While many executives have every moment scheduled from dawn to dusk, Microsoft Chairman Bill Gates was unhurried when Context arrived to interview him. He had been managing his burgeoning empire by reading and answering e-mail--a task he returned to at the close of the 70-minute conversation.

The Microsoft offices still feel almost as college-like as they did a decade ago, when Mr. Gates had just begun to generate headlines. A series of low buildings nestle among the pines--and the construction cranes that have not left the site over the years. Volleyball courts and picnic tables sit between many buildings. In Mr. Gates' modest office, the dominant piece of art is a colorful poster of the circuitry in an Intel Pentium chip. Mr. Gates, almost as casual about dress as he was 10 years ago, was wearing a pastel pullover.

Much has, in fact, changed. Mr. Gates is no longer the nervous kid who rocked back and forth while his old friend and colleague Steve Ballmer rocked side to side. Not only has Microsoft thrived, but information technology has risen to ever higher levels of prominence. Technology has become a core piece of what most businesses do, of what they are. So Mr. Gates' office, physically on the second floor of Building No. 8, is actually sitting at the summit of the world of commerce.

In our conversation, Mr. Gates provided some glimpses of what the digital future will look like and made some sweeping observations about how companies can prepare for that future.

CONTEXT:Where are you spending most of your energies these days?

GATES: Well, first is what I refer to as the "digital nervous system" issue. We are trying to improve our products in such a way that any company, of any size, can run its business on our software, using solutions built on top of PC hardware. Closely coupled with that is our work on reducing complexity and on lowering the cost of ownership of personal computers. A third thing is our research.

Computers today are phenomenal in their inability to see, listen, or learn. They can't speak to you. They can't read your handwriting. They can't read a piece of paper. They can't learn how to behave better in the future based on how you've used them in the past.

All the value in software over, let's say, the next five years will go to the company that solves those problems.

CONTEXT:How do you do that?

GATES: We're somewhat dependent on Moore's Law (which states that microprocessors will double in power every 18 months). Our solutions to today's problems probably won't run on today's hardware, so we count on (Intel CEO) Andy Grove and his competitors to continue the magic of semiconductors. Most people are a little disillusioned in that it's been more than 20 years since people began promising them that computers would do some of those things, such as speech recognition. They even saw demos where it seemed the computer was doing those things. But I know that, in the future, the bulk of the value of an operating system will come in these kinds of capabilities.

CONTEXT: What are the main technology issues that senior executives need to think about over the next two or three years?

GATES: Everything that has to do with information movement. I mean, people are underestimating how easily information will move around in a 10-year time frame.

Just to talk about this general subject a bit: These books that try and tell businesses how to play for the future, most of them are pretty bad. I just read Competing for the Future, which was written in 1993. The authors are two smart guys. They're probably as good as there is in the field. So what examples did they pick? General Magic. Yeah--they understand the future. Apple Computer? Every example they gave, with the exception of Hewlett-Packard, was a total joke.

Why isn't it easy to give examples of companies that plan for their future against fairly dramatic change? Well, that's the nice thing about the way capitalism works. New companies generally come in. Yet, nowhere in any of these books is there even a single sentence that says that when you reach a strategic inflection point, there will be a lot of cases where your company's skills and background will no longer be relevant and, so, in a sense, you shouldn't even try. Andy Grove's book talks about Wal-Mart coming into town. He has this paragraph saying, specialize a bit, do this or do that. I say, hope your real estate is valuable. Don't try to resist a strategic inflection point that's so dramatic.

I think the easiest thing to say about technology and how people should think of it is to say: Imagine that information was just totally available. All the information you wanted about what is going on with customers, the market, resources, and possible partnerships was totally available to you. In 10 years, you'll just take it for granted.

CONTEXT: Are you suggesting there will be very low transaction costs?

GATES: Even more than low transaction costs. Today, you find the market's not mediated very well because it's very complicated to catalog all the information about buyers and sellers. And, if the resources you want to use are at a distance, the ability to take advantage of them is very limited as well. Those limitations will go away.

