Copyright 1998 CAUSE. From CAUSE/EFFECT Volume 20, Number 4, Winter 1997-98, pp. 15-20. Permission to copy or disseminate all or part of this material is granted provided that the copies are not made or distributed for commercial advantage, the CAUSE copyright and its date appear, and notice is given that copying is by permission of CAUSE, the association for managing and using information resources in higher education. To disseminate otherwise, or to republish, requires written permission. For further information, contact Jim Roche at CAUSE, 4840 Pearl East Circle, Suite 302E, Boulder, CO 80301 USA; 303-939-0308; e-mail: [email protected]

Technology and Change: An Interview with John Gage

John Gage, Director of the Science Office at Sun Microsystems, was the opening session speaker at CAUSE97. After his presentation, he and Ken Klingenstein, Director of Information Technology Systems at the University of Colorado at Boulder, sat and talked about a number of technology-related issues. Following are excerpts of that conversation.

A sample of the actual interview is available in RealVideo (28.8k)

To take advantage of RealVideo you will need to obtain RealPlayer 5.0 from the RealMedia Website. See the CAUSE file help page for details on configuring your browser to view these files.

Klingenstein: I�ve come to believe that managing complexity is going to be the biggest task at the turn of the millennium, in terms of technology and our personal lives. Do you see any way to manage complexity?

Gage: The tools we�ve always used, personal relationships with people you respect, are going to be what we rely on. But managing, where you�re trying to understand where things are going, is different than judging what�s amusing or personally beneficial.

In the management world, complexity is a complexity of people, and it�s becoming far more complicated. If you look at the structure Sun has built as an industrial company, we have people who came from Taiwan, or from the Indian Institute of Technology in Delhi. They come to California, they make friends, they leave. They become part of an American company for awhile. The Taiwanese pass through two or three years in Silicon Valley, meet everybody, then go back to Taiwan and start companies there, using these friendships and links.

Suddenly, we�ve been relying upon friends and co-workers. Now it�s five or ten years later, and we�re relying on this web of people for parts, technical predictions about what they�re going to do. We�re telling them what we need, but they�re in different time zones, speak different languages, and have different reliance on governmental policies. That kind of complexity, we�re beginning to layer, as we break up this centralized notion of who�s in charge. It�s a bit like the original Internet in the academic days when one smart graduate student would post code and someone in New South Wales would fix it in an hour and send back a fix. It�s a distributed global environment. The tools we have to make this work must be the same tools that, in theory, we in education develop. We need the ability to tell the good from the bad. Good taste is what is required to manage complexity. Without good taste, the cacophony will rule.

Klingenstein: We don�t have a course in good taste or critical thinking anymore.

Gage: That�s a shame. When I look at how we manage, 90 percent of what�s presented in management meetings is either wrong or not thought through. The way the company works globally, you have to instill in people some way of judging what each action does in relation to the mission of the overall organization. If you have no coherent mission, then there�s no coherent behavior. People just go off and do things.

How do you pay people in a way that gets them to do certain things that are profitable overall for the company? This is art. I used to think this was boring. Turns out it�s not boring. It�s motivating people. The Sun ethos of open access is more believed in the remote ramparts of the Roman empire than it is in Rome. In Rome, you have people who came from DEC and want to charge you for everything. In China and Ireland are the true believers in the Sun way of life, and they�re saying it�s not proprietary. It�s open. You like the source code? -- here�s the source code. That overall sense of direction is the one way you can use as a touchstone. Do I raise the price to the customer, or do I decrease the price and help them so it seeds something else? Little distinctions can change overall behavior enormously.

Klingenstein: How do you create openness out of proprietary attitudes? How does one take a Microsoft and reintroduce competition?

Gage: The pathway we�ve embarked on relies on breaking up all the large monolithic pieces of software into smaller objects. Recently I saw a Web page of a series of calendar JavaBeans that this company is now offering for sale. You incorporate them into your application. They�ve done a very sophisticated job of keeping track of dates on many different calendar systems. You don�t have to learn any of that. You just take their Bean and incorporate it into your application and pay them a little bit.

As we build an object economy, which goes down through the operating system, a scheduler could be an object you could use. You don�t need to use Sun�s. We�d begin to build new markets in components. That�s the pathway that will allow us to break up the existing software monopolies. Because the software provides functions that are multifaceted, you can break them up into things that are useful. For example, I�d love to be able to take the full-text indexing systems that exist, and index by meaning the natural language that allows you to break words up. That should run in conjunction with the word processor. As I�m writing something, in the background, the words I�m writing are being broken out, morphologically the endings are taken off and it�s compared. Now I get a deeper kind of help in writing in a literate way. It would point out a certain word isn�t really in a right spot. I can�t do that today in any easy way on a word processor.

If they�re defined properly, the interfaces allow add-on capability. Java is the particular vehicle for this. We�ll bring out something that allows you to put a layer into your existing system that deals with all services provided to your system, whether disk storage services or networking services. They all will come to you in this distributed system. Software talks to them. It slowly eats away the functionality that�s now embedded in a large operating system.

