build such a picture, and automatic map-building will supply a useful tool to the developer of an information space. But there are thousands of possible map designs for a given information space, based on different graphic arrangements of the objects in the database. Only a human being with a talent for extracting order from chaos can draw an understandable map, with subject-oriented symbology, with straight paths for the "main highways" that readers will often take, and with meandering branches for the links that are less traveled.
The Gentle Seduction Page 26