Workshop on Emerging Opportunities
For Information Technology

Marriott Casa Marina Resort, Key West, Florida
April 30 - May 2, 1995

Table of Contents

Mission and Objectives

 Because of the explosion in computing and communication technology, many new products and services are now possible. However, many of the proposed visions for these new products and services are mutually inconsistent and some still have significant technical and social obstacles. Technical elegance will often not be the principal factor in determining success.

Many of us, in our own organizations, prepare "ten-year roadmaps" based on our guesses about where technology is going, and use these roadmaps to identify critical areas in which research is needed to stay on track. In the workshop we will identify and critically examine such issues and opportunities in information technology from a broad perspective. The goal is to consider these issues with more vision and greater technical accuracy than we each bring to the discussion individually.

This year, questions that will be discussed include:

  1. How will digital services be delivered to home / business / mobile users worldwide? What services? How will these uses affect the world?
  2. How do people today look at the world differently because of the ubiquity of computing? What new uses will be created by people who have been immersed in computing since childhood?

Final Agenda

 Sunday Evening (April 30)

6:30 - 7:30 pm, on the beach opening reception

Monday Morning (May 1)

Session Topic: High Bandwidth, Wide Area Infrastructure
Session Chair: John Denker, ATT

Monday Evening (May 1)

Session Topic: New Services
Session Chair: Rich Howard

Tuesday Morning (May 2)
Session Topic: New Services
Session Chair: Steve White IBM

Tuesday Evenging (May 2)
Session Topic: How are users and user interfaces changing?
Session Chair: Scott Kirkpatrick IBM

Speakers and discussion moderators will address these areas at a level and in terms that will permit all the attendees to participate in the discussion. We ask speakers to prepare about 30 minutes' material, and plan for questions during their presentations and during the period at the end of each talk. The "reality testing" sessions each evening are panels and discussion intended to bring out views not presented, and to identify critical events which must happen to make these promises a reality.

Reality Session Feedback

No procedings were kept of the formal presentations. However, discussion sessions and many questions from the audience kept introducing reality tests. Three of the "reality testers" have provided summaries of their comments, linked below. These thought-provoking questions are a guide to the program we hope to put together for next year's workshop.

Monday Evening (May 1)
Rich Howard's Feedback

Ran-Hong Yan's Feedback
Tuesday (May 2)
Rich Lippmann's Feedback
  1. There should have been some presentations on the impact of the Internet on the way people work, on productivity, and on how they interact with others.

  2. There should have been some presentations on how user interfaces affect how a user views an application and can use that application.

  3. A talk or two on working at home (telecommuting) would have been useful.

  4. A discussion of low-bandwidth versus high-bandwidth solutions would have been interesting:
    • Present top few videos every 1/2 hour intead of video on demand every second of the day.

    • Download interesting items at night into a local cache instead of putting up with long interactive delays (Most of the interesting stuff on the internet fits into 3 Gbytes)

    • Use LYNX net browser at home (omits pictures uses dumb terminal) instead of netscape.

  5. A discussion of a user's understanding and retention of information presented with real-time video, sound, still photos, text, or all of this but with interaction would be useful. It would help determine how much you add with higher bandwidth and more complex displays.

  6. More compete surveys of what the internet is really used for would be useful.

  7. I would have liked an update on how well groups of workers can cooperate over a long distance using high-speed communications links to maintain contact.

  8. In keeping with the conference theme I would like to know if students coming out of highschool with computer experience at home from grade school on up can use computers more effectively (solve tasks quicker).

  9. It would be interesting to know how much value added users perceive is provided by real-time video and audio and interaction and by higher resolution displays. How much are users willing to pay for these features applied to different tasks (obtaining information, communicating with others, writing documents, personal finance, programming)?

Jim Kondziela's Feedback

Operations:

  1. Operational Integration. There was no treatment at all of integrating new multimedia technologies into current, mundane (but complex) operational business environments. By that, I mean interfacing with critical billing and order entry systems, network monitoring and surveillance systems, etc. On several occasions, I have personally witnessed new projects be significantly delayed because the technology folks did not adequately consider the impact of the technology on these other systems. Perhaps we should pay some attention to these operational issues.

