After leaving Silicon Graphics, Jim Clark wanted to get into the interactive-television business, but wasn't sure where fire would next strike. With Mosaic, Clark thinks he has found the spark.
__When Wired last spoke with Jim Clark, in the fall of 1993 for a profile of Silicon Graphics Inc. (issue 2.01, page 116), he was chair of the red-hot company that continues to dominate 3-D interactive computer graphics. But Clark was clearly unhappy with how he'd been treated at his own company, and during the course of a lengthy interview, he let his hair down, revealing that he almost resigned when SGI executives, including company president Ed McCracken, were resistant to his ideas for bringing the company into the future. What he didn't say then was that while it appeared that SGI was adapting to Clark's view of the future, a future in which SGI technology would fuel interactive-TV networks and Nintendo videogames, he was viewed less like a visionary than as an eccentric uncle - tolerated, but not taken very seriously. "I was not happy there for about four years," Clark now says. "It was too much of a struggle to get anything done."
So in a February 1994 SGI board of directors meeting, Clark handed in his resignation. Within weeks, he was talking vaguely about a software company that would develop graphical user interfaces. During one conversation with Wired more than a month before he made public his new company, his response to a casual question about his thoughts on Mosaic elicited a telling reply: "Why are you asking me that?" He suddenly sounded wary, almost paranoid.
In April, Clark announced that he had formed Mosaic Communications Corporation - which would be producing a commercial version of Mosaic with new and improved features - with Mosaic author Marc Andreessen.
Michael Goldberg met with Clark at his new offices in Mountain View, California. After years of frustration at SGI, he seemed content.__
Wired: Six months ago, you were chair of the board of Silicon Graphics; now you're heading a new software company, Mosaic Communications Corporation. How did that come about?
Jim Clark:
A lot of the things that I was trying to get done at SGI were going as rapidly as a company that size was going to be able to push them. I wasn't happy with the pace.
What was wrong?
SGI isn't pushing volume into the marketplace. Volume is going to set the standard and it may be already too late. My challenge to the company is make that technology available at prices that are competitive with PCs. It's doable; just do it. That's what I used to tell them, but no one ever got it. If I wanted to right now, I could just blow SGI to smithereens. I'd never do it. But all that's got to happen is somebody takes the basic graphics library - there's no patent on it, I had that technology, parts of it, before SGI ever started - and builds an absolutely killer chip. It does take some knowledge, and - fortunately for the company - most of that knowledge is held within SGI. But just build it, design it. Design it for ultra-high volume, strike a deal with Bill Gates. Say: "You embed this set of graphics calls into your operating system, and I will sell this chip set, just like Intel sells the X86. We have a balanced CPU and graphics system; you support it in your software. And there would be no reason to buy these high-end expensive things that SGI sells." But I'm not going to do that. And I don't think anybody else has either the knowledge or the ability to muster what would be required. But it could be done. That's their most serious vulnerability. It's an excellent group of people but they - and people at every other company - begin to define themselves by what they have been doing, not by what they can do.
That seems to be human nature. It's comfortable.
Something has gone wrong when you stop valuing the person that tries to set a little bit of farsightedness. I didn't feel valued there. I feel incredibly happy to be out of that place. It was too much of a struggle to get anything done. And that's too bad. It's a sign of bad things when a company 154 c loses its founders. I think it was bad when Steve Jobs left Apple. If Steve had done the NeXT product inside Apple, it would have been wonderful. But he was going against the grain when he was doing it outside of Apple. Just as I would if I had left SGI and started another graphics workstation company. If I had gone down the road of trying to make this chip and do a PC with advanced graphics, it would have been a real struggle, because I would have been going against the whole culture over there. Right now no one over there has anything against me. I'm not trying to hurt them. None of this is aimed at hurting; it's more just observation.
