Interactive Television
Project at AT&T
VINCENT GROSSO

I have a very simple message: I would like to share with you some of our knowledge and experience at AT&T, a very big technology and telecommunications company in the US.

I work for a very small division that is examining opportunities in the home. I want to share some insights, so you can begin to think about what the implications of the emergence of this media are for the home. I feel I'm talking to the future leaders of interactivity, whether it involves the television, computer or something wireless. Essentially, all of you are here to learn. I'm very gratified by that.

I'd like to talk about `the road to discovery' When people ask me what I do, I say I'm in the learning business.

T h e . C h i c a g o . T r i a l
We went out in New York City and asked people what they thought of the `information superhighway'. What and where is it? The term is a rallying point in the United States, involving legislation, point of view and influence on the home, among other things. We discovered that society at large doesn't really know what the information superhighway is or what to expect. We fight this problem all the time. A couple of years ago, when I started looking at interactive television, we decided to ask people on the streets: what would you do with your television set? How would you use it? Could you imagine playing a game with someone across town? Can you imagine shopping? People didn't know what to say. They idealised their behaviour. They said: I'm going to use it for information, for health related issues. We just didn't believe them. So we did a trial in Chicago. We created content of all types: information, shopping, games, all sorts of things. We put them in homes. We mailed cameras out to people and had them take pictures of themselves and send them back to us. We put them on the system used them to identify people. We had an interesting topology that created content in New York City (with about one hundred designers) and transmitted the content over T1 Carrier to Chicago. We believe that that's how this is going to work. The brain will power sits where RAM power is and networks will move the pictures and sounds.

The first menu included various options--sports, weather, news dates. We were looking for something called a `killer application'. Was it news? Weather? Sports? What is that one application so powerful that people are going to want it in their homes? After the first nine weeks, usage started to drop. We thought to ourselves: does this mean that only ten percent of the homes will use this for an average of about seven minutes?

We started with a category called Sports Scan. They looked at fourteen teams and those fourteen team scores were updated every three hours. We doubled it to twenty eight teams and nothing happened. Usage didn't go up. Then we decided to update the scores every hour. Nothing happened. At this point, the very structured researchers got into an argument with the very unstructured designers. The designers said: this is boring, let's add a game. They added something called a rotisserie game, a very popular game in which one can buy individual players with certain statistics and put them together into a fantasy team. Usage started to go up. Then people started to call us and complain that they couldn't trade players easily enough. They asked us to provide a communication function. We did that and usage went up again. Then people began calling to ask why there was no place to simply buy tickets, when people were trading tickets to games on the communications part of the system? So we added a transactional function. Usage rose. We invented a rule: like Sports Scan, all these categories started off basically as information. We added communications (the Buzz) and transactions (Rotisserie Baseball).

Upbeat was a place that we sold CDs. You click on the CD and you can listen to twenty seconds and decide whether you want to buy it or not. It started as a transaction. We added a communications piece: let's talk about that CD, is it worth buying? What's the performance about? We added information, critical reviews--what does the press say about the CD? We added a music trivia game. Usage rose again.

Something phenomenal happened. We started out looking for the `killer application'. And we found out what I call the `rules of engagement': in order to engage the public that doesn't perceive boundaries on what it wants to do, entertainment, communications and transactions must be combined in successful applications.

So our definition of interactive tv is: the entertainment of the television set, the transaction and communication power of the telephone and the information processing capacity of the computer. It's a new consumer experience without a single killer application, but with only killer attributes. When we said that about nine months ago, people were still saying they could find the killer application. I'm beginning to believe there is no killer application.

Another thing that has emerged is our hypothesis that the television set is not going to merge with the computer. The infrastructure we call the super highway is going to make the television set very smart. It is also going to make the computer very smart. So both the television set and the computer will be more interactive and allow you to receive pictures and sounds. Some applications will migrate to the computer: those that require more text and downloading. Those that require entertainment, visualisation of moving video and socialisation will happen on the TV set. We found that interactive television is not a task medium, but an entertainment medium. Everything they saw on the television set had to be entertaining. When we put too much text on the screen, they said: take the text off, give us more pictures and sounds. A interesting thing is that interactive television is a socialising rather than a privatising event. We offered several simple games: whole families would sit in their living rooms and play whole other families at seven o'clock at night. It was unbelievable.

Our hypothesis going into this was that television is a privatising event. We were surprised to find out that this was not the case. We were surprised by most of what happened. And that is the whole point: you have to put these services in front of people and watch them use them in order to understand anything about interactivity.

