Speech Transcript – Craig Mundie, National Association of Broadcasters Conference

Remarks by Craig Mundie, Group Vice President, Platforms and Applications, Microsoft Corporation

National Assocation of Broadcasters Conference
April 7, 1998, Las Vegas, Nevada

MR. Mundie: Thank you.Good afternoon.In the next 45 minutes, I’m going to share with you a little bit of the vision that Microsoft has in the business dimension for the impact of the changes that are coming to television. And, with the help of a few demonstrations, I’ll talk about how we see some of this being realized.

I’d like to do this in three parts.One, to just look back retrospectively almost exactly one year to the speech I gave at NAB ’97, and talk about my view of what has happened in the industry collectively in that year.I want to talk about defining digital television — what is it — because in some sense the vision we have of it may be a little different from that which prevailed for quite a few years.And then how does this affect the business of digital television.

A year ago, I gave a speech in the Sands at NAB ’97.And in that speech, we announced three different things.One, that was the day we signed the definitive agreement to acquire WebTV, then a company less than two years old which had been one of the pioneers in developing interactivity for television.

Second, we announced the formation of a loose affiliation of companies called the DTV Team, which at the time was Compaq, Intel and Microsoft, and has since been joined by Lucent and several others.But the idea was to get a group of companies together that could work to educate people toward a different way of thinking about bootstrapping into the world of digital television, both at the standard definition, if you will, and ultimately higher definition levels.And we proposed a subset specification for which we coined the term
, and a layering strategy I’ll talk about a bit more, so that you understand the reasons for that.We also indicated at the time that we would include a broadcast infrastructure capability in Windows technology going forward.And, in fact, in 1998, we expected that digital-TV-ready PCs would arrive during 1998.

As we look back now at each of those different aspects of the announcement, WebTV has certainly met our expectations.We indicated at the time that we would begin to reconstruct WebTV on using the Windows-CE operating system, but that it would take place over time. What we focused on at the beginning was the natural migration of WebTV from a system that didn’t really include television features, but rather just used the television as a display and put the Internet on your television.

We changed it to a system that really integrates completely with TV, and I’ll demonstrate that in a few minutes.Compared to where we were a year ago, we’ve had over a factor of six in increasing growth.We had something under 50,000 customers last year, and only in the Internet access capability.And today we have about 325,000 customers for WebTV.These are active online users.

Interestingly, the demographics of the WebTV customer base spans the entire demographic

or normal demographic profile — of the U.S. population.About two-thirds of them do not own a personal computer.Those that acquire WebTV actually spend two-and-a-half times as much time online each month as the typical PC user does who is connecting to the traditional online services or Internet services.And with over 300,000 people, we certainly feel that we’re getting a very, very solid sample in terms of what the population is like; what the differences are in the reaction of the people who got started in the
version, which is just the Internet services; and the over 30,000 WebTV Plus receiver customers. WebTV plus, which we introduced just before Christmas and have been selling since the end of November, integrates the TV capability.

The theory behind WebTV and, in general, what we’ve been trying to do in digital television is to give the consumer more choice and more control over their television experience.We want to give them easy access to information, and we believe using, for example, the vertical blanking interval as a transport, we can begin to provide interactive programming today.I’ll give you a demonstration of some of the things people are doing in that area.

What you see here in this screen shot actually is a feature we’ve now dubbed
“WebTV for Windows,”
which is the first step in building the same type of TV capability into the PC platform as we have already in the WebTV platform.In the case of WebTV and other appliances we’re building with Windows CE, the goal indeed is to build appliances — very low-cost devices, very simple to use — and with them be able to complement the very rich but sometimes more complex environment of the full personal computer.

We know that to ultimately get PC penetration above the 40-odd percent level that it is in U.S. households today, it ultimately has to become less complicated.And we’re doing a lot of work in that area already.But to round out its multimedia capabilities, Windows 98 has a big focus on adding state-of-the-art capabilities in multimedia — not just CD-ROMs, and music-CD playing as we’ve had in other standard PC products;but DVD capability and AC3 audio decode facilities. And now, today, both analog television and, in an exactly identical architecture, full digital television capabilities built into the standard PC Windows 98 operating system.

Here is a screen shot which is very, very similar to the facilities and cosmetics of what we have as the guide in a WebTV box today and the guide that we will introduce in Windows 98 in June.The users of the WebTV features in Windows we expect will actually be largely a different population than those who would use a WebTV product. If you think of WebTV primarily as a television appliance in the home, then people should enjoy the PC’s ability to be a television — and, in fact, we do expect some people to deploy them in their entertainment centers as, for example, a fairly high-grade, fully capable personal computer and entertainment center. But we think that many uses of video, and many potential new businesses in broadcasting, can be done using the digital spectrum and providing television services not only to homes, but basically to offices, and schools, and potentially other applications, for example in training.

