Valedictory

The Autodesk annual shareholders' meeting on June 10th, 1988 marked the point where I relinquished the title of chairman (or as I usually put it, always afraid of offending, “chairbeing”). Since I had held that title since the inception of the company, and since my transition to total focus on the technological future of the company was now complete, the meeting seemed an apt time to sum up the first six years and comment on the perceived opportunities which compelled me to concentrate on software development as my best contribution to Autodesk's future.

Since this will be my last meeting as chairman, I'd like to take a few minutes to share with you my view of where the company is, how we got there, where we're going, and what I'm going to be doing in the future to see that we arrive safely.

When I was thinking about starting Autodesk in 1982, the local paper ran an article about software companies in Marin County. It said that the opportunity to start new ones had closed because “all the basic tools had been put on the computer”.

This didn't make any sense to me. It seemed obvious that the new wave of mass-produced and mass-marketed personal computers would create ample opportunities for new products to perform tasks made possible, for the first time, by those new machines. In fact, the small group of people who founded Autodesk identified five such products, and we decided to work on all of them and see which one took off. One of our five products was AutoCAD—and its success in creating a new market for a new tool on the personal computer is what has brought us all here today.

Over the six-year history of Autodesk, technological progress in delivering affordable computer power to the mass market has, if anything, accelerated. The products we're developing today can expect, in the middle of their lifecycles, to run on computers 100 times faster than the computer that ran the first copy of AutoCAD—yet they'll cost no more than that first system. By the time Autodesk celebrates its first decade, we can expect to see affordable computers ten times faster than that entering the market, and there is no reason to believe it will stop there, either.

All of this computer power is pointless unless we can think of something to do with it. Who cares if you can recalculate your spreadsheet in 10 microseconds instead of 10 milliseconds? Fortunately, there is no shortage of tasks to which this kind of computational power can be applied—and Autodesk is superbly positioned to take advantage of this opportunity, which will create many entirely new product categories and markets, just as the IBM PC generation enabled us to create AutoCAD.

Let's think for a moment about what it is that Autodesk does—at the highest level. All of our products are basically in the business of putting models of real-world objects into a computer, and then letting you do things with them. This is a fundamentally different, and more interesting, business than twiddling numbers on a spreadsheet, shuffling text in a word processor, or whatever boring things a database does. It's different because every increment of computer power, graphics performance, large low-cost storage, or new computer architecture lets us get closer to a complete model of a real world system. As we continue to approach the ideal of a complete model, we're able to do more things with the computer model, and these things turn into products we can sell, to users who already understand the need for those new tools.

AutoCAD started out as a simple two dimensional drafting system. When we got computers with more memory and faster processors like the 80286, we were able to make AutoCAD programmable. This allowed us, and hundreds of third parties, to create applications that customise AutoCAD for vertical markets. Our own AutoCAD AEC Architectural has become the leading architectural design product in the United States, and we expect our recently-introduced AutoCAD AEC Mechanical to similarly dominate the heating, ventilation, and air conditioning market.

As AutoCAD was extended into a three-dimensional modeling system, we were able to use those models to generate realistic shaded pictures. The advent of low-cost, high-resolution, colour displays enabled us to introduce AutoShade, which is placing shaded rendering in the hands of thousands of customers who previously thought it far out of their reach. And recently, we've shipped AutoFlix, which adds animation to the list of things you can do with that model inside your computer.

And still our models continue to become more accurate. AutoCAD Release 10, announced at the recent AutoCAD Expo trade show, dramatically increases AutoCAD's three-dimensional capabilities, adding surface modeling and multiple views. AutoSolid, now shipping, delivers true solid modeling to the mechanical designer. From an AutoSolid model, you can calculate mass properties, perform finite element analysis, and begin to forge the direct links from design to manufacturing which are so essential to increasing manufacturing productivity.

We also recognised that many problems don't require a tool as powerful and complicated as AutoCAD, so we introduced AutoSketch for less than $100. It can put computer aided design in the hands of anybody who draws. And we've sold more than 60,000 of them.

But you know all that…what about the future?

I believe that technological progress now underway is creating opportunities, as large or larger than the 1982 prospects for AutoCAD. Further, I believe that Autodesk is uniquely positioned to exploit these opportunities and thereby to prosper.