Economics:

  1. Upper Limit of Cost to Consumer. What is the upper limit that customers will pay for such services? Is it $100 (Ed Zurkowski), $62 (Ran-Hong Yang), or something else? Is there a limit? In the phone business, we plan numerous AIN (Advanced Intelligent Network) services, but I wonder how the customer will react to large monthly bills for all these services (talk about a reality check!) One approach might be to try to distribute the costs across a number of categories. The idea is to try to influence customers' perceptions, so that they categorize A as their telephone costs, B as their information services cost, C as their e-mailing costs, D as their entertainment services cost, etc. If the customer categorizes all these costs as, say, "telephone" costs or "computing" costs or "entertainment" costs, the upper limit that he is willing to pay may be disappointing. Can we architect in order to influence customers' perceptions?

  2. Value of Retaining Customers. Even if you charge nothing for these services, there may be great value to these services simply as a means of helping to retain a customer base. All the RBOCs, for example, will face declining market share as competition at the local access level rises. If the long distance market is any gauge, the RBOCs could lose about what AT&T lost, viz., about a third of their market to competion! If these services help retain market share, they would be very valuable for that reason alone. It is very costly to win new customers, as AT&T and others know.

  3. Third-Party Development Tools. Although the computer industry is accustomed to developing operating systems and tools for third parties to develop applications, the telcos are not. However, the RBOCs are being forced into third-party access through, for example, AIN architectures. My sense is that this is a positive development, enabling capitalism to do what it does best--allow niche developers to meet the needs of niche markets. My point is that our architectures should explicitly set as a goal development of tools for third parties to develop new services.

  4. Low-Tech Alternatives. Given the success of the Net and the Web, should we look as much for low-tech as for high-tech multimedia solutions? For example, the RBOCs are facing renewed interest in an old technology, ISDN. There is an alternative to fiber broadband, viz., ADSL, that Ed Zurkowski mentioned. Eric Sumner is clearly moving in this direction. Low-tech solutions may save on costs, allow us to deliver more quickly, and spur market demand for high-tech solutions.

Psychology:

  1. Psychology of Home Life. There is not a large amount known about the psychology of home life. For example, do people currently work and play in the same part of the house? If they don't, will we be pushing them into doing something that does not come naturally by, say, developing technologies that all operate over a TV screen? In other words, will we develop technologies consistent with or in conflict with these conventions? Personally, I think that it would make sense to develop technologies that are consistent with current human behavior in the home. (Researchers such as Arnie Lund at Ameritech, I believe, are investigating the psychology of home life.)

  2. Concealing Complexity. It was good to see efforts at concealing complexity from users, such as Rajiv Ramaswami's "transparent network protocols" and Jim Pitkow's network visualization techniques. I would encourage more such efforts.

  3. "Satisficing" and the Versatility-Usability Tradeoff. A naive and erroneous view of users is that they continuously learn more and more about how to use a product, technology, or service. In actuality, users tend to learn only as much as they need to in order to perform a desired task. In other words, they learn what they need to, and then reach asymptote, after which there is not much learning. This phenomenon is known as "satisficing," and it is quite rational because users are continually bombarded with new technologies that are ever-changing. Another, related phenomenon in user interfaces is what I have termed the "versatility-usability tradeoff." This tradeoff refers to the fact that, in general, the greater the functionality available to users, the more difficult to use is the product. Some of our biggest gains in usability occur by simply reducing the amount of functionality! Yet, we technologists and, sometimes, marketeers tend to engage in "feature inflation" through our committee work.

    Think about it. In committees, individuals may have pet features, about which they feel strongly. Others on the committee may not feel the same way about these features--indeed, they may even feel somewhat negatively towards these features--but may be reluctant to oppose the inclusion of the features for various reasons (e.g., they do not want to alienate team members, they may fear retribution against their own pet features). My point, then, is that we should always do as much as we can to fight feature inflation, especially in genuinely new technologies. I got a sense that Eric Sumner, for example, was fighting this, the good fight, in regard to AT&T's new products.