CONTEXT: Are you saying that businesses are going to be very tightly defined and that it won't be possible to be vertically integrated any more?

GATES: This is a very tough problem. The scale issue. It's always interesting to look at different businesses and say, What scale of firm makes sense? National law firms never succeeded, but global accounting companies have a certain part of the business they dominate. You almost have to look industry by industry. People have always argued against scale. Well, sometimes scale works. I mean, for GE, scale works.

CONTEXT: You can certainly have tightly defined businesses of great scale. You have one.

GATES: Yes, you can have a very high share within your tight definition and, depending on the market, you can do very well. I mean, Coke's scale will not change. They are a very focused, tightly defined business, and nothing we're talking about here is really going to change the nature of their business. Maybe they should learn how to run ads on the Internet, but so what if they're a couple of years late?

But there are a lot of businesses that aren't tightly defined today, and to become tightly defined they have to spin off things or shrink.

You have to pick your core competencies pretty well.

CONTEXT: It's a question of what you want to leverage, right? We're seeing a number of industries really change. For instance, the airlines have always leveraged their capital equipment.

GATES: There's something irrational there. More capital is put into that business than should have been put into that business.

CONTEXT: What does Microsoft leverage?

GATES: If we have anything, it's the ability to develop and market and support and sell software on a global basis, and we do these high-volume products. That's us.

It's pretty tightly defined. We've always been fairly tightly defined. That was our advantage over other companies in the computer industry. And that's what's so funny. You get these new businesses looking at the Internet and wondering where the value is going to be. Therefore, you have a temptation to have a strategy like people have in communications. Right now, people in communications are saying, well, your strategy should be to package local access, long distance, wireless, maybe some movies in there. Packaging is supposed to be the big value-added.

I always think it's pretty tenuous whenever you define a business according to that sort of bundle of things. I think markets are awfully damned efficient. Even at the consumer level, people notice whether, if they buy a la carte, they can get a much better deal. I think people are sort of kidding themselves. No one knows which piece of a bundle is going to have a lot of value. Maybe all the pieces will have less value than they've had in the past.

CONTEXT: What about leveraging the relationship with the customer rather than leveraging capital or equipment?

GATES: But nobody owns the customer. Nobody owns them.

CONTEXT: For companies that leverage a skill like you do, that's one thing. But for companies that have played the game across a wide variety of fronts and have no particular skill that's a clear leverage point, the one thing they do have is relationships with customers. Can't companies do something to hold customers through mass customization?

GATES: I totally agree, and that ties back to the Internet in a big way. To map this into Microsoft terms, we have a business called Microsoft Office, one of the most profitable businesses around. It makes Coke look bad. Today it's being characterized by the one-time sale of the physical package that you buy. In the future, we'll get your permission to send you mail once a month and collect bits that describe how you use Office. We think we'll get 90% to 95% of the people to agree to that. We'll make sure what we send is very informative to them and tells them what new pieces of Office they might want to download. And, if they're paying our annual fee, of course, they get everything in there. Our relationship with that customer will be drastically different. We'll know the nature of their current machine.

CONTEXT: You will mass customize, is that what you're saying?

GATES: Absolutely. We have to. That is the future of the Office business.

Our relationship with our customer will be our big asset. We'll use that asset and the technology asset to, hopefully, create something that's very hard for people to duplicate.

CONTEXT: Let's go back for a minute to this idea of ubiquitous information. It seemed to me that an assumption you made very early on that was right is that computing power would essentially be free. That's one reason why you were willing to develop complicated, graphical software while IBM wasn't. In your book, you say that you now work on the assumption that communication capacity is free. Are you saying that five, 10 years out, information will be free?

GATES: Not free. But any executive with help that is at all competent will be able to take advantage of the incredible building blocks that are out there. They'll be able to analyze their sales, analyze who's working with a customer, make it easy for customers to give feedback, or make sure that all of their customers who really care are instantaneously aware of the things that are going on in a way that's tailored to who those customers are. It's not like just one company will figure out how to do it. But between now and 10 years out, there will be a few who get it faster than the others.