You're known for speaking your mind. You've made comments about Apple and Trip Hawkins and IBM -
What I said about Apple was that they made a serious mistake by not capturing something else besides user interface stuff and trying to protect that when, in fact, Steve Jobs copied that from Xerox. He had no patent. And had Jobs been there, he'd have been pushing them to create something new, rather than trying to protect their current territory. Apple circled the wagons around GUI, and that was the wrong thing to do. GUI wasn't protectable. Rather than creating a new thing like the NeXT box and like object-oriented programming or any of that kind of stuff, they circled the wagons. Later they started to make notebooks, that was a good thing. And what I said about Trip Hawkins was that he was hyping the 3DO player too much. I still believe that Trip got way out ahead of himself. He basically started believing all of the things he was saying. He complains about me in some recent article, "I didn't ever do anything to Jim Clark." No, he didn't ever do anything to me except he trivialized 3-D. Claimed it was easy. It isn't trivial. It's hard to do. It's hard to pull off all those tricks with smoke and mirrors. So he trivialized it, and I think it's going to catch up with him. He made the world believe he was going to deliver 3-D for $500 to $600. And I don't think he will. No, I don't think I've gotten myself in trouble.
Why Mosaic?
Mainly because I ran into Marc Andreessen immediately upon leaving SGI, and Marc was the author of Mosaic. The original thought was to go into interactive television in some fashion. But what is Mosaic? A navigator for interactive stuff on a network. The network happens to be called the Internet, but the physical network is slowly improving in bandwidth and someday it will be capable of carrying video. The slope of growth of the television industry is zero. It's even negative. Television doesn't change. Cable is nothing but an overlay of a physical delivery scheme for broadcast television that has been around for 40 years. There's no two-way interactivity, nothing. So I began to think about the difficulties of the transition of the cable industry into interactivity. It's a completely daunting task. First of all, you have to cause the television industry to get accustomed to digital technology, which is a major change. Then it has to think of two-way interactivity, which is a major change. And it has to make the physical network carry switched-video capability. That's just a whole group of major changes.
That's not even talking about the consumer.
Right. People don't have problems with interactivity on computers. More and more, computers are being built so you can see video on them. You look at those dynamics, and you look at the dynamics on the other side of the ledger, and you say, What are you doing over here? Get over there.
I'm sure you've followed all the concern about the coming commercialization for the Internet.
That to me is peculiar. When the phone system was invented it was primarily for voice. We commercialized it when we began to use it for business, and we commercialized it further when we began to do data transfers over the wire, money transfers over telephone lines. It's exactly the same thing. Commercialization of the Internet is as inevitable as the sun coming up tomorrow.
While heading SGI you came to believe that the future lies in interactive TV. You made the deal with Time Warner for servers and set-top boxes. This is a big change.
Tele-Communications Inc. is a great company. But it represents a blockage in the system because it doesn't have the bandwidth to deal with all the deals. If I'm sitting here waiting to do a deal with the cable industry, I have to do it with Time Warner or TCI. And that's just too small a market. I'd rather deal with something that's more organic, that has lots of ways you can get into it. That's how I view the Internet. Even by the end of the year, based on current growth rates, the total number of users on the Internet is going to be as big as cable. US cable serves 60 million homes; we're currently at 25 million Internet users. By the end of the year, that will be 50 million. By the middle of next year, you'll be over and above what US cable operations touch. It's a big market, and no single entity is in control of it. It's much more organic and grass-roots, which I find appealing.
Ten years from now, are movies-on-demand going to come in over the Internet to computers?
I think what you call a computer today will mutate. A stripped-down version of today's computer will be used as an interface to a television. The television display will become higher quality. The television display will be able to display text. In other words, it will look like a computer screen. So the television as we know it goes away. And now the ordinary consumer says, Well it's a better-looking picture, and I've got this different remote, and I can point to things on the screen, and it's a higher-quality picture. But I can watch movies and see video - ordinary television.
That kind of user you could call the couch-potato user. Meanwhile, there are the 25 million plus with computers. Even the average person who uses a computer is sophisticated, compared to the couch-potato user.
And that audience is growing every day. The cable industry, or traditional television, counts on the stupidity of the consumer. The PC counts on the intelligence of the consumer. But the point is, even a nonliterate person will be able to use what the PC becomes, because when you remove the keyboard and put a big display on it, you have television, but you also have computing. You have things that you associate with a computer today. Those things you traditionally associate with each medium are going to come together, and I think they're going to come together with the center of gravity around computing, not around the old television technology. When you change the entire screen, the entire analog system to digital, the underlying broadcast to an interactive schema, you have computing, not television. So the computer becomes a TV, rather than the TV becoming the computer. Maybe the flaw in my thinking was that I felt television and the cable industry could move quickly. I thought they could. I was wrong. I don't believe they can. The rate at which cable can change the installed plant doesn't compare to the rate at which computing technology is going to become a television.