The features of WebTV for Windows include video reception. (At the discretion of the OEM obviously: they have to incorporate a tuner in the product, or with some external mechanism.) We believe starting this year, more and more PC manufacturers will incorporate this capability directly into most residential model personal computers.The PC also provides the same interactive programming model now as WebTV Plus.It also handles streaming media, inheriting that capability from the work that was done in the Internet activities.And we’ve also added an architecture to support data delivery over any of the broadcast networks, whether cable, satellite, or terrestrial.

Last year, we talked about digital television and tried to highlight some of the differences in the view the computer industry held, which was different than that which the consumer electronics industry held. It became clear, particularly after we made the presentation, that it is a complicated system, this end-to-end thing we call television.And in a way, it’s getting more complicated, particularly for the people who have to produce it and transport it, than what we’ve known in the past as NTSC or PAL.

In particular, we put up this picture that showed we viewed this as largely a three-stage process.One was the production phase, where with content, capture storage, and ultimately production processes, you needed to produce the product that you were going to send to the consumer.The second phase we think of as the broadcasting phase.Here the key question is, how do you do the compression?Ultimately, what’s the format for transmission?And how does that impinge on the receivers?And then the presentation phase, where the question is, how do you display this?Do you put on a flat panel like this one here?Do you put it on a projection system?Do you put it on a traditional CRT?And what are the costs and, if you will, price/performance tradeoffs in that?And how does that affect particularly the integration with computing devices?

Many people, you know, have characterized the debate between the computer industry and the television industry as being over this question of interlace versus progressive.And that, in a way, is an unfortunate simplification.In fact, the thing that we are really concerned about, or have been concerned about, out of this entire process, is really the question of what goes into the transmitter.In particular, what goes into the compressor. And the reason we’ve been focused on this is largely because we think that it has the most effect downstream in terms of what the ultimate cost of the receiving system is.And it does have some effect, ultimately, in the display systems as well.

And so, when we’ve talked about emphasizing progressive transmission, we really were talking about emphasizing only putting progressive inputs into the compressor.In fact, in support of this, there was an announcement this morning between Sony and Microsoft where we’ve agreed we’re going to collaborate in a number of areas regarding the use of Windows-CE and their home networking architecture.But, we also reiterated that Microsoft is a strong proponent of capturing video and producing it in the highest quality that is possible and economical.And certainly today that probably still remains 10 ADI, no matter what your predilection is in terms of the transmission and reception environment.And while that may change over time, it’s also clear that in this digital world, you don’t have the same isomorphic process that you had the world of NTSC analog television.

At each of the stages in this pipeline, in fact, you can make some type of conversion, perhaps value-added, perhaps just for convenience or efficiency, as you go down the chain.And so, for us, the most important part always has been the question of what gets put into the compressor and, hence, what is transmitted.

For us, progressive encoding has been the key, recognizing that every television system will have to deal with a multiplicity of formats.You have all the Legacy 601 compatible video to deal with and to produce and post-produce.And, of course, now it’s pretty clear that there are going to be a wide range of capture and recording systems available.The important thing to us was, if you happen to have interlaced material to put into this pipeline, the most beneficial way to do it was to convert once — at the end of the production cycle, and code only a progressive picture, and transmit it down to this receiving environment.The reason is that that allows the integration of the computing and display environment at a minimum cost in the coding hardware for these devices.

We also had recommended, and only time will tell whether people ultimately adopt any of these recommendations, that people start with a subset of these formats, and over time work up.The steady progress in computational capabilities, the declining cost in memory, and more sophisticated compression algorithms

for example, MPEG 4, and other things that are already in use in the Internet environment

might allow even the digital terrestrial broadcaster to aspire to a 1080 60p picture, which people now clearly recognize.

In fact, Sony reiterated in our press conference this morning that the ultimate objective in their next three to five years is to provide 1080 60p production capabilities.And that will leave us with the question as an industry, if you don’t adopt something like this at some point, how do you, at least in the terrestrial channel where you’re bandwidth constrained, ever be able to go beyond the top of the current APSC format chart?

It’s also true that in this environment it costs nothing essentially in any device, whether it’s a PC or a TV, to output to an interlaced display system.So, if people still want to buy high-quality interlaced CRT-based products as their display systems in the home, there’s really very little cost associated with converting to that environment.

