Dr. Dobb's Journal December 1999
The clinical term for the dilemma of modern man is mania. So some would say. Like someone on speed, many of us find that we can multitask like mad, but can't sit still to listen to a symphony. Stewart Brand says that we only do short-term memory. He's talking about both our thought processes and our data-storage media, and he's got a point either way. But even if we are manic, is that necessarily a bad thing? James Gleick looks at our mania in a new book called Faster, which I review this month. And Gregory Benford, in another venue but on the same wavelength, predicts the end of prediction.
Some would call a 40-year commitment to a product that has never shipped a kind of mania. I'm talking, of course, about Ted Nelson and Xanadu, his grand vision of computing that just went open source this year. The open sourcing of Xanadu is a genuinely historic event in the history of computing, of communication, of art. At least it will be if it results in something like the Xanadu vision someday becoming reality. For that, some of the rest of us must now join Ted in his mania, and that's just what he's inviting us to do.
"Ideas have a long half-life with me," Ted Nelson told me over dinner in a Monterey restaurant in September.
Indeed they do. Ted was in town to give a speech in which he would announce that his 40-year-old baby, Xanadu, was going open source. I was in town to hear him, and to have dinner with Ted and the current Xanadu team the night before the speech. To understand why I had driven 400 miles that day to meet Ted, and why Eric Raymond, the guy who gave open source software its name, was chosen to introduce Ted the next day, you need to know the background.
In 1960, Theodor Holme Nelson had a vision.
As articulated brilliantly and entertainingly in several books and magazine articles and white papers and many unforgettable speeches over the succeeding years, it became a devastating critique of computing as it is done, and Ted's list of things that are broken in the current model is legendary: the desktop metaphor, hierarchical file structures, applications, the clipboard. Ted is no fan of the Macintosh, which he says took away from most of us the ability to program the machines. But the vision in 1960 was simpler: He wanted to edit text on a cassette tape, and thought he saw how to do it. Familiar with video editing techniques, he extrapolated from them to text. Store all text exactly once, somewhere, he said, and maintain links to the elements of the text. Then just edit the links.
It addressed the problem of editing in a sequential medium-like tape. But by leaving the text alone and doing everything with the links, you gain many other capabilities. You can maintain multiple versions, in fact every version, of the document; you can have exactly one copy of any text and merely clone the links. Other insights included copyright protection and micropayments, and documents distributed across many machines.
He called it Xanadu, and it has been built twice, the second time under the auspices of Autodesk. Programmers who must be mentioned in the long history of Xanadu include Roger Gregory, Mark Miller, Hugh Daniel, and K. Eric Drexler.
Although it is a system of linked documents intended to hold all human knowledge, Xanadu is fundamentally different from the World Wide Web, at least in the Web's current form. Xanadu links are not just navigational but structural. With HTML or XML you manually insert tags, with Xanadu you just write and edit. The links are not merely consciously created visible user-interface devices, but are the basis of how documents are created, copied, quoted, and edited.
It's all going open source. Ted is retaining the trademarked name Xanadu, the open-source code being licensed under the name Udanax. Both versions of Xanadu are posted at the http://www .udanax.com/ web site, the original version, now called "Udanax Green," and the Autodesk version, now called "Udanax Gold." Both bodies of code are available but you're pretty much on your own with the Gold version.
Three months ago I wrote about Stewart Brand's The Clock of the Long Now, a book about how we deal with time. This month I review another time book, this one by James Gleick, whose past books, Chaos and Genius, impressed me. His latest, Faster (Random House, 1999; ISBN 0-679-40837-1), is also impressive.
Gleick does a great job of documenting the little evidences of the speedup of our lives. Overnight, it seems, Peter Jennings and Tom Brokaw have become reflective elder statesmen of the news, the real reporting happening online. What does it mean that these sound-bite spewing talking heads (and I mean that in the nicest possible way) now seem deliberative, and television has come to be seen as a slow medium? TV producers know fits happened: Witness the phenomenon of ironic commentary overlaid on what once passed for hot programming. Beavis and Butthead commenting on last year's videos on MTV, Pop-up Video captions annotating last year's videos on VH-1, and the best of the artform, Mystery Science Theater 3000, ridiculing old science fiction movies. These things, once cutting edge, are now too slow and boring and old fashioned to watch without a separate show overlaid on top of them. The floor of the stock exchange, long taken as a symbol of frenetic activity, has come to the living room: A whole pastime (profession? addiction?) has developed of people sitting at home in front of their computers all day trading stocks. The businesses represented by these stocks are no less frenetic: Companies are expected to go from startup to IPO in 18 months, and product cycles have shrunken dramatically. A widely cited result says that it's better to bring a project in 50 percent over budget than six months late. Just-in-time delivery spawns just-in-time accounting and just-in-time training as the gears of industry grind toward a fabled frictionless economy in which an order from a customer doesn't just trigger an order to a supplier, it becomes an order to the supplier. Doctors complain that they are doing beeper medicine: When you live an interrupt-driven life, all activity becomes crisis management. The instantaneous nature of communications media make us so connected that news can become locked into positive feedback loops, giving us stories that won't die: O.J., Monica, El Nino, Y2K. We have fast-track legislation and fast-track drug trials, and if you can't stand waiting out the intertrack gap between songs on your CDs, Sony will fast-track those for you.