One of the great things about the CAD business is that you never run out of a desire for more powerful computers and more comprehensive modeling software. Extending the degree to which our products approximate reality, making models easier to build and manipulate, and enriching the ways in which one can interact with these models can easily consume all of the computer power we can envision, as far as we can foresee. This means that adding capabilities to all of our current products, adapting them to take advantage of developments in computer hardware, and extending our product line by adding additional modeling, rendering, and analysis tools will keep Autodesk's product family growing in size, and will open new markets and application areas for our products.

In addition, more powerful computers and networks will allow our customers to integrate the design, drafting, analysis, and manufacturing functions much more closely. Autodesk is the only software vendor whose products maintain complete compatibility across all computer hardware, and our commitment to open architecture makes our products the obvious choice for this integration. We are developing our products so they work equally well for the customer with one isolated desktop computer, and the customer with 10,000 workstations distributed around the world. Both are key to our success; both are central to our strategy; both deserve the best products we can design.

Most industries are founded on a single, simple idea, such as “cheap cars for everybody” or “machines to automate business”. Autodesk remains the only software company committed to “putting the real world inside the computer”, and we believe this idea has potential as great as those that spawned the great industries of the past.

There's a lot more to the real world than houses, turbine blades, and circuit boards.

The real world is also a world of words and ideas, images and information. When we recently acquired 80% of Xanadu Operating Company, we undertook the task of putting that world into the computer, as well. The dream of hypertext, and the ability to store, organise, retrieve, annotate, and present information in ways that do not mimic paper, but are fundamentally better, has inspired a generation of thinkers. We believe that Xanadu can benefit anyone who reads, writes, and thinks, and is applicable at scales ranging from an individual's personal computer to a global library storing all of human knowledge. We expect to ship the first commercial Xanadu system in less than 18 months. And Theodor Nelson, who invented Xanadu and coined the word “hypertext”, has joined Autodesk to help us do it.

Ideas are precious and rare. They should be immortal. Most ideas are lost, most discoveries ignored, most interrelationships unremarked upon because we have no effective way to store and then find them. Galileo observed Neptune, but nobody noticed until the 1980's. Most ideas are committed to bits of paper, obscure professional journals, or computer media made obsolete by the next wave of innovation and are lost forever. We can't afford to lose 90% of the products of our collective minds. Xanadu is intended to fix this, and we believe that in time it will become as universal a product for people who think as television is for people who don't. And if this happens, we will sell lots of them.

Autodesk's current products as well as Xanadu are converging to address another large opportunity—building tools that help people work together more effectively. We talk about personal computers. But how many people, even free-lancers, work in total isolation? The computer will come to play a role as fundamental in communication and collaboration among people as in facilitating the work of an individual. To succeed, it must enhance individual creativity, not stifle it.

Too many computer systems seem like one-lane roads where all must proceed at the pace of the slowest. We need instead to design freeways for the mind, where people can work together, each proceeding at his own pace, without impeding or imperiling others. Products that do this effectively will yield rapid and dramatic productivity gains, and create new product categories in the software industry.

As our computer models become richer, and the ways we can manipulate them increase, creativity in devising how we interact with these models will be rewarded by making them accessible to many more people, broadening the market for our products. The AutoCAD customer who, in 1982, squinted at a low-resolution screen while typing in commands from a keyboard, can today create animated colour movies of three dimensional models built largely by pointing to a screen. I believe that progress in user interfaces has only begun.

While the faltering leaders of the last technological revolution are suing each other over how their screens look, the pioneers of the next are developing tools that will take the user through the screen, and allow direct interaction with the data in the computer, whether as concrete as a connecting rod, or as abstract as the history of revisions to a document. Using a computer involves suspension of disbelief just as much as reading a novel. The ways people will use the tools and models we create will continue to evolve, just as styles in writing and filmmaking change with the times. Innovation spawns progress; litigation impedes it.

So what does the future look like for Autodesk? We're a software company. We build and sell tools. As long as there is demand for new tools, there will always be opportunities for great successes with software products. Anybody who thinks that all possible software tools have been invented is unimaginative, uninformed, or just stupid—man has been called the animal that creates tools—we've been at it for millions of years, and I don't think we're going to tire of it this quarter, this year, or this millennium. Autodesk has prospered so far from a tool that solves a fundamental problem—drawing and design. Autodesk will prosper in the future by continuing to develop that tool, and by creating and selling new ones that solve other problems as profound and fundamental.

