While other CEOs are talking about ideas, Steve Jobs, CEO of NeXT Computer, says he's talking about reality
In 1976, at the age of 20, Steve Jobs co-founded Apple Computer, where he not only built the Apple II, but helped develop and manufacture the Macintosh and LaserWriter. Since '85 his entrepreneurial energies have been focused on running NeXT Computer, a leader in the object-oriented software market. Mr. Jobs spent an afternoon with us philosophizing about Netscape versus Microsoft, completing the computing loop, and the pros and cons of youthfulness.
The Herring: Beginning when Jim Clark met Marc Andreessen and founded Netscape, every technology executive seems to have had their own "road to Damascus" experience regarding their newfound faith in the Internet. How long have you been thinking about it?
Jobs: For probably seven or eight years. I don't know if you have tried the NeXT e-mail system, but it is really the best in the world. So we've been using the Internet for a long time to send mail to people. They never did that at Apple. The more relevant question, I think, is when did we start recognizing the value of the World Wide Web. NeXT has had a long association with the Web. Tim Berners-Lee, the European physicist who led the team that developed the original foundation for the Web, used NEXTSTEP. So we were somewhat exposed to it from the very beginning. But I don't think we quite got it until maybe two years ago. That's when we started to see that the Web was going to be phenomenal, and it was going to change the way people think of computing.
The Herring: How so?
Jobs: The old way to look at computing was as a straight line between the desktop and the enterprise, with the primary focus on improving desktop productivity. That world, as we all know, is owned by Microsoft. But the Web is changing all of that. One way to view the Web is as the ultimate direct-to-customer distribution channel. At least that's how NeXT looks at it. Now who cares about that? Businesses! Suppliers! They are the people who can best leverage the Web by using it to conduct business and make money. So the Web completes the computing loop by providing businesses with a new way to interact with their customers.
The Herring: Perhaps ironically one could say it's like the old mainframe computing model, but with all of your customers hooked into your network, too.
Jobs: Exactly! The browser is just a 3270 terminal [IBM workstation] on multimedia steroids. Right?
The Herring: What do you think about Netscape's vision that someday soon we will all be automatically hooked to the Net when we boot up our computers, and their Navigator platform will be our primary interface to the world?
Jobs: I wish the world could work that easily, but it doesn't. You are talking about ideas, I am talking about reality. Look, I love Marc Andreessen, he's a great guy. But he's young, and he's got Microsoft to deal with.
The Herring: But Netscape does have 10 million customers using the Navigator, and that is reality.
Jobs: Yeah, but they give it away. They have probably made $20 million off their browser business. Do you think they will make a lot of money on 2.0? It just ain't gonna happen. They don't have 10 million customers anyway, they have four million.
The Herring: We haven't personally checked the numbers, but Netscape claims to have a system in Mountain View that identifies users every time they fire up Navigator, and so they can verify those numbers.
Jobs: Okay. So maybe they do. And I think that's wonderful. But, by the way, I couldn't give a shit about the browser. We are not going to make any money by selling browsers, and I personally don't think they are going to make any money from it either. If you can get a browser from Microsoft for free, why are you going to pay $39 to Netscape?
The Herring: But if Navigator is platform-independent, and...
Jobs: But everybody uses Windows. Come on -- 90%+ of the people use Windows, so 90%+ of the people are going to hook into the Internet using Microsoft. Now, you know me, I love the Mac too, but I am trying to be really objective here. Plus Microsoft is...
The Herring: ...busy making all its apps Web-friendly.
Jobs: Microsoft is busy trying to kill Netscape. And it has a certain track record of being successful at those kind of things. So I wouldn't write off Microsoft right now. But all I am trying to say is that no one is going to make money by selling browsers. I do think a lot of people are going to make money off the pipes, but that ain't us. The pipe is going to be owned by the RBOCs. Pac Bell and all those guys are going to provide cheap ISDN lines into the home that come with a little box that turns it into Ethernet, and they are going to be impossible to compete with. But, as we've been talking about, the new Web set-up is just like the mainframe computing model, where all the apps will run off the server, and these will mostly be custom apps.