At the beginning of the trial, we did ethnographics. We put a psychologist in the home to watch their behaviour and we measured the distance from the front of the chair to the television set. It was fourteen feet. And at the end of the trial, the average was fourteen feet. They never moved the set. They didn't change their behaviour around the television. Behaviour around the television is different from behaviour around the computer. Ease of use was a major factor (the remote control). One thing that I couldn't emphasise enough: it's got to be easy to use. At the end of the trial, we collected all the books. Forty percent of the books still had the shrink wrap on them. We asked them how they had learned the system. Over half of them said: my children taught me. Seven-, eight- and nine-year-olds were teaching their parents how to use interactive television. The rest went through it hit or miss. As concerns content providers, they don't really care about technology. They want the most engaging applications. As designers you will translate that to your respective communities and stake-holders. The look and feel on television is nowhere near the computer.

What we found to be important in consumer needs was ease of use and updating content. This is important. We take it for granted on the Internet that you can update text quickly. In this world of the interactive television set in the home, you have to update pictures and sounds and they have to move across networks as fast as ASCII text. It's simple to say, but difficult to back up. We're talking about a world where you digitise something and can send it as a file. It can become a part of an informative or a shopping application. It can be modified. I think that all of you instinctually understand that, but the network behind it has to move the images for you and give you choice control and convenience, communications and emotional closeness; personalised service with the over choice which can happen when the television set is your gate keeper to various services.

I want to talk about the idea of the personal assistant and about entertainment and fun. If it's not entertaining and fun on the television, it doesn't happen. There is also a connection to work at home and mobility.

We're going to take some of our learning from Chicago and do a more robust trial in which we'll use moving video in various planes. We're going to have greater content access. And we're going to look again because we believe that you can't believe what people tell you until you watch them.

In the final focus groups in this Chicago trial, I asked people: well, what did you use? People said: Well, you know, I spent a lot time in news and entertainment and information... (while I could see that the person giving this answer is the biggest game player in the whole group). They idealise their behaviour before and after the test.

C o n t e n t . f o r . I n t e r a c t i v e . T V
Interactive television is a social event. This is an important concept. We thought it was a privatising event. The television is a social instrument--a social appliance. The home is a social place and the television has a very special place in the home. It's usually in the living room, a place for a common experience. In the combination of entertainment, information, communications and transactions, we found that communications is very important. That communications component started off as text on the television set. Text is imperfect on the television. It is quickly going to lead to voice and video telepathy on the television set. We're going to start with interactive applications with a communications component that is very difficult to use--keyboards that sit in your lap. And that's rapidly going to turn into the remote control which is going to be the telephone. So you're going to download voice and interact with voice on the television set in a variety of ways.

This is especially interesting from the point of view of telephone companies. Telephone companies only know the end point model of perfect pairs. The perfect pairs model is: I know you, am going to call you and we are going to talk to each other about content which is derived on either side.

In an interactive program about dinosaurs, there are many children interacting with those dinosaurs. That is the communication component. A child could meet another child in another part of the town or country. If you are a sports fanatic in soccer, you may find a community around a soccer game or a soccer experience. You can see others experience it with you. Content has created a relationship. For those of you on the Internet, this is obvious. In the news groups, you create an experience with someone you've never met. You share something special and have a social connection with them. On the television set in the mass media this is going to lead to an enormous swelling of communications. If there's anything I want to leave you with, it is that there are two forms of engagement.

I've showed you that entertainment, information, communications and transactions must exist in the application at the same time.

I'd like to give you and example of an interactive tv server that is actually manipulating video elements. It's right off the television set, so it really works. It's demonstrative of the kind of work that we'd like to do in Castro Valley, California. This interactive server offers the new movies on demand. This is what getting a movie will be like. You can order, browse or play a game. We're going to browse and find categories of viewer ratings (a form of communication). People download their own points of view. Before you make a decision, you can see Actors, Directors, Genre, Reviews. Once again, there is lots of moving video and very big text. If you pick a movie, you'll get a movie in real time and have control of it. Not only movies, but educational applications are also possible. One application for children is called Clicky Corners. It takes the informational processing of a computer and combines it with the wonderful graphic processing of interactive television and the computer.

Children can send electronic mail. This is designed for children from ages four to ten. And this is actually working. Lots of moving video is being typed on the television set. Social communication between the parent and the child and another parent and child. They can make an appointment and talk about food. This is real. The background is constant moving video.

We also have an interactive game for adults that includes video telepathy. It's a play-along. It opens with an animation. You will see the opportunity to have video telepathy. This is a game where you have to answer questions very quickly against someone else in the network. This is a video telephone experience. The video telephone can stay up during the whole experience or we can take it away. If we take it away, the representation of your brain changes according to your answers. If you answer before your opponent, your brain gets bigger. If you don't, your brain gets smaller. The point is the rules of engagement for interactive television: moving images, less and bigger text, sound, engagement.

For the people now, television is a part of their home. We have to think about how to make valuable programming for the home. The technology is neither positive nor negative. We need to give it value, rules of engagement: form, colour, motion and sound. Motion pictures are very friendly to the human mind. Full, high resolution thirty-five millimetre film is very easy to take. Next comes television with its smaller screen. Then comes the CD-ROM. The CD-ROM wants to give us motion. It gives us sound and lots of form. At the bottom of the pyramid are the early PC'S that just gave you form.