One of the few things actually slowing down in Western countries is driving. Gleick has some fascinating information on traffic, including the proper definition of gridlock, the concept of memory in traffic flows, and the average number of vehicles actually moving at any given moment in downtown Manhattan (9000).
Four minutes a day: That's the amount of time the average American gives to sex, according to Gleick's source. From other sources he comes across that 4-minute figure again: It's also the amount of time per day the average American spends filling out government forms (his source on this is the government itself). Four minutes a day is the amount of time that using a microwave will save you, so thank your microwave that you have any sex life at all. Four minutes is also how long you'll wait today for Windows to start up and shut down, at least.
We don't experience time the way our ancestors did. When did our relationship with time start to change? Movies and the telegraph and trains had something to do with it: The telegraph made time zones necessary, and destroyed the illusion that noon was noon wherever you were; local time was just that, local. To people who went through that transition, did it seem like a gain or a loss, I wonder? To catch a train, you had to own a watch, and you had to start thinking of minutes and seconds as meaningful units of time. That kind of thinking couldn't help but spill over into other aspects of your life. Photography and movies did something more: they allowed us to chop up time, examine instants. Muybridge's careful time-sequence photography demonstrated that horses don't run the way everyone thought they did, and slo-mo movies let us see apples exploding when struck by bullets, raindrops deforming on striking earth, flowers opening: real-life scenes formerly beyond human experience.
Such alterations of the normal flow of time gave rise to the term "real time," a term that is used more and more these days even as less and less of human experience is lived in real time. The TV remote was originally conceived as a device to let you turn off the set without getting up, maybe to change channels once or twice an evening. Nobody anticipated channel surfing. Networks have responded to it, though, by altering the content of the programs. It is no longer adequate to hook the viewer at the beginning of the program; every second of a show must be hook material, to catch the channel surfer. What channel surfers are doing, typically, is watching more than one show at once. Multitasking. Pop-up Video and Mystery Science Theater 3000 are examples of programming that builds in multitasking, two shows in one. And now at the end of the show the credits roll on one half of the screen while promos for other shows run on the other half: The producers understand that you can -- that you prefer to -- handle multiple streams of information simultaneously. You're probably eating dinner and carrying on a conversation or two, too; and many of you are also on the net. Multitasking: You can't get through the day without hearing someone brag that they are multitasking, but if anything, we do it more than we realize.
"Let's give ourselves credit," Gleick quotes an NYU professor, "We have learned to grasp quickly. We can read signs, change lanes and avoid other vehicles at 70 miles per hour while also listening to a song and planning our weekend..." Our eye is quicker than our grandparents'. Folk wisdom used to say that we use only 10 percent of our mental capacity. Are we using more today? Or is quickness bought at a cost of depth? And whatever the cost, is it worth it?
Gleick gives a pretty hilarious survey of the contradictions of self-help time-saving books, but he makes a serious point: Saving time is an ill-defined concept. If you skip the 4 seconds it takes to fasten your seatbelt, is that 4 seconds you can use to make money at your marginal rate of income? Is it a billable 4 seconds? Or does it just disappear? Try to detail exactly how you spend your time during a day, or during an hour, or the last 10 seconds, and you'll find the task fractally infinite. How many hours did you work today? How long is a coastline?
There are other nuggets in this book: The Strong Law of Small Numbers, the Future Packaging Industry. Gleick has done his research well. Finally, he notes, as Brand does in his book, the seemingly incomprehensible fact of the singularity: The fact that many present-day phenomena, graphed against time, go asymptotic a few years from now. Neither Brand nor Gleick give much help in interpreting this fact. Clearly, the number of commercial Internet hosts, the number of software patents, the MIPS of a desktop computer, and the number of chest-pain emergencies will not go infinite in the year 2004, even though the graphs say they will. But what is the message of the graphs? Does all science or civilization break down at the singularity? Surely not, but what happens?
I found the best answer to that in another source. Scientist Gregory Benford, in his science column in the October/ November Magazine of Fantasy and Science Fiction, dealt with it as clearly as anything I've seen.
"One idea that shall surely not survive this century," Benford says, "is that of the readily foreseeable future." Right; the message of the singularity, of the graphs going asymptotic, is simply this: We are running smack up against the massive failure of the art of predicting the future. All prediction, whether linear or geometric or whatever, occurs within a paradigm. Future progress, I read into Benford's words, will chiefly be across paradigms. If that's true, the pace of change will only increase.
How can we keep up with an evermore-rapidly changing world? By changing, of course. Benford says that just as the 20th century was the century of physics, the 21st will be the century of biology. We ourselves will change, as will our vision of ourselves. "The rate of change of our conception of ourselves will probably speed up from its presently already breakneck pace," Benford predicts, hoping we won't call him on the fact that he just said prediction was impossible. But whether his predictions hold or not, he concludes, "one thing is certain: the ride will be interesting. Hold onto your hat."
DDJ