Klingenstein: The politics of technology is very hard. I�m wondering if the Justice Department has the wherewithal to understand this and go in carefully with scissors and snip apart the monopolies?

Gage: I�m a technological determinist in the sense that I think that if we have the framework for the breakup of the existing monolithic software into objects, then there�s an economic and technical inevitability. We�ll want to be sure that the objects we compose to make a particular application are from people we trust, that have some kind of a liability link, some way we believe that this will all work. That structure will be a survivable commercial structure.

Today, you buy from Microsoft because you�ve heard of them, they do a pretty good job. If you�re confronted with someone with a similar piece of software and you�ve never heard of them, you�re much less likely to buy from them.

Klingenstein: It�s an interesting premise that the fruit will fall from the tree and we don�t have to cut it.

Gage: That�s our hope; it may be naïve. On the politics of technology, the poor lawyers only have blunt scissors. There�s no mechanism they have to get inside this world. In some sense, we�ve just got to do it ourselves.

Klingenstein: I�m curious about how we�re going to preserve interoperability against proprietary standards. It feels like the standards processes are breaking down.

Gage: It turns out it�s really quite painful. In things you can define clearly, it�s not changing. To pick twenty-three core objects that must be shipped, that seems pretty clear but is becoming more fuzzy. Each of the objects could have inside it some changes, which is what Microsoft did. They altered some of the methods to do things, which just don�t work. You expect the object to behave one way and it doesn�t behave that way. Then they eliminated a couple of objects from what they shipped. The remote method indication object, which is one they left out of Internet Explorer 4, you have to get into your own machine and embed it in the class hierarchy. This is definitely subverting the ad hoc standards process.

The parallel track is to put the standards under International Organization for Standardization (ISO) control with existing mechanisms for countries and companies to participate. That was redone when ISO recognized that eight years to standardize C++ was a crime. ISO, to develop standards in a rapidly changing world, must alter its procedures.

How do you do this? We plan to let anybody submit a standard and let everybody vote and we�ll argue about maintenance of the standard. However, we want the cycle time to be a year or six months, not eight years. Sun will play. OMG [Object Management Group, a consortium of software vendors and end users] said they�d play. A variety of different groups went into this procedure and submitted the work they�d already done as a standard and found it accepted by an international standards body. We want to figure out a way to maintain it. That�s the language we�re trying to work out now. We�re all in agreement, pretty much, about the directions, with the possible exception of the part of Microsoft that�s trying to derail the whole thing. It seems we can move together collegially. The benefit to people will be of such a magnitude that there�s no reason to fight it.

Since we beat Microsoft thoroughly, globally, on this first-round vote, they�ve learned now that people really do want to have something that is not under any particular company�s control, including Sun�s.

Klingenstein: This ISO process is kind of the equivalent of the Request for Comments (RFC) used in the development of the Internet. It may be that the RFC process, this meta discovery mechanism, is the enduring legacy of the Internet.

Gage: That�s exactly right. The brilliance of the RFC mechanism, the populism of it, the openness of it, is the model for all of this. When we went to ISO, we said, �You must modify how you behave. When we followed the pattern of all requests for information the way the Internet worked, when we did that with Java and put code up, anyone could comment. What we discovered in that process was the student at Upsala had a comment that was more valuable than the comment from Microsoft, or Intel. We need you, ISO, to alter your procedures to allow comments from anyone, not just a country committee.� ISO blinked twice and said, �We�ll do it.� We�ve altered procedures for establishing global standards by allowing any person to comment.

Klingenstein: When they talked about creating a footprint in the floor of a telephone company, it was a visible thing they could understand. How do you create a footprint inside an operating system?

Gage: Sun Vice President Ivan Sutherland has a wonderful idea. What we should do is eliminate all restrictions on wiretap, so it�s completely legal. You can wiretap your wife, your wife can wiretap you, you can wiretap your neighbor. However, also make it perfectly legal to have completely powerful encryption to protect yourself. That will create a market overnight. You get a little device you plug in to your phone. Now you�re secure.

We�re in this current plunge toward massively accessible�it�s just a boon to law enforcement beyond anything they ever dreamed. Law enforcement doesn�t know how much power they have. My cell phone, I turn it on, you know where I am. We�re locating ourselves, every call we make. This phone stores in it my last ten calls and how long they were. I call somebody or they call me, their phone number pops up on the screen. There are a lot of instances where you don�t want somebody to have your phone number, but you�re not going to go through the mechanism of having the call cut off at the phone company. We�re weaving a web of complete disclosure.

Privacy and knowledge of your environment, the boundaries of civilized behavior, are being eroded. It used to be possible for the two of us to talk to each other, with complete certainty that no one else could hear. The walk in the woods is your only alternative now. However, with parabolic microphones or with any number of devices, you can�t even do a walk in the woods anymore. Your location, every interaction you have with this expanding web of electronic commerce, every interaction you have with any component of technology, is now being marked. I saw a number recently that in New York City a typical working resident has his or her picture taken twenty times a day. It can happen at a street intersection, when entering an office building, in an elevator, or through a fire alarm system. These video surveillance systems are everywhere for reasons of security.