CONTEXT: Maybe we can talk about ubiquitous information as Gates' Law. But we need a measurement to go with it, such as the one with Moore's Law.

GATES: Yeah, but for most business people it's almost frustrating to tell them what the future is. For everything they learn, they are saying: What do I do about this in the next month? It's hard for them to do anything with it if you give them something wild--such as that computers will talk or listen--and tell them that's 10 years out. They need to know what that will look like year by year, what the effects will be.

CONTEXT: In your travels, have you run across particular people or companies that are thinking about technology in an interesting way?

GATES: I don't think many companies are doing it very well. There are companies whose whole business relies on information so you'd think they'd be super, super strong--like people in the financial sector. But I can't say there's anything super impressive. People are smart enough to use the current building blocks. You know, their kids come home and tell them they ought to be doing something or other with the Internet, so they go back to work and ask around about it.

What you like in a business is to have a set of customers represent the future. That way, you can test whatever it is you do against that customer set. In my view, that customer set is kids on college campuses where you have Internet terminals everywhere--where people expect to get their class assignments that way, where they can communicate with their friends, and where it's just part of their life. That mirrors the Web lifestyle people will take for granted a decade from now.

A lot of companies don't even have electronic mail set up so you can count on everybody reading a message within one working day of when you sent it. If you don't have that, you know, it doesn't matter what sort of dilettante project they take you over to show you that they're working on. Without that basic ability, a company isn't at first base in terms of the Information Age.

But who's really great at using information technology? We're pretty damn good. There are some things we do, like our sales tracking system, where we can look at our sales in some incredible ways. Who else would we expect to be good? Intel? Not really. Various computer companies? Not really. They're just sort of catching up. Everybody buys SAP. How can you be differentiated when you're all buying and using SAP? So, people have decided that they're not going to be much ahead of anybody else.

People using their Web sites? Yeah, there are some nice dilettante Web sites where somebody got a huge budget so they could say GM is going to have a cooler Web site than Ford. That's nice. Is it really changing the culture of the company? Are they redesigning the product development process around that? If you take these huge data processes like a new car, a new plane, you think somebody would just take a clean sheet of paper and do it right.

Boeing claims to have done this. But the way they go about it is just so tradition-bound. Even though the files are digital, they didn't change most of the way they organized. They liked the way they did it, they knew they could get the safety requirements done right that way, and they don't like to take risks. They're not the right guys to prove to the world how new design processes should be done.

CONTEXT: In my experience, there are two places that push the envelope. One is government, because government always deals with volumes that are incredible. They don't always spend wisely and all the rest of it, but they push the envelope. The other place is Wall Street. The reason is the volumes they deal with, and also because they have to strike a balance every night or else they're out of business

GATES: O.K. I'd say Schwab and Fidelity use information technology pretty well. They are very smart buyers of information technology. They know what they're doing. They're willing to try out all the new things and decide for themselves which ones don't work. They are very aware of what the other one is doing. And there are parts of banks that are like that. But I don't think Wall Street is very good. Take tracking and everything that's going on with a particular customer. They don't have a system to do that. It's pretty fundamental.

CONTEXT: USAA does a reasonable job in the insurance area.

GATES: Yeah, we finally got them off of a pure blue (IBM) strategy. I think they might have been the last company in the world to go away from a pure blue strategy. But when they switched, they switched in a pretty dramatic way.

You know, there are well-run companies in this world, companies like Geico that don't make mistakes in their use of technology. But most of the people you think of as leaders got in and used pieces that are now obsolete, and they're not quite sure how to re-do them. Take Sabre. In its day, it was a very nice thing. But now they've got a real challenge. They actually understand that, but it's all a big mess. We're doing some work in that area, so we have a bias.

CONTEXT: Let's turn to the Internet. What are the three or four most surprising things you've learned about how people use it?

