The Frontier Between Us

Jaron Lanier

The easiest predictions to make about the next fifty years of computer science are those centering on the computers, ignoring the people. For example, I can predict with a fair degree of confidence that there will be fabulous increases in hardware capability, which will be largely consumed by a corresponding decrease in software elegance as decades of legacy systems tangle like parasitic vines. It's when people are brought into the equation, however, that the business of prediction becomes difficult.

Who would have thought that not only computers, but computer science would become a centerpiece of pop-culture? Who imagined that one would be able to buy magazines with articles on caches and dithering at the supermarket checkout counter? Or that UNIX path punctuation would become a vernacular element of advertising (in the form of web addresses)?

The biggest surprise from the first fifty years of computers is that computation turns out to be a cultural object in its own right, with all its warts and semicolons. Many a visionary had imagined that computers and networks would have a transformative effect on culture, but it was usually assumed that the nasty details would become invisible as their influence increased. It is still part of the marketing orthodoxy of the computer industry that hardware and software must in some misty future become "consumer products", as unobtrusive as toasters. And yet it turns out people love to obsess about the insides of their computers. Children who build elaborate web pages in HTML and Java routinely burn toast.

The public has often warmed to the surface of science and engineering, but never before to the depth. While there are tens of millions of people who love dinosaurs and black holes, how many of them have gone on a dig or analyzed spectrum data? When it comes to computers, though, a mass culture of technical literacy is being born, especially among children. We always thought computers had to become popularized, and instead the public has decided to become surprisingly technical.

This is due in part to the stalwart marketing of awkward software by Microsoft, and in part to the economic pressures favoring open systems, which will always have rougher edges. But those cannot be the only reasons. There is an emotional draw. Maybe it is the ability to control a microworld that is more predictable and less filled with pain and ambiguity than real life. An abstract aquarium, a theater of numbers.

Whatever the reason, I would want to celebrate the public's embrace of computer arcana, except for one thing. The material itself is unrelentingly ugly. I want to cry when I see those toast burning kids endlessly tweaking HTML source code. This is the kind of soul numbing tedium that I once believed would be forever banished by the end of the nineteen-eighties. In those ancient days I thought that by now there would be widespread adoption of brilliant visual programming tools.

Computer science is, alas, the only engine of culture that has not concerned itself with beauty. Why should we have? We didn't know we were making culture. We thought we were making invisible tools. We've been granted a surprise franchise as culture creators. In the next fifty years we have an opportunity, and a responsibility, to contribute in ways we never anticipated.

Our art is abstract, but has a profound emotional and social effect on our audience. There are already masterpieces. TCP/IP, along with the related code that runs the Internet, is perhaps the most dramatic. It is beautiful, and it embodies a profound conception of openness, and therefore of faith. It is rare indeed that an unsuspected and positive attribute of human nature is exposed for the first time through a work of art, but that is exactly what happened in response to TCP/IP. The internet is such fertile earth that it practically commanded the blossoming of exquisite new organisms like the World Wide Web. Never before did we know that millions of people could cooperate almost instantaneously to build something (the Web) merely because they wanted to, with no planning, lines of authority, advertising, or finance. It turns out that in the right conditions, people are good enough for anarchy.

Unfortunately, TCP/IP is the exception. There are far more numerous examples of ugliness, such as MS DOS. A hard fact of life is that ugliness in software is worse than ugliness in other art forms because it is less perishable. Layers of software become locked in place when new layers refer to them, and ugliness from lower layers percolates upwards. So we'll be stuck with MS DOS for many, many years, and it will reduce the beauty of all the software created on top of it.

How do we make beautiful software? General engineering principles, like openness, are good enough to create elegance, but not beauty. Beauty requires an awareness of human affairs outside the computer. When considering the relationship of people and computers, we're sometimes subject to a figure/ground illusion. We can easily see the computer as the center and the person as the peripheral. This illusion is encouraged by the public obsession with computers. It reaches its extreme in the Artificial Intelligence approach. Such disconnects from reality are, I believe, among of the primary ways that computer ugliness comes into being. When software design decisions aren't made in reference to human concerns, they can only be made in reference to each other, leading to a self-referential bundle of nonsense suspended by a sky hook. The simple way to notice the illusion is to point out that computers don't function independently of people. They are cultural artifacts, like language, intelligible only to those who know them. To a Martian, a computer and a toaster are the same.

When we treat information systems as no more than conduits between human imaginations, grand vistas open up. The pleasant news computer scientists can infer from the public's early embrace of computer tinkering is that we will not be serving a population of consumers, but rather of creators. In the next fifty years, computer science will give birth to a delightful new vernacular art form that combines the three great art forms of the twentieth century; cinema, jazz, and programming. The result will be a mass theater of spontaneous shared imagination and dreaming. My fond hope is that it will take the form of networked VR with inspirational authoring tools that are capable of quick, improvisatory creation. But whatever the specific form, what we are building will encourage people to share interior vision and treat it as a tangible, worthy thing, even into adulthood.

This is the frontier that information science opens up to mankind. There are other frontiers enabled by science, of course; the exploration of space, the study of the brain. But only ours will continue to reveal unsuspected potential in the most precious of natural phenomena, relationships between human beings.

Go back to Jaron's home page.