The major move forward in image processing capability means your face could be recognized fleetingly at the bank, the ATM, or any number of places during the day. Every time you do a transaction they�re going to take your picture, or scan your iris.

This web of information can be used for good or for evil. Law enforcement loves it because they can catch the bad guys, but the fundamental, ideological conviction of law enforcement is that everybody is a bad guy, or a potential bad guy, so they want to know. It leads to these extraordinarily restrictive environments, as biological, genetic analysis becomes more powerful. It leads to a breakdown of society�s absorbing responsibility, spreading the risk. I can now identify people who have genetically very advanced risk. If I can identify them, why should the rest of us pay for those people? You lead into some very difficult ethical and political environments. Knowing too much about each other can be a very serious danger to being able to hold a normal conversation. There�s a reason why our thoughts are not transferred to everyone else.

Klingenstein: There�s a story about Enrico Fermi walking the grounds of the Institute for Advanced Studies at Princeton a couple of years after the bomb. He saw a turtle crossing his path in the woods, and picked it up to take home to his children. Fermi walked a few feet down the path, stopped, turned around, left the turtle where he found it and explained to a friend, �I think I�ve done enough to disturb the universe for one man�s lifetime.� What turtles would you put back?

Gage: The turtles are out of the bag. We�ve unleashed this capability. The American ethos, the engineering ethos, were pragmatic: Here�s a new tool, use it. That is carrying us in a path towards, I don�t want to call it invasions of privacy, but enormous accumulations of information about things, which can destroy completely the universe, the balance, the harmony that�s hard enough to create.

Two people attempting to get along is a serious problem. Now with information about everyone easily accessible, it lets the fringe components of our society have leverage and power over institutions and people in ways that have never been possible before. In some sense, you want to cut the power and pull the plug.

There�s a second side to it that is good. In attempting to understand the power of the industrial mechanisms we�ve built, we never took into account emission of chlorine atoms, or any number of things that resulted in ozone holes or the global carbon monoxide and dioxide. Alterations will kill you fast or kill you later. These off-the-balance-book results of people multiplying and industrial growth, if made visible, can cause behavior that could counterbalance them. In that sense, knowing more about the processes of environments we create and are involved in can be very helpful.

Given the existing technologies of powerful encryption, I believe we can regain the control over information about us, restrict the amount that is usable to others, and focus on the good side of all this information, which might allow us to truly account for all of our activities in a healthy way. Every time there�s a polluter, there�s a polluter who�s not paying the cost of the pollution. It�s the others that bear the brunt. It�s these off-the-book expenditures which are supported by the society as a whole. Now we can get a grip on some of these things and perhaps manage them better.

Klingenstein: It�s pretty clear that we�re a technology-driven society. How do you get away from this personally? How do you escape?

Gage: I have simple routines that remove me from this. I find that it gets to a point -- and the point is arriving more and more frequently -- where I simply push away all this. The 200th e-mail in a day. My fiftieth voice-mail message. This constant sense of equal priority for everything. It forces you to distinguish things. I just turn it off. I don�t listen to my voice mail. I don�t read my e-mail. I go to the coffee shop near my house and I read something. I�ll go to a lecture or a concert.

It�s very different ideas and rhythms, dance and harmonies, different language and vocabularies, that let me suddenly think of something new. It�s odd juxtapositions and richness of culture that lead me to think of something new. The environment of e-mail and quotidian messaging of day-to-day life, it�s all the same. I don�t get that richness and mixture.

I�ve begun to learn that it�s vital to mix in your life the ability to go out and confront the unexpected by breaking out of all the established patterns and being a human being, trying to see how other human beings live. That is the renewing experience. Often I�ll go to a lecture and I�ll listen to the words but I use them as a counterpoint to stimulate something that I�ve been thinking about. C. Wright Mills did a wonderful thing. He would write, think, ponder some question. He�d work out some theory. Yearly, he�d take all his files and writings, everything filed chronologically or by subject, and dump them on the floor. All the clippings; he�d dump them on the floor. For a week, he�d get on the floor and pick up a piece of paper. The juxtaposition, completely serendipitous, would cause him to think of something new. That was almost mechanical, but very powerful in source of inspiration. It�s like poetry, where an odd juxtaposition of words has an incredible power to invoke new thought.

I spent a lot of time with contractors. I discovered they have eyes I don�t have. They can look at a wall, a room, a building and see exactly what�s behind the wall. They know what is needed. They know what they�ll find in what to me is a perfectly impenetrable surface. They see through it. I began to see as they do. Then I began to watch still photographers and video photographers. They�re very different. The people taking color pictures will move to a certain place to take pictures. Other people will move to a different place, the ones with the black and white motorized cameras. You begin to watch their positioning. I�m not there. Even if I were physically there, I wouldn�t be in the same place. They have an amazing fusion of seeing things and this ability to get these pictures. I began to appreciate those plays where some character will describe what happened and the next character describes it very differently, and the third character is very different again. That�s our lives. If we pay attention to the most minute detail, there�s an enormous richness. It comes from accident.

...to the table of contents