So, that was the environment that we had a year ago; those were the recommendations we had made.It’s been interesting for me personally, and for those of us that have been close to this question, to look at the progress that’s been made with progressive transmission strategies in the course of the last year.

First, there was a lot of confusion about whether or not these layered formats could actually work.And back in September, Panasonic actually demonstrated in laboratory tests that, with virtually no loss in efficiency or picture quality, you could send things layered, and as a result have a family of TV products over time, all of which could increase the capability and be backward compatible.

In November, we demonstrated that 480 progressive video could, in fact, be directly integrated into the video subsystem of then-standard personal computers.And some of the demonstrations I’ll give you today show what results from that kind of integration.But, in fact, it was some of the first time that anybody had really been able to see, essentially, native 480 progressive digital television.

We also have seen pretty much conclusive proof, in the technical community in the last year that if, in fact, you only put progressive signals into your encoder, it actually has a higher coding efficiency.So, if you look forward to this question of what is the business of the broadcaster in the future, and you determine that it’s other than uni-channel video-only services, then you will become more and more parsimonious about the use of your bandwidth. Each of these things ultimately accumulates, in terms of enabling you to have different businesses in the future.

Some of this is semantics, of course, but in January, Consumer Electronics Manufacturers Association agreed in their own statements that they would consider 720 progressive to be a high definition television format.That was also an interesting time, because two announcements outside the world of digital terrestrial broadcast also happened, which have an interesting impact and analysis on the relationship between the businesses that the direct TV people are in, and companies like TCI, and how their decision making process differs a little bit from that of the terrestrial broadcaster.

Thompson and Direct TV chose to announce that they were going to offer, later this year, a high definition service.But, they didn’t choose any of the formats that were in the ATSC standard that’s being used for digital terrestrial.And Microsoft and TCI announced, in concert with General Instruments and several others, that TCI had chosen Windows CE as an operating system for at least 5 million of their upcoming advanced set top boxes.But, they also agreed that these HT-0 formats, this subset built around primary 480 P, would be what they would focus on for their digital TV.

And here I think that these companies are proxies for the consumer to some extent, particularly TCI or any other cable company who has to put the set top box on their balance sheet.They become incredibly sensitive to the incremental cost of adding features.They also found — given that they felt they had a fixed amount of money they could finance in a set top box — they had to make a cost trade off, between how much you would spend, if you will, to improve the video service, and how much you would spend to get the computing capabilities that they thought were the enabler for the other services that they view as part of their future.And in that fairly pure environment, and outside the political context that has wrapped around the question of digital terrestrial broadcast, and the spectrum allocations for that, you see a fairly direct decision to optimize in a little different way, between video and computing capabilities in these boxes.


a lot of the work we’re currently doing in high definition television had its roots in some of the analog high vision work in Japan.And it was actually only the beginning of early last year that the Japanese government agencies finally did indicate that, in fact, analog high vision was probably going to be eliminated, and they would, in fact, go to a digital television system.

The remainder of the year was interesting, and in February of this year, they formalized, in fact, five formats that would be done

— that would be used in Japan now and in the future for digital television.Which not only included the conversion to digital transport of legacy 480 interlace, and of course the carrying forward of the 1080 interlace stuff, that had its heritage in high vision, and the work here in the United States, but also 480 P, 720 P, and even 1080 P, on a prospective basis, as the basis of digital TV in Japan.And this covers all of their transport systems.

More recently, Sony and Panasonic announced support for 480 progressive broadcast equipment, saying that they would be, in fact, making this readily available.And interestingly after more than a year of review, the U.S. government video working group, which covers setting specifications for video usage throughout the U.S. government activities, in both military and non-military applications, has promulgated a standard that is all progressive.

So as we approached NAB this year world was, in fact, changing a bit.We feel good about the work that the DTV team collectively was able to do, in concert with many other people.And we are starting to see, completely independent of any business relationship that any of us have with these companies, a lot of decisions on the part of broadcasters, certainly not uniform, but the whole intention is to get the FCC to agree, and the broadcasters, and consumer electronics companies to agree that the marketplace would set the video standards,to allow this choice on the part of each business.

So Fox has now chosen to go 480 30 P, at least for their introductory digital TV services.ABC announced last week that they’ll do all progressive in 480 and 720.Yesterday, KCPS, which is a PBS station in the Seattle area has indicated two things that I think are interesting.One is that they’re going to do progressive only transmission.But, also that working with an affiliate company, they’re going to start to do programming for these Web TV Plus boxes, focusing in the public interest area.And I think that, particularly with the uncertainly around the continued government funding of public broadcast television, they’re taking steps to become more self-sufficient, and to help to promulgate both high definition programming, for which they’re quite famous, and also some of the interactive programming, which I think they aspire to equal fame.