GATES: It's worked out to be an even better customer communication vehicle than we expected. We're in a business where we have 200 million people using our operating system. But there is a core set of only about a million people who other users look to to find out what's the latest, how should they do things. Maybe it's two million. The percentage is still quite small. Those people are on the Internet, are willing to go to our Web sites, and spend lots of time. Our site has taken on a different character, and the richness of communication has become more like what we have with a developer.

We have a pretty unique customer set. By definition, they have a computer and use software. But, in the same sense that I said university campuses are a leading edge, the way we use the Internet as a marketing vehicle should be a leading edge. People should look at that. How many e-mail names of people do we have? How many do we have permission to really interact with? How much money are we spending on our advertising? In 1997, we'll spend about $30 million on the Internet, which will make us, I think, the largest.

CONTEXT: How big is the medium for commerce? When will real profits come?

GATES: The thing on the Net is that a lot of people are just losing money hand over fist. We found that 1% of Internet sites claim to make money. Many of those people that have claimed to make money absolutely do not make money. I guess it's prestigious to say you make money. Or to say whatever it takes to get into Business Week.

Anyway, whenever people are putting together an Internet thing, they have a tendency to think of how they compete with non-Internet companies. But they don't take the next step and say that the barriers to entry on the Internet are just so low that, once they are clearly there doing well, somebody else could come in. What is it about the way they present information that the other guy can't look at it and do too? And have they really thought through how they compete against electronic competition?

You know, we have to go out of our way saying a million times we will not be a bank. We're not going to be a bank. If somebody handed me a certificate that said that I could be the only bank on the Internet, though, I'd be very tempted to be a bank, I have to say.

Who wins in the long run in these electronic things? That depends basically on brand and on customer relationships--your understanding of the customer and how you work with the customer.

In any business that doesn't have friction in it, it's harder to make money. And the real world has a lot more friction in distribution and geography than the electronic world. But it's all good for customers. If you decide that the airline industry is a total net loser, should you try to stop those losses from happening? No. It's a great thing for customers. Many, many businesses on the Internet are like that. You just know that the customers will be better off because of it.

There are going to be a lot of articles about how tough it is to make money doing content on the Internet. The next four years are going to be pretty barren for everybody. A lot of money will be lost. A lot of people will decide to give up. And so it's all a belief that as this thing becomes mainstream, as the devices become easier to use, cheaper, and portable, as this becomes so mainstream that it's part of your everyday activity, then the economics will be there. The interesting thing is that you can't wait to start investing if you believe that, if you're that optimistic.

MYHRVOLD'S COROLLARIES

Gates' Law about ubiquitous information isn't the only big idea floating around Microsoft these days.

Nathan Myhrvold, a Microsoft senior vice president and its semiofficial visionary, says-as a sort of corollary to Mr. Gates' idea-that all kinds of everyday devices will have computing power added to them and will be linked into computer networks, whether locally within offices or globally over the Internet. He's talking, for instance, about processors that are increasingly being built into industrial equipment and that can warn a technician of a looming problem in time to prevent a breakdown. Or cars with processors that detect when an airbag deploys, place a cellular phone call to "911," and then connect the emergency operator to the driver so the operator can determine whether to dispatch an ambulance. Certainly, many in the computer industry see cheap processors-referred to as "jellybeans"-moving into all sorts of new types of products. But Mr. Myhrvold is even more ambitious with his first corollary.

COROLLARY 1: Any device that uses electricity is likely to be revised in the future to have a central processor, memory, software, and a global network connection.

On a more subdued note, Mr. Myhrvold indicates that Moore's Law can create too rosy a picture about the progress the computing industry is making.

COROLLARY 2: Users' expectations and desire grow faster than the ability of either software or hardware to satisfy them.

Mr. Myhrvold says computer users won't even see their applications run much faster despite the more powerful computer that are made possible by Moore's Law.

COROLLARY 3: Software developers will use up new computing resources slightly faster than they become available.

Of the corollaries, this is the most widely accepted in the computer world-although it's often stated more irreverently. In a reference to Andy Grove, chief executive of chip giant Intel, and to the ever more complicated software produced by Mr. Gates and Microsoft, the line in the industry is: "What Andy giveth, Bill taketh away."