The newspaper's claim in 1982 that “all the tools had been done” was actually profound as well as silly. The software that can be done at any point in time depends upon the computer power then available in the mass market. Computer power is now rising exponentially at constant price, and Autodesk is one of the few companies working seriously on products to take advantage of it.

Developing new products is difficult. I know—we've done 8 since 1982, but if we apply the same humility we had during the evolution of AutoCAD, I believe we'll continue to develop products which solve new problems and create new markets for the personal computer. In 1983 people used to ask us about “our vision of the future of CAD”. We used to say in all candour, “we don't really know enough about CAD to have one”. But we had a secret. We knew somebody who did know—our customers. No vendor in any market knows as much about his product, how it is being applied, where its shortcomings are, and what is needed to make it better, as the customers using it. And today, we have more CAD customers than any other company in the business, and they continue to guide the development of AutoCAD and all our other products. Hubris, borne of detachment from the real problems facing customers, has doomed many technology companies. That's not going to happen at Autodesk, because we still don't know enough to chart the future of anything. But we listen patiently, and we take lots of notes.

So we don't know what the future will bring, but we do think we know how to get there. This company was founded on the idea of developing 5 products and getting behind whichever one took off. I like to be able to be wrong 80% of the time and still do well. It worked, and we will continue to follow that evolutionary approach to the marketplace as we have for six years. Many things we try may not work, but we expect those that do to form the foundation of our future prosperity. This isn't all that unusual an approach—every investor with a diversified portfolio is doing the same thing. I think that companies don't like to admit they're being guided by the market because you get credit for being a brilliant planner, not for being a careful observer. Evolution looks like planning when you look back in time—that's why so many people have trouble understanding it. In fact, market-driven pragmatism not only works; it pays well.

Getting there won't be easy. But getting here wasn't easy. We've seen competitive products from IBM, CalComp, AT&T, and Computervision come and go. We've lived through the worst stock market crash in history. We've seen overt theft of our products, and we've discovered how hard it is to stop it. We've made misjudgements of the market and we've corrected our course. Sometimes we've been wrong, but we never stayed wrong too long. I'm not going to stand here and tell you that Autodesk will never have a bad quarter or never screw up. But I do believe that as long as Autodesk remains focused on the enormous opportunities we face, as long as we act prudently and responsibly to take advantage of them, what will be called Autodesk's period of great growth and success lies in the future, not in the past six years. And it is a measure of my confidence that the management and directors of Autodesk will achieve this that I have no hesitation whatsoever in entrusting the operation of our company to them.

Many of the things we need to do are obvious. But just because they're obvious doesn't mean they're easy. It is because I believe that the technical work ahead of us is so great, so important, and the benefits to Autodesk are so enormous, that I will be focusing all of my energies on these tasks, swearing off meetings with lawyers, giving speeches, signing checks, filling out forms, talking to analysts, and all of the other fun things that I got to do while chairman.

For whatever reasons, we've all wound up owning a company that's right in the middle of an exponentially growing technological revolution that's remaking our world. Many companies share this position with us. But we know where we are—and we think we know what to do about it.

Many companies have abandoned an optimistic view of the future. Starting a company out of thin air at the bottom of a recession with almost no money and seeing it grow into this in six years, largely as a result of simply plugging away day and night on the mundane, near-term tasks that had to be done, is a wonderful antidote for pessimism. It gives you great confidence in continuing to reach for further success and achievement that same way—getting all the details right, but never losing sight of where success can carry us. Without imagination we are all doomed to live in a grey world devoid of hope and excitement, fortune and glory. We talk about the Golden Age of Engineering. It's our age—if we work to build it. To achieve it, we must toil tirelessly at the gory details. If we invest that effort, our achievements will be unbounded, our horizons endless, and the potential for Autodesk incalculable.

Thank you for the confidence and trust you've placed in me the past six years.

The ongoing development of AutoCAD is extending it into a general three-dimensional surface modeler. AutoCAD Release 10 adds polygon meshes, providing a general surface object. This allows new applications, such as modeling the surface defined by functions of two variables. Here's an interference pattern generated by two exponentially damped cosine waves. I made this with a little AutoLisp program to test huge polygon meshes with interesting cases.