  4. Initialization/Configuration/Set-up/Getting Started Routines. We have consistently found that any technology, product, or service which requires end users to perform an initialization activity, even a very simple one, provides a significant obstacle to its usage. For example, with voice mail, an initialization routine that requires users to enter a password, recorded name, and personal greeting is typically an obstacle to usage of the service. The point here is that we should always seek the ideal case: to eliminate initialization routines in our technologies. For this reason, Yacov Rekhter's mention of "host autoconfiguration" was commendable.

  5. Uniformity & Diversity to the User Interface. A significant problem for new, distributed multimedia technologies is to maintain a sense of uniformity at the same time these technologies welcome great diversity. The kind of diversity and uniformity that seems desirable is analogous to that of automobiles. Certainly, consumers enjoy the diversity in differing "makes and models" of automobiles, yet they would just as certainly not appreciate variation in, say, the pedals of cars (i.e., changing which pedal is acceleration and braking, right or left). We anticipate that this uniformity-diversity issue will grow in importance with these new technologies.

  6. Role of Human Drives and Emotions in Technology Development. I am struck, but not surprised, by the degree to which human drives and emotions are playing a role in our development of emerging technologies. In general, there is enormous interest in the "media entertainment" value of these technologies. Also playing on human drives and emotions is the repeatedly acknowledged desire of users to communicate (e.g., recall Ed Zurkowski's amusing comment about how much people want to talk with each other: "after they shop, they want to talk about their shopping"). And the sex drive of "horny nerds" was repeatedly cited. In regard to sexual interest on the Net, I recently saw a survey which claimed that the top two Internet search topics were "sex" and "erotica" (I'm obviously naive, because I don't understand the difference). But there are some who actually claim that sex has played a positive role in the development of many emerging technologies historically, such as the development of the VCR and the Polaroid camera.

  7. Organizational Psychology Issues. These technologies attempt to bring together the telcos, computer industry, cable companies, and media industry. Yet it appears that these industries possess differing priorities. For example, it was mentioned (by Chris Heegard?) that cable companies and telcos treat the reliability of the network differently (reliability is hugely important to the telcos, but perhaps of a little lesser concern to cable companies historically). I have also seen differences in traffic "over-engineering" on the part of the telcos, i.e., being overwhelmingly concerned with adequate capacity for the network. This concern, I sense, is not always shared by other mentioned industries. My point here is that perhaps we should try to make some of these differences more explicit among ourselves, so as to aid communication and working with one another.

  8. Psychology of Aesthetics. What makes something beautiful to behold? Speaking more crassly, what makes for a "slick" user interface? Clearly, multimedia and computer user interfaces have become slicker. Take, for example, the extensive use of texture in graphical user interfaces. Does texture have any major effect on usability? No, the reason for many of these pretty features is simply aesthetic: to make the product more attractive and, therefore, more enjoyable for the user. Yet we know little about the aesthetic dimensions of interfaces. I sense that this ignorance is about to change, especially when media companies and the MTV generation are involved in the buying decisions.

  9. Expect Surprises in User Interfaces. We are likely to be surprised by the kinds of user interfaces that are developed. For example, Bloomberg Information TV, a successful, predominantly business information news television service, provides a user interface which most human factors people would probably have condemned at the outset. The screen is divided into about 4 windows. In the upper right quadrant is a talking head reading business and other news. In the upper left quadrant might be the top 25 college basketball teams listed, with current win-loss records, or a "quote of the day," etc. At the bottom of the screen are 3 lines of moving stock tickers for various national and international markets. In the middle of the screen might be weather information or currency values. All the screens are active at once, independently changing, and on simultaneously! Is this information overload or a shrewd response to their customers' voracious information needs and short attention spans? My point: Expect the unexpected in user interfaces.

Final Exercise:

  1. A Role-Playing Exercise. Suppose that you were a financial portfolio manager wanting to invest in this emerging field. Where would you invest? Frankly, I find it very difficult even to speculate, and the fact that I have difficulty even speculating means that we're all in for one helluva ride. We better buckle our seat belts!

List of Attendees

Organizing Committee

Contact Information
Denise Prull
451 North Sycamore
Monticello, IA 52310

Email: DLPrull@aol.com