Interactive television fits somewhere between television and motion pictures in its ability to engage and to convey pictures and sounds to the individual.

What we're really talking about here is the convergence of networks: ATM (a coding system for digital bits which can become images or sounds); an interactive network which runs over fibre coax to your home through your cable system; the Narrow Band Network that delivers your telephone service and a cellular network which delivers sound and voice to you anywhere. The connection of these networks will allow you to request information anywhere. The cellular phone will give way very rapidly to the Personal Digital Assistant. Personal Digital Assistant will allow you to see on a little screen whether you can make an appointment with a friend; order a movie where you can spend some time with a friend; have information sent to you; ask for information on the cellular network--it can be sent on the television set because it's in video form. Or it can be sent on a computer on the Narrow Band Network. But these networks are rapidly merging. The computer is already connected to the telephone network. The tv is connected to the fibre coax network. When the fibre coax network starts to move data, you'll be able to connect your computer to the fibre coax network and access the Internet much more quickly. You'll be able to get pictures and sounds much better on your computer. As I said, pictures, sound, and data combine to form applications to define their respective networks.

What is advertising going to be? Television advertising today uses thirty-five millimetre film, wonderful audio, excellent actors and actresses. In a world of interactive TV, advertising is going to use all of those attributes: form, colour, motion and sound--and it is going to be very difficult to tell the difference between an advertisement and an application.

How do you make content in an interactive world? You digitise it to M-tech and create navigation. It becomes an application. You then check it and send somewhere to be sent to a series of networks. That is AT&T's business, but that's not important. The important thing to note about broadband interactive television is that the server has to be very close to your home. That's a big difference with the Internet, for which the server can be very far from your home. In the broad band world of interactive television, servers are going to be very local, with very regional content under your particular control. Which means designers like yourselves are going to have to educate others to create content at the local level. And that content is going to happen on PC's. In the United States there are going to be regional production centres that are hubs which fed by thousands of individual PC's and others creating content. You will be able to create content on very inexpensive desktop machines for servers that are locally deployed.

As consumer choices increase, you'll need to have more control in the home in order to be able to filter the things that you want and don't want. The role of the personal assistant will probably be very controversial. This event-driven `spirit' in the network will go out and get you financial news, important information about health or whatever you want. It will reach into those three networks: the broad band television network, the narrow band network of the computer and text and the wireless network. It will allow you to get what you want. As a consumer, a viewer and a person in your home, you must be able to do that. There will probably be others who will want to program personal assistants for you.

Today, voice and ASCII is standard. Tomorrow, video will travel as far as ASCII. As those networks emerge, images and sounds will become a language. Those of you working in it now realize that you can take an image, paint it and send it to someone else. Images from around the world will soon begin to converge and be edited together on the desktop for applications. Networks will get much better at moving those images. Inpeg II encoders will get much cheaper for almost anyone to encode. Images will become a language.

The home information package is going to change rapidly for many reasons that we've heard for the past day and a half here. But the home information habit on the television set is particularly sensitive, with its rules of engagement, of form, colour, motion and sound--information, communications and transactions.

The new sense of time we find on the Internet is not critical. Pictures and sounds can be downloaded and retrieved whenever you want. There is potential here for a much freer sense of exchange. Globalization is already happening inside the home with the Internet. It will happen on a much grander scale when the ubiquitous device of television is included.

The home can become the originator of content. Consider the video telephone picture: I think we're going to see that people don't want to see each other on the television set. They will want to create images that stand for feelings that stand for events that they can send. I believe the home will be a great originator of picture information as well as text information and that we will have to improve the upstream pathway back up to these servers.

The impact on the global society is fairly obvious: the creation of virtual communities. However, virtual communities will have to exchange not just words but many, many pictures because the television set demands those kinds of access. It is probably naive to speak of the elimination of geographical boundaries, what may be the potential impact of moving pictures when pictures become a language? And those pictures can be remanipulated and resent.

I was recently at the film studios in Hollywood. Hollywood studios in the next couple of months will be putting home pages on the Internet. You will be able to download Mpeg II movie clips on the Internet. My first response was: that's terrific, I'm going to take the Mpeg II movie clip, change it and send it to someone as a greeting card. And they said: uh-oh!

The impact on the television media is going to be extensive. A hundred designers and producers in New York are creating digital images for these trials. A lot of different kinds of specialists must work together. The most important thing is that these specialists are team-oriented and multi-discipline, have a balance of the right and the left brain and can think creatively. Gone are the days when a programmer can say: I'm the programmer. I'm the one who's going to do this for you. Leave me alone and pick it up in two weeks. When we meet programmers like that we tell them to go away. If they can't deal with the writers and the producers evenly, we just don't want them, because interactive television requires this interdisciplinary approach.

Content or context? With so much content coming, we're going to need that personal navigator or sense of criticism and context will become the rule.