The Herring: Enter NeXT Computer.
Jobs: Well, as it turns out, the businesses that can best use the Web are the exact same people we have been talking to for several years about NEXTSTEP and Enterprise Objects. Those customers now have a real need to build custom apps on the Web so they can vend products, information, and services to their customers. As we started to think about it, we came up with four categories of things these customers are going to want to do with the Web. [Mr. Jobs gets up and starts drawing on his whiteboard.] One, they are going to do static publishing. That's where somebody makes a Web page and vends it. Anybody can look at it at the same time as 3,000 other people, and it doesn't change until someone goes in there and changes it by hand. The second thing people will be able to do, which is going to be a lot more exciting, is what we call dynamic publishing. There are already a few examples of people doing this -- like the Federal Express package-tracking Web page. You give it a number, it goes into four or five different databases and finds the information you need, and then presents it to you so you can browse it. Now there isn't a little gnome in there that makes up this page for you, the computer makes the page for you. It's a custom page answering your custom request, dynamically created on the fly. And this is just the tip of the iceberg. Have you seen our Chrysler demo?
The Herring: Only when you demonstrated it on CNN.
Jobs: I will show it to you in more detail in a few minutes, so you can have a better idea of what I'm talking about. The third big application for the Web will be commerce. The security issue here is the red herring, so to speak. And frankly, it is going to take Visa and MasterCard to solve this problem. Netscape can't solve this problem; we can't solve it. Before you start sending your credit card number all over the Web, you want someone to guarantee that if there is fraud, you aren't going to be held responsible. And who can guarantee that? Not Netscape. Only Visa can say that, and it will solve that problem. The real issue here, however, is that if you are going to sell something over the Web, and you're a medium-to-big company, you've got to have an order management system. But guess what? You already have an order management system that you've been running your company on! So, to be efficient, you need to tie the Web into your existing order management system. It becomes multi-platform that way, right?
The Herring: Right!
Jobs: Now you are constrained with the UI [user interface], but a lot of apps can be written in constrained UI. Look at the number of 3270 apps that have been written in the world and are still used. So, if you can constrain yourself to the existing UI today and write your app, not only will you get multi-platform capability, but you can roll your apps out to the contracted agents working for your company, and eventually out to your end-customers. For example, Merrill Lynch works with over 10,000 people who do not work directly for them, but help Merrill Lynch sell its products and services.
The Herring: How long do you think it'll be until this Web-centered world fully comes into play?
Jobs: Static publishing is already happening today. Dynamic publishing is just beginning to happen, but is really going to be the big thing in 1996. Web-based commerce should also start kicking in 1996, and, in my opinion, building internal apps for the Web won't really get going until 1997. When we looked at these developments, we realized that the final three require custom software. And that's what we do here at NeXT, Custom 'R' Us, right? So we created this thing called WebObjects to help make it easier for people to build custom apps for the Web. For example, it took Federal Express four months to build its Web site -- using WebOBjects, you could build that same site in four hours.
[Mr. Jobs then showed us the Chrysler Corporation Web site that the NeXT team built with WebObjects and an Oracle 7 database. During this demonstration, Mr. Jobs searched the site for several specific models of cars, at different price ranges, in different colors, and sorted in different ways, and each time he was instantly presented with a Web page that included all the cars he had requested. He also showed how the NeXT team had built a custom function into the site that allows customers to calculate their own financing options and identify which dealers have the exact models they are looking for.]
The Herring: That's pretty cool. How did you do that?
Jobs: It takes your request, parses it in WebObjects, grabs all the data, and dynamically builds it into a Web page for you to browse. The way we set up the car financing feature is that it actually sends an OLE call to another Windows computer that launches an Excel spreadsheet that does the calculation for you, and then OLE messages the information back and shoves it onto the Web page. You can't possibly do this in a static environment. I would think that this site is, what, an order of magnitude or two more dynamic than any other Web site out there right now. Wouldn't you agree? And we set up the whole Oracle database, we built the whole app, we scanned in all the pictures -- everything, in about 48 hours with four people. And, to reinforce something I talked about earlier, the site I just showed you will help Chrysler sell cars, because it distributes information to customers far better than Chrysler's dealers can.
