Why Cellular Automata?     Batting .300

Understanding AMIX

The American Information Exchange (AMIX) is perhaps the most innovative venture in which Autodesk has become involved to date. Without doubt is has been the hardest to explain. Autodesk invested in AMIX in June of 1988, with the intent of implementing and launching the world's first on-line market for information and expertise. Within a year, while the development of the server and access software was still underway, it seemed like almost everybody had forgotten the opportunity that existed and why we had become involved. I wrote this memo as a reminder and to persuade Autodesk to maintain its commitment to AMIX.

To: Al Green, Malcolm Davies, Ron McElhaney, Greg Lutz
From: John Walker
Date: September 7th, 1989
Copies: Eric Lyons, Phil Salin, Kern Sibbald

Subject: Understanding AMIX

I've been peripherally involved in several recent discussions about the development status of AMIX and the soundness of Autodesk's ongoing commitment of resources to AMIX as approved at the end of fiscal year board meeting. Some of these discussions have indicated to me a lack of complete understanding of the fundamentals of the AMIX business, and consequently of the nature of the risks and potential rewards of our investment therein. This note is an attempt to dispel, as best I can, any confusion about AMIX by stating, as clearly as I can, the essentials of the business opportunity it represents.

First, let me make explicit my biases. At the time we decided to fund AMIX I believed it was a sound investment opportunity, more for the public service than for the in-house product. I continue to believe this. For AMIX to succeed, it must run the gauntlet of any new venture; it can fail technologically, in its marketing and sales plans, or just because the time is not right. The considerations that apply here are identical to any efforts we undertake other than self-evident derivatives of current products which are inherently incapable of generating major increments in revenue. AMIX lacks the technological excitement of being a “neat hack” in the sense of Autodesk Animator or The Eagle Project. Instead, I believe it is a Good Business, in the sense of Jiffy Lube or Federal Express; one that may be properly positioned in time and the development of technology to grow and prosper beyond the estimates of its creators.

The Information Age

To call our time “The Information Age” is so obvious it has become a cliché. Our Information Age is the age in which mouthing of the hoary maxim “knowledge is power” is supplanted by the reality that “knowledge is wealth”. The success of Autodesk and other software companies which have built businesses upon intellectual property—the product of the human mind, with minimal financial investment, physical plant and other capital equipment, are testimony of the advent of the information age. My paper “The New Technological Corporation” explained the ways in which I believe that intellectual property and its consequence, technological leverage, have scouted a new course to the summit of business success.

Signs of the emergence of the information age abound for the acute observer. Warren Buffett, the shrewdest investor of our age and foremost believer in “no-tech” investment; the man who proudly proclaims that he still doesn't know what a transistor is, buys a major stake in ABC/Capital Cities because he believes that control of channels of information is capital just as much as ownership of a steel mill. The acrimonious battles over patent and copyright protection of algorithms, software, “look-and-feel”, and chip architectures are in fact nothing more or less than the gropings of a system designed for machines of steel and steam forced to come to terms with silicon and software.

As intellectual property supplants physical property, the ability to obtain access to accurate, timely information has become one of the keys to success. Our world is increasingly webbed by many different means of delivering information: some old, some new. In a given week I usually partake of the following:

Newspapers (Wall Street Journal, New York Times)
Magazines (Aviation Leak, Electronics)
Scientific Journals (Nature, Science, IEEE Computer, …)
Newsletters (HSL, Dow Theory Letters, Foresight, Commercial Space, …)
Television (McNeil Lehrer, Nightline)
Radio (BBC World Service)
Computer Broadcast   (USENET NetNews: comp.graphics, comp…)
Computer Online (CompuServe CIS, Dow Jones News/Retrieval)
Computer Mail (Technet, USENET, Internet)

All of these give me information. Information keeps me from being a dull boy, points out things worthy of further study and the like, but seldom addresses my chief need when I have a goal at hand and a pressing deadline to achieve it. I do not need information, I need answers.


We live in what is supposed to be the exemplar of a “free market economy” yet virtually nobody, even many who have achieved great success in business, understand the basics of the dynamics of a market. Markets cannot be easily analysed. They are simply places where the collective wisdom of the buyers and sellers of the moment duke it out and somehow, in a chaotic yet historically effective process, arrive at prices and hence values for the goods they exchange.

Markets are so messy and confusing that it takes an intellectual leap to propose a market as the solution for anything. Typically a dealer and/or retail distribution system will predate a market. For example, stock options were available in a dealer market for decades before the emergence of exchange-traded stock options in the 1970s. Although futures markets for grain, cotton, and sugar existed in the 19th century, livestock futures were considered a radical innovation in the 1950s and 1960s, as were financial futures in the 1970s and index futures in the 1980s. Each of these developments created an efficient market in which prices could be assigned to commodities traded in far less visible ways before.