CBS and NBC here last week had indicated that they will, in fact, support 1080-I, for a lot of their programming.But, in fact, their commitments, at least toward the end of this year, when they go on the air, are quite minimal.I think they both indicated, you know, single digit hours per week, or something in that vicinity, of actual 1080-I originated material.And of course, Sony, Panasonic, Phillips here at the show are all showing a full range of equipment in all of these formats.

So it’s pretty clear that we are facing a split decision, if you will, as to how digital terrestrial television services will be inaugurated.But, it’s pretty clear that people are choosing different business strategies.And ultimately different balance points in the use of their spectrum.So when we define digital television today, we think of it as really a three-part program.One, clearly, and obviously, has to be high quality video and audio.And if you haven’t seen it, we encourage you to go see the demonstrations.We have some in the Microsoft booth, and clearly there are lots of other ones around the show, where you can look at even side-by-side comparisons of these.

It is clear that even going up to 480 progressive represents a pretty dramatic step, in terms of the picture quality people see in the home, as a delivered video system.And as a result, no matter what choice a broadcaster makes, they’re clearly going to be giving people a much better television picture, and audio system than that which they have known for the last 40 years.

We also think, though, that the future of television is about using this appliance for information access, and communication, and the ability, ultimately, to blend those together into a new form of produced television experience, which includes interactive content.Many people talk about interactivity and they fear it, to some extent.In fact, over the last four or five years that I’ve been involved in Microsoft’s efforts to work toward a new form of television, people always told me that they were

— there were two things that they didn’t want to do, they didn’t want to reboot their televisions.And we’re pretty clearly aligned with them on that.We wouldn’t want that to happen either.

But, the other thing they always said is, hey, you know, I want to come home, and I’m tired, and I want to be entertained.So there’s all this talk about having computing intersecting my television, meaning that I somehow have to be sitting on the edge of my chair all the time, and interacting.And, of course, the answer is, no, the benefit of having intelligence in the receiver can be just as powerful as

— I’ll show you in a minute, enhancing the produced, but passively enjoyed, television experience of the future, because it allows the ability to do personalization.

I think we live in a society today where people more and more are expecting personalization of services, and I think that it’s only true that, with the combination of data broadcasting and video broadcasting and intelligent receivers, the content and broadcast community is going to be able to play to this trend.And, of course, you do have the opportunity then to provide interactivity, which could either be locally done in the device, or if you have some type of connection to the Internet, then, in fact, it could be a more complete experience.

One of the big questions, and I’ll come back to this again later in my remarks, is how is the public really going to adopt to this world of digital television?There are really two premises that are put forward.One says I can take a high definition television, put it on the air, and say that that is compelling enough in its own right that we’re going to get people to go down to the store, spend quite a bit of money, and upgrade their television system.

The other premise is really one that has its roots more in the Web TV class of products, where we’re saying, no, the consumer will find more value, ultimately, in a balance between the communication, interactivity, and video services that are possible in these televisions, as long as we don’t overwhelm them in terms of the cost of doing that.

It’s also interesting to look at this chart, which shows some long-term adoption cycles, or rates of adoption of different technologies, as far back as the 1800s, you know, where the automobile came into play.Here, what you’re looking at is the percentage of homes in the United States that have taken up this technology, and have it present in their households.PCs here, after just less than 20 years,have now actually just crossed over the 40 percent mark.Cellphones, down here, are pushing 25 to 30 percent, after an even shorter period of time.One of the things you note is that there is very, very little adoption of technologies for fairly long periods of time.And then all of a sudden, the thing becomes a phenomenon, and as the volumes go up and the costs get driven down, then you see a very, very rapid increase in the acceptance of this in the population.

Another thing to note is that the population is becoming more technology literate, or at least more accepting of these technology changes.And, in fact, some of the more recent things that have been introduced have actually had much shorter periods of time between the time the technology is first introduced, in sort of any form, and the time that it begins to go up this ramp toward population adoption.

And we think this is a critical issue, particularly when we recognize that television isn’t a new medium.It enjoys virtually 100 percent penetration.Personal computing is still on the rise, and particularly when put in the form of these appliance products, we expect it to continue to rise quite quickly.And so the question is, what’s going to happen as we convert the personal computing experience and the television experience, both important and popular phenomena in society today, into this new thing that we call digital television, to supplement what we know as the world of computing?