The Herring: And pretty soon, with 3D immersion on the Web, you'll be able to get in the car and test drive it.
Jobs: I personally don't think that will happen for a long time. But what will come soon, when we have MPEG decoder chips in every computer, is the ability to download a high-quality video so you can watch the car drive around.
The Herring: Does WebObjects work across all platforms?
Jobs: It's very portable. It can run on our Mach operating system, it runs on Solaris, H-P UX, Digital UNIX, and it now runs on Windows NT. It's also fully distributed, so you can have objects on different machines and one object can send a message to another object without even knowing where it is. In fact, you can move an object from one machine to another without ever changing the app, it just automatically works itself out.
The Herring: It also seems to work pretty seamlessly with the Oracle database.
Jobs: We discovered over the years that almost all mission-critical apps make extensive use of databases. So we tried to figure out the coolest way to integrate data sources with objects, and we came up with this thing called the Enterprise Objects Framework [EOF]. In essence, EOF allows you to graphically connect the data-structures in your objects with any SQL database, and it will automatically -- automatically -- make the data in your object persistent and coherent with that data in your database without any programming. And it is exceptionally powerful. EOF is far-and-away the most aggressive database technology out there for objects. It's really slick. With EOF, you don't have to know about SQL. It has full TCP/IP communications built in, so you don't have to know about that. It also has Sybase and Oracle client libraries in it, so you don't have go out and buy those. You literally just point it at the database on the network and it works! So when we wrote the WebObjects framework, we based it on our experience with EOF. What that means is that your object doesn't have to know anything about the Web -- it literally doesn't have to know anything about its UI, it doesn't have to know about HTML, it doesn't have to know about URLs. And, like EOF, it doesn't have to know about the database or the connectivity. Everything is taken care of automatically. So WebObjects is far ahead of anything anybody else is doing out there.
The Herring: Why do you think you have built up such an advantage?
Jobs: The reason we are ahead, I think, is because our understanding of the fundamental business model of the Web is more advanced than Netscape's, or that of anybody else we've talked to. We've spent eight or nine years developing PDO [Portable Distributed Objects, NeXT's object model], we've spent four years developing EOF, and we have just leveraged that by spending about 1.5 years developing the WebObjects framework. The other guys haven't even gotten started yet.
The Herring: When will WebObjects ship?
Jobs: WebObjects is in alpha right now, it will go into beta by the end of this year, and we are shipping in production in the first quarter of 1996 -- my guess is by February.
The Herring: Do you worry about Microsoft?
Jobs: My goal over the next few years is to stay far ahead of Microsoft, until the Web is so ubiquitous, that even Microsoft can't own it.
The Herring: Could a Windows-compatible-only Web strategy become Microsoft's Achilles' heel?
Jobs: I gotta tell you, multi-platform compatibility ain't what it used to be. Windows has won. It beat the Mac unfortunately, it beat UNIX, it beat OS/2.
The Herring: But it took 10 years. [Laughs]
Jobs: We can all laugh at how long it took, but then we can all cry about the fact that it did happen. An inferior product won, but it won. And there is no changing that. I still think multi-platform is important, but not as much as it used to be.
The Herring: Netscape's vision is that multi-platform capability is important, and...
Jobs: Wait a minute, let's zoom back for a moment. It's not Netscape's vision, it's Tim Berners-Lee's vision. His original idea was that the Web would become the circulation system connecting us all together. Netscape embraced this vision, and it has done a better job than any other company in doing so. But as we all know, Microsoft has embraced the same vision Netscape did two years ago. So to state that Microsoft and Netscape have diametrically-opposed views would be foolish. They have both bought into the Tim Berners-Lee vision, and Microsoft is going to be a force on the Web, whether you like it or not. Look, I remember the day when Microsoft entered the application business for the first time -- its first programs ran on the Mac, not the PC. It was back in 1984, when we launched the Mac. Today, half of Microsoft's revenues come from application software, and it is the leader in that business. Now I am not a cheerleader for Microsoft, but I think it would be stupid to think it isn't going to be a big player with the Web. And don't get me wrong. I take my hat off to Netscape. I love Netscape.