How did you conclude that?
Well, I wasn't allowed to come to that conclusion when I was the chair of SGI. I couldn't say, Let's take our technology and become a PC. That would have meant taking the MIPS machine and building a PC out of it. That means that you can't add any value because the people in Redmond, Washington, determine the flow at which you can add value or add capability to DOS and Windows. Those are the constraining things. They constrain that market.
Do you think Time Warner and TCI and the others are wrong about the infohighway? Are their tests going to fail?
If they wanted to be successful, they would have had to act in unison. Time Warner and TCI had to say, We hereby endorse this thing. Had they said that, the whole world would have focused behind it. Everyone would have started applying intense energy into making it low cost, et cetera, et cetera. But Time Warner and TCI couldn't see eye-to-eye. I tried my best. Even with complete unison they would have had to run like hell, because the computing industry is changing in a nonstop way for many, many more years.
Has Malone actually spent a day on the Internet?
I don't know. Malone is a very smart guy. He'll pick up on this very quickly. I just don't think he's figured out yet how to leverage his current holdings, a combination of content via Liberty Media and the wire business, into this new area. By the time they get interactive switched-video circuits, the Internet will have scaled into something that can do the same thing. Or the Internet will be able to ride on top of it. The way to think of the Internet then is a protocol riding on top of whatever physical network is there.
Would it be difficult if, right now, he wanted to put computers on his cable systems?
He could offer Internet access right now. Intel has just introduced a modem to let anybody connect a PC to the cable. They fully intend to do that. The fascinating thing is, once they get there, how do they control the content?
They can't.
So they become a utility. A data-carrying utility.
Look down the road 15 years: what's coming?
Fifteen years is adequate time to have lightweight portable things that give you most of the media you want to read or view. The notion of media will be completely blended in one digital stream. Moving pictures and static pictures and text and audio - the mix of any of those will be commonplace. Imagine turning the pages of a magazine: every one of the images is a sound clip, and when you put your cursor over one of them, the audio comes. That's absolutely going to be happening in 15 years. Full convergence of media and television and computers.
Do big companies dominate?
I don't have any way to predict that. If they act properly, they can. If I were Sony, just to give you an example, I would be acquiring more software. Not content, but more underlying software. Because software is where the added value is going to be, and they're not acquiring any. There's a layer of software that's enabling, that connects content to hardware, that they don't know anything about.
If you had to start a hardware company, what would you do?
I've got an idea there, but I'm not going to start a hardware company. I would build an extremely inexpensive box that connected to cable television and that had enough power to take in decoded video signals and programmability and the ability to run Mosaic. Now, such a box could be built today for very little cost. I think what I'm talking about is what the set-top box should become. But whether it will is a completely open issue.
Is it true that you were unfamiliar with the Internet until recently?
That was a misinterpretation. Of course I knew what the Internet was. But I hadn't thought about what the implications were in terms of its growth rate. The Internet is a protocol running on top of a physical connection network. That underlying physical network can be changed and the protocol can still be called the Internet. As the underlying physical network changes, all of these applications you develop for the Internet port right over there. You would be astounded at the number of terabytes of data that are served by Mosaic today, daily, running on numerous platforms. It's sort of the standard network tool.
What are you going to do with it? Right now it's available for free.
It's like saying you get 1.0 for free. What about 2.0? Do you want it? Of course you want the upgrades and the new features. So the first release was free. Second release? Depends on how we structure it. It's a new game, and we're going to build a business around it.
How long before a commercial version will be available with security ad billing?
Six months to a year. Actually, it will be later this year.
You started SGI with half a million dollars and six students and turned it into a billion-dollar company. Can you do it again?
I'm relying on Andreessen to strike fire the second time. The vision is really Marc Andreessen's, not mine. And I think he can strike fire the first time. He kind of struck it already, and now he's going to do it in a commercial sense.