So the question really is, in our mind, there is going to be a family of digital television receiving devices.And, in fact, this simplified chart shows, in a way, the way we think about it.At this end of the chart, there’s one approach that says, we’re just going to have a television, like the televisions we’ve always known, in the sense that they aren’t computers, they aren’t connected to the Internet, they have really, really nice display systems, and they provide this high definition television experience.And this, if you will, has been the plan of record in the consumer electronics area for quite some time.And, in fact, little focus, virtually no focus was given to the question of data broadcasting, or the integration of computing into these devices.

On the other hand, the computer world, you could say, was over here, starting at the other side.We had the computers.They were already becoming very, very capable multimedia devices.And with this release of Windows 98, and in this calendar year, the release of new PC home products for the residential market, you are going to see televisions that are

— or television services that can be had in the form of a personal computer.Now, that device can also play video games, be a DVD player, connect to the Internet, and run Word, Excel, and PowerPoint.So it’s essentially the uber appliance.

Now, we don’t believe that every appliance, just because it can be emulated by a personal computer, should be replaced by a full personal computer.So that’s why my division, for five and a half years at Microsoft, has been trying to create the computing architecture and the software to go in these low cost dedicated appliances, that would have to essentially work to provide these services more along the lines that people already know and love.

So the real question ends up, how do these two things move to meet in the middle?And you can say, there are really two approaches.One is to think, if you ultimately do come to believe, as we do, that computers should be part of the equation.If we’re going to go out and tell the public that they should upgrade their television system in the home, we think, of course, that they should be buying some computing in the process, not in the sense that they’ve traditionally bought personal computing, but more in the sense of the kind of things you see in a Web TV product.

So then the question is, do you propose this as something that’s integrated into the device, and at the outset propose that it has to be purchased along side the high definition display capability or, in fact, do you believe that there is some type of advanced set top box, for whatever transport, including the terrestrial transports, that allow people to get this blended service, better delivery just assatellite represented better delivery of analog television, and the new cable systems represent digital delivery of analog television.Certainly, even that step represents a big step forward, compared to what we’ve known in pure analog delivery.

And so the question is, what is going to be the strategy of using set top boxes, in order to get the public to embrace this combined capability of computing, information access, and entertainment, all through their television appliances.

So, let me stop now and give you a couple of demonstrations, and I’ll start with the newest capabilities of the WebTV product.What you see here is

— this little black box right here on the stage with me is a WebTV plus receiver.These boxes sell at retail for $200, and basically you can either buy them, for example, in concert with MCI if you want to use them as your Internet service provider, or WebTV will provide it directly.So, you pay a monthly service fee that ranges from $19 to $24.95 a month for

which provides all of the Internet connections, all of the TV services, the electronic program guide, andimportantly, the maintenance of the box from a software point of view.

I mention that because while this box is the same one that people have been buying since the end of November, in fact, the software I’m demonstrating is actually a beta test version of the next release of the software that we will download into all those existing boxes.So, probably in June of this year, some of the new features that I’ll show and mention here today that are not part of the product people bought when they took it home, will actually one night be downloaded into their boxes without them having to do anything or worry about it, and they’ll wake up the next morning with an introductory piece of email, or video mail that will actually explain to them the new features that they got in the box through the software download.

Now, what you actually see here is the local cable service here in the convention center being fed directly into this box and, as you see, we have here what we call the TV Home Page.This complements, and I’ll come back to this, what we think about as the Internet home page.What you see here is actually the original WebTV capability.You navigate with a remote control.The big yellow box indicates what is actionable at this point.And so, anything that the box will stop on as you move around is something that you could, in fact, tell it you want to go do.

The system provides electronic mail capability, and to show just another way of thinking about the use of broadcasting and video bits, here’s actually a piece of video email that was sent to us by one of the guys in our Japanese subsidiary.So, I open up a standard piece of email, there’s a little piece of video clip here, and what you’re looking at here is essentially low bit rate video.

To put this in perspective, this is 180 kilobits a second that is software decompressed in this box using just a standard microprocessor.And so, while no one would say, hey, is this high definition television, the goal isn’t that.The goal is to begin to take a step towards the day when people will be able to send messages back and forth to each other, not because they’re expert with the keyboard, but because they can sit in front of their television, have a little video camera stuck on the top, or their camcorder on the top, and in fact that feature, the first step in that, will come out with this new version of software.

There’s a video input in the back.You can put your camcorder in there, and you can essentially use this to take video stills, embed them in email, and send pictures of the kids home to grandma.So, this is not science fiction anymore, this is the kind of thing that will be happening.And increasingly, in probably less than a year, that type of low-bitrate video could actually be captured and delivered even with only the phone wires as a way to do the two-way communication.