Markets have evolved, then, from concrete commodities such as a carload of wheat, to abstractions such as a 100 shares of Consolidated Diversified Industries, to concepts such as December XYZ 40 Calls, to financial vapourware such as options on futures on international shipping rates.

And yet, we have no markets for information.

Information Markets

AMIX is nothing more and nothing less than an attempt to create the first efficient market for information. If its current prospectus and goals seem modest, I suggest this is not a consequence of the potential of the venture but rather of the belief on the part of its principals that the concept is so robust that it can establish itself as a viable, self-sustaining business even with a very limited initial scope.

The information age has endowed information—knowledge, with the value previously reserved for physical property. Yet, as usual, our institutions—the software of civilisation—have lagged in evolving an efficient market in which this new commodity can be exchanged and hence assigned a value. AMIX is such a market. It was conceived and designed by people with a deep understanding of the role of markets in an economy and of their functioning in valuing abstract commodities.

To the person seeking an answer, existing online services provide plenty of information but few answers. To a potential provider of answers, there is no market to which he can bring his knowledge to offer to those in need of his talents. In economic terms, the market for information and answers is inefficient; it has high transaction costs for both buyer and seller and fails to provide the valuation feedback to both which is essential to make informed decisions. AMIX is an efficient computer-mediated market for matching buyers and sellers of information. The utility AMIX provides to the Information Age is as self-evident as the function of currency futures in the post-Bretton Woods era, yet equally difficult to invent and understand before the concept becomes so commonplace it's considered trivial and obvious.

Obvious Markets

In May of 1968, I visited the Battelle Institute in Columbus, Ohio and first heard a description of the “Xerox Necessity”. What was the size of the market for perfect copies of black and white documents, produced on plain paper in less than 60 seconds, the year before anybody realised such a thing was possible? Effectively zero, of course. What was the size of that market 5 years after the appearance of the Xerox 914? Look it up.

What was the size of the market for overnight small package and document delivery in all the years before Federal Express figured out how to make such a service work, and gained the confidence of their customer base that the letter would actually be there the next day? One can imagine the entrepreneur who conceived of Federal Express fielding questions such as, “How is this different from First Class Mail?”, and “Even if you succeed, why won't the Post Office jump in and put you out of business?”. And yet….

The rationale for our investment in AMIX is the belief that AMIX is the first well-conceived efficient market of the information age. Those who have created markets for the predominant commodities of the past have done very well, indeed. If this is, in fact, the information age, then eventually there will emerge an efficient market for information; one that supplants the retail or dealer market of existing information utilities with the give and take of an exchange floor and thereby assigns correct values to information, reduces the price of information to the purchaser, and opens the market to many more sellers, to the benefit of all and to the profit of the market maker.

Non-critical Issues

It seems to me that our concerns and evaluations of AMIX have often focused on issues that appear peripheral. The implementation of AMIX has encountered its share of the unforeseen technical delays that befall any engineering project. The delays which have ensued, and the way in which the problems have been addressed, are on the order of our experience in such projects as AutoCAD 2.6, Abbey Road, and AutoCAD/Macintosh. Let's not confuse the inevitable technical difficulties in realising a complex project with our belief or lack of belief in the value of the goal we're working toward. In all of the AutoCAD projects I named, at no point would we have considered a schedule slippage as grounds for cancelling the project as long as we believed in the product we were developing. So it should be with AMIX. If we understand and believe in the goal, then manageable schedule slippages should be accommodated. If we believe the goal to be unachievable or illusory, then we should expend no further effort or money in its pursuit, regardless of the state of its realisation.

I do not intend by any of this to demean the demonstrated competence and manifest dedication of Chip Morningstar, Randy Farmer, and the other members of the AMIX development team. They have undertaken a difficult task and are well on the way to completing it. I think, though, they would be the first to agree that while failure to complete the software development tasks they're doing can spell doom for AMIX, success can only grant AMIX a chance to succeed. I am confident that the AMIX development team will complete a product within an acceptable time and budget, albeit having encountered slippages on the order of those experienced by our other development projects, adequate to grant the AMIX business concept a fair trial in the marketplace. Consequently I'm not particularly concerned with the details of AMIX software development.


To invest in AMIX is to believe that we truly live in the information age; to accept that in our age information is a commodity as tangible as wheat, live hogs, Swiss francs, or the S&P500 index; that a self-regulated market in which information can be exchanged will inevitably emerge and prosper just as markets have developed throughout history for whatever was valued in that era; and that embodied in AMIX is the basis for the first true efficient market of the information age.

I believe in all of these things. I think AMIX was a sound investment when we funded it, and I believe it merits continuing our investment as planned.

I hope we will see AMIX through to the ultimate verdict of the marketplace, the information market that AMIX hopes to create.[Footnote]

Why Cellular Automata?     Batting .300