The Herring: Why do you love Netscape?
Jobs: I love Netscape because I love any group of people willing to work their butts off for 18 months to get something done and take a new concept to market. I love that!
The Herring: How about Sun's Java software?
Jobs: My view is that putting a programming language like Java in the client will slow the Web down, and allow Microsoft to catch up. So while NeXT thinks Java is a fine language, and eventually it would be great to see it in the client, I actually feel that for the good of the Web, and for the good of the industry, the Web ain't broke, so let's not fix it. The most important thing right now is to let the Web accumulate users and establish ubiquity, until it's so entrenched that even Microsoft can't own it, and then let's add in all the cool stuff. Now I am not denying that the UI on the Web strains its use, but I am a little worried that in the microcosmic lust for perfection, macrocosmically we will give Microsoft the time it needs to own the Web. I hope that doesn't happen.
The Herring: We suppose that's one way to look at it.
Jobs: As an example, I predict that by the end of this year, Microsoft will announce that it has a Visual Basic variant or deviant that it proposes as the Web-client language. And Sun and Microsoft will have a war. And Microsoft will put everything it has into that war, because if it can win, it will have killed Netscape along the way. Netscape will put everything it has into that war, because if it loses, it is in trouble. So I ask you, who will win that war? Probably Microsoft. I hate to say it, but it has a lot more resources. So, in a way, Java may be the undoing of some very good things that are happening with the Web right now. I want to emphatically say that I like Java, but I am looking at it from a geopolitical perspective, not a technology perspective. So having said all of that, WebObjects works perfectly with Java. [Laughs]
The Herring: In our interview with Jim Clark, he said that Java "has a facility to protect you against the transmission of viruses, and a cryptography envelope that can wrap around and protect programs delivered over the Internet."
Jobs: Well, that is just not a true statement. We know a lot about cryptography here. We have invented far-and-away the best public key encryption technology in the world outside of what's inside the NSA. It blows RSA away. We have been told by people who know. Therefore we know that any language in the client is going to be susceptible to viruses.
The Herring: Any thoughts on @Home? You've already stated that the RBOCs will own the pipeline.
Jobs: I think several of the RBOCs will make a lot of money selling unlimited use of ISDN and an Internet account for $20 to $25. That's reality, that's product you can have in your home in January, and @Home is talking about a cable modem product that I may be able to get in a year or two. I mean, fine, asymmetric cable modems are very interesting. But all I'm saying is that you have these multi-billion RBOCs who already have customers in every home in their territory, and they have trucks with people who can install new products, and they have bought a zillion servers and have set up Internet farms, and they are ready to roll. I'm just mentioning that that is a fact.
The Herring: What about Apple?
Jobs: Well, I love Apple. I hope they make it.
The Herring: You sound concerned.
The Herring: So what is NeXT's growth strategy?
Jobs: We have three major new things coming out in the next nine months. One is D'OLE, the distributed OLE product which will be out by the end of the year. We have WebObjects shipping in the first quarter, which takes us into the Web market in a very big way. And we have OpenStep for Windows shipping in the second quarter. Our big initiatives then are really the Web and Windows. With the introduction of our Windows product, we are really going from having 10% of the seats available to us, to having over 90% of the seats available to us. And we think WebObjects should become pretty big, because we think the Web is going to be pretty big.
The Herring: Is an IPO in the near future for NeXT?
Jobs: We don't have to go public, but there are other issues such as employee liquidity and credibility with our customers. So I see an IPO sometime down the road.
The Herring: Closing comments?
Jobs: The Web is great because it breaks down two big barriers. It breaks down the platform barrier, because it is multi-platform, and it breaks down the internal/external barrier. Small to medium-sized customers will be able to share information seamlessly across the Web with their customers, and that should increase everybody's productivity. But, again, I think the biggest issues that are going to determine the fate of the Web are not technical issues, they are business and political issues. Maybe I am getting too old, but that's what I think.