We have the ability to store favorites, and I’ll start one of these and come back to it in a minute.Let me just go to some personal favorites that were put in here.So, here we have pages that

they’re little thumbnails that represent the places that we’ve stored.Let just, for example, go to the MSNBC page.So here, while this goes off and sucks that page down from the Internet, we’ll go back to the other side for a minute.

Here what we’re doing is digitizing the video on the fly because we have only analog television coming in.And, of course, if I just say, select that, I go back to full screen video.I have the capability to channel surf.But here you see the introduction of new capabilities in this environment, where we have the program guide built in.We can essentially do translucent overlays on the screen.And, in fact, if I go up here to this channel, this is a prototype that we’ve developed with people in the baseball environment.That little “i”s pops up, that means that there’s actually some automatic content that is available through the Internet side that corresponds to what’s happening in this game.Now, this is obviously a prototype, but it shows the kind of things that will become possible.So there’s a little piece of data that’s downloaded.It says, oh, there’s actually another menu that’s available.And I can select on that and get a video overlay of the batter and his personal statistics.

So, these kind of things are all possible by combining the ability to do compositing of new services in the box on a personalized basis.It’s only done really at the discretion of the viewer, but it gives the ability to have many, many new capabilities that really just weren’t possible before.

I won’t wait and show it now, but, for example, another thing that happens would be, perhaps, particularly interesting to this audience, and maybe we’ll come back and find it in this tape in a minute, is the same kind of capabilities are available to advertisers.And, in fact, there’s one ad here I know on this tape that was developed, a prototype with Southwest Airlines, where, in fact, when the Southwest Airline ads comes up, if you have this WebTV type of receiver, that actually puts the video of their ad spot

— positions it on the page at about three-quarters of the screen, and then essentially provides other interactive capabilities in real-time direct for the advertiser’s benefit.

And so, we think that these kind of things represent a tremendous opportunity not only for the content developer, new services and forms of entertainment for the consumer, but they also represent, essentially, a new world in terms of the advertising environment.And we’ll come back and show you just a couple of other examples of that in a minute.

Here we have a number of capabilities, we can keep track of favorites, so as we talk about a world, particularly where there are multiple transports, for example, terrestrial and satellite, in the product that we’re developing, they can be integrated into a single guide.And, as a result, you break what I call the
“six remote controls”
problem, where if you have multiple video sources, you ultimately have multiple remote controls and guides.So, we’ve been working on a canonical way where the people who develop the guide data and programming can essentially put it in a single system, and the consumer can seamlessly navigate between all those different forms of entertainment.And so, if I would just select that out of that, it would immediately go back and play that capability.

If, for example, I go back to this other side here, for example, I pull down this MSNBC news home page. You can scroll around in this environment to read the headlines that are on the web page at this point.The thing that we couldn’t do before, which we now can do, however, is, we can use this digital video capability, even though we’re digitizing analog to get it, to actually have the ability to composite the video window on top of the Web page.So now I have the ability, even if I want to do it manually, (in other words, if the programming environment hasn’t yet gotten to the point where they’re going to provide this kind of linkage to me as an automatic service,) I could decide to go to watch NBC news, and then read the Web page at the same time.And this is, again, at the discretion of the viewer.

If I go back to this side, and just show you a few other things, one of the things that we have behind every program is essentially a Web page that not only contains the basic guide information, but in fact provides two capabilities.One is single-button recording, so this device is capable of controlling an external VCR.So, you can pick any program out of the guide and say, record it, much like people do with VCR plus, but here there’s no number to enter or anything else, the linkage is automatic.

You also have a reminder capability.So, in fact, as you go through the guide and this product has seven days worth of guide data always available instantaneously because it comes down through the telephone wire at night while you sleep, and so you can essentially have what we call a TV planning capability, where if you ask for reminders –I don’t think there areany in here now — this TV planner would essentially be the list of programs that you’ve identified for yourself that you want to watch.And you can set them up so that they either interrupt you to watch them, or just give you a notice that the programs you were interested are available.

So, there are many new ways to get people involved.Here’s another example where, we recognize that while people are begin entertained, they don’t want to necessarily always immediately follow these links into the interactive environment.And so we have the ability with a single push of the remote control, when one of those little “i”s comes up and says, oh, there’s some interesting thing behind this, you can just push the button and have it remember it.And so, this is essentially a list of the TV-related Web sites that the broadcaster sent to you embedded in the vertical blanking interval, and which you can at your leisure go back and collect, and you’ll know what show produced that link, and you can just push on the button, and away you go.You’ll be taken to the Internet to watch that capability.

There arelots and lots of other features, and time is short, so let me just go on and show you a few other demos.As we move toward the digital TV implementation of this, which we have today for analog terrestrial and basic cable, and which we hope to have for satellite and clearly there will be cable-like products like this coming within less than a year’s time.But this is the kind of thing we think of as digital television.It just happens to be minus the really good digital picture right now.

So, let’s switch now to the demonstration of Windows 98.And I’ll leave that there for just a second.If you’ll look at the big screens now, what you actually see in many ways resembles that electronic program guide I just showed you on the WebTV Plus box, but what you’re actually looking at here is the release candidate that’s in beta test now for Microsoft’s Windows 98 personal computer operating system.And you’ll see that it operates in much the same way.I happen to be using a mouse to do it, but we also have remote controls that allow you to do this with a PC from across the room.And, as a result, it doesn’t really matter, in our mind, whether you have a small screen PC or a big screen PC, or a WebTV box.At least in our products, we’re trying to do everything we can to make the television viewing experience, and the types of programming that you could enjoy as similar as possible in this environment.

So, we call these features WebTV for Windows because, in many ways, we’re trying to make them as similar as possible for the user.So here, as I click around, I could watch any of these programs.And, in fact, let me click up here on this one.And let’s go ahead and run something that you might find interesting.Here we’ll start a tape that actually puts an ad together.A number of ad companies, in this case Colby Eppler (sp), has been working with Suzuki to develop what might be an interactive ad.So, I’ll go full screen and we’ll watch this ad.

So, here

— the video will start in a second

— this is essentially a broadcast video spot that can be done over analog television, and the data necessary to support the rendering of the Web page that you see they’ve wrapped around the video actually comes through the vertical blanking interval.And, in fact, it’s being transmitted now, received by this personal computer, and these graphics were essentially rendered by the PC locally, based on the data that was transported alongside the video.

And what you see here is that I could sit here and never do anything, and just watch this ad.And I have these little motorcycles wrapped around it.But, in fact, each of these things represents an interactive button that I can push.And if I push it, then it shows me in this case the control panel of that particular model of motorcycle superimposed on the video, and it gives me some very basic information about that particular bike.I could go down here and look at the GSX and R-750, and get a similar kind of presentation.

So, more and more we are finding that advertisers are intrigued not only with the ability to provide more entertainment, but to provide interactivity, and ultimately in devices that are really connected it gives us the ability to get immediate feedback from the consumer.

In addition, it’s also clear that the data provides a way to personalize these ads for individuals, because not every individual has to see all the same information.So here, for example, if we click on this “i” you’ll see another ad.This one has been under development for Toshiba Corporation, where we, again, have buttons that are composited on the screen locally in the computer.The same thing would be done in a WebTV box, and the architecture for how this transport works is the same in all these products.

You can essentially click on this tab, reduce the video, you only have two choices, big or little, in their design.But if you reduce this, then, in fact, you get an interactive Web page.It provides more detailed information about their DVD line of products.You could, in fact, click on any of these links.And, if you’re in an online environment, like WebTV or a PC that has connectivity to the Internet, then, in fact, you could immediately go to the Web page.You could, in fact, have a store locator to tell people where to go.You could do an online transaction to show how to buy it now.Or, in fact, you can continue to read more information.

So, these are just examples of how the advertising community, rather than thinking about the potential problems that interactivity and intelligence represent, is looking forward to try to figure out howthey can take advantage of these things and deliver more interesting advertising,more directly to consumers on a targeted basis.And, as the world becomes more connected, take direct advantage of that.

Let me go and give you one last demonstration here of some actual digital television.This, by the way, next to me is a Fujitsu flat panel display.It’s a 480-line progressive display that has a 16-by-90 aspect ratio.What I’m going to show you here are some demonstrations that show, again, how

— the kind of capabilities

— this video will start, again, in just a second, it loops

— what you’re actually looking at is fairly rare today, not only in that it’s actual native 480 progressive originated material on video.We went to a zoo with a camera and took these pictures. I hope we did.The capabilities that are represented here are essentially coming from compositing graphics on top of this.

There it goes.It was at the end of the loop.One of the things that is interesting about this is not only that it was a complete 480 progressive chain, all the way to these displays, but it’s the first time you can see the effects that start to happen when the native graphics in these devices are rendered at exactly the same resolution as the display.

If you go back and think a minute ago about the WebTV demo, we were completely limited by the resolution of the NTSC display.And even in a high quality environment, you could clearly see the the resolution of the analog video; it was always obvious where the video started and where the computer graphics started, or ended.

Here’s an idea

— I’ll just give you an example of how someone might produce a program, perhaps along the lines of a nature program, or an educational program, and using my remote control I can essentially turn on interactive panels that essentially overlay on the video.And what happens here is you can see that the graphics quality and the actual native video quality are now the same.And so for people who want to produce this kind of environment, where you can navigate between the different capabilities, let’s say go to the land-based version, all of these things become possible by sending data down to the computing devices in either a WebTV or PC-like product, and then allowing this type of local interactivity.

So here, even if you aren’t talking about the Internet, if you wanted to send educational programs down

— here, I’ll get rid of this

— so this video is very, very high quality.And even blown up to these 20 foot screens, many people are quite pleased with this level of capability.And it was these kind of things that I think have led at least Fox and ABC to decide that doing a tremendous amount of their programming in 480 progressive, and then conserving bandwidth in order to provide these type of other interactive, or entertainment oriented services,may be, in fact, a great way to get into the digital TV era.

So let me just return to the PowerPoint slides, and make my closing comments.The business of digital television then is really going to be a two-step process.One is, how do we move from these intelligent analog receivers, that we seein things like WebTV, and this year with Windows 98, into this world of full digital receivers?We expect that our OEM companies, like Sony, Phillips, and Mitsubishi, and others that are building WebTV systems going forward, both as set top boxes, and other devices, will produce digital versions of these in 1999.The TCI order for 5 million set top boxes that do all progressive to be deployed early in 1999, I think, will continue to build momentum for people who produce this kind of programming.And we expect 4 or 5 million Windows PCs to be shipped every month, supporting that.

So as you look forward to a world of digital television, I think there are new opportunities for content creators and advertisers, in order to provide this kind of interactivity, for example, associated with sports programming.The advertiser has new places in which they can do advertising.They have new ways to reach out and touch the consumer.They can essentially deliver personal advertising, because in fact, you can deliver either all or part of that ad, in concert with some other non-video data delivery system.

And we have the ability now to have transactions, where the consumer can interact directly, to order a ticket, or buy a jersey, or whatever might be appropriate in the context of the program that’s presented.And, in fact, even for digital terrestrial broadcasters it may be possible to contemplate a set top box environment that would allow them to offer not only their traditional free to air ad-supported product but, in fact, to have a form of subscription based programming as well.And as such become more competitive with the cable and satellite offerings.

So when we look at this question of digital television receivers, and the strategies that range from the personal computer, which we think will apply in specialty cases, both in the home and in the office and school environment, you contrast that to this question of just launching against high definition with receivers that don’t have an architecture for computing, and we actually think that ultimately the most important thing in the next couple of years is going to be low cost, advanced set top boxes, for cable, satellite, and digital terrestrial broadcasts, and that that will allow you to do all the things I’ve shown here in the WebTV and PC kind of environment.You’ll be able to do it in the context of high speed digital data broadcasting, and supplementing your choice, essentially, of high quality digital video.

And so the reason that we’ve been so focused over the last couple of years on this question of what’s the right format to choose to go into the compressor is because, in fact, if you’re worried about selling set top boxes, like this one down at the $200 price point, or the $300 or $400 price point for full digital TV capabilities, and building a bridge between the world of the display people own in their current television, and the future world where they’ll make an investment for some nice, big, new, fancy display, the set top box that can essentially output to either of those turns out to be the low cost way to get the consuming public involved in the experience of digital television.And that, of course, is what we are really the most interested in.

So at the end of the day, the challenge each of you face, whether you’re operating the business of broadcasting, are an advertiser or a content developer, is how do we collectively move to DTV adoption, which way is it going to be?Are we going to essentially go on pretty much a slow track, because it isn’t a compelling enough experience for people to go out and spend a lot of money, and we start with very high priced offers, and gradually work our way down?If you lived in a world where everybody was just a digital terrestrial consumer, you might be able to live with that strategy.

It’s our belief that, in fact, you’re in a highly competitive market against satellite and cable delivered video.They are clearly moving forward with these kind of set top box solutions.And in fact, we think that the consuming public, courtesy of personal computers and the Internet and low cost devices like this, are in fact prepared to go on one of these very short cycle, very vertical ramps.But, not, in fact, if you aren’t giving them the prospect of all this interactivity, and not, in fact, if you’re not giving them the prospect of getting started for a couple of hundred bucks.

Even if you believed that people were going to go out and buy six and seven thousand dollar televisions in the next couple of years, or even if that was halved in price in the next five years, there’s still the problem of the other 2.5 televisions in every household, and how do you get them to be something where you can experience your digital television offering?

The answer is going to be set top boxes, in one form or another, and that’s why the format question has been and will continue to be an important question going forward.

Thanks a lot.I appreciate your time.