some responses from Hal

Real-World Application?

Michael Otte – Has the EDC been used in a real world application since this paper was written in 2000? if so, do you have any comments from the users about their experiences?

Soumya Ghosh – The article mentions that future work involves applying EDC to real life situations. Has there been any progress on this front since the article was published?

Not as much as hoped: we did use it in a City Council/Board of Regents meeting (the feedback was positive) and are working on a Williams Village/Martin Acres project. Most of the application has been within urban-planning methods courses, to develop tools that might be used. We also have a tech-transfer project underway with the Univ of Costa Rica.

Design and Collaboration

David Gnabasik – Externalizations are used to extend our cognitive abilities by allowing all stakeholders to engage in a "conversation with the materials". Would it be the primary purpose of artificial intelligence to make the materials "speak" to the domain problem and its participants, or to bridge the gulf between the physical and the computational environments?

Our perspective on AI has been to think of it from the perspective if IA (Intelligence Augmentation)—how can we have the computer do things that it does well in ways that enhance humans in doing what they do well. This can take many forms

Jeffrey LaMarche – The prototype for the EDC is presented in an ideal fashion, but there is generally disagreement in group collaborations. There are people who are not team players and who have their own non-cooperative agendas; they will try to halt or impede the progress that can be achieved using the EDC method. Is there a fail-safe in the EDC to prevent destruction of the shared knowledge or from people poisoning the pool of knowledge with their own agendas?

Good question. I wouldn’t say that the EDC would fundamentally change human behavior; these sorts of things would still be part of how people choose to engage in these sorts of processes. I would say that the general approach we have tried to take in developing scenarios in the EDC is to try to find ways to use such potentially negative aspects and turn them into opportunities rather than try to suppress them. How can we take areas of conflict and make them into opportunities for drawing out tacit perspectives?

John Michalakes – The constraints and underlying assumptions built into collaborative environments, not to mention their sheer novelty, may cloud or otherwise affect the outcome of the collaborative process. Further, since extending the environment to address unanticipated aspects of the design problem may require technical skill or system access unavailable to many stakeholders, the fairness of the results might be in question. Lastly, one may not be able to assume that all participants are honest agents in the process. I would be interested in reactions to these concerns.

This is always a challenge with new technologies—that the “gee whiz” factor can get in the way. The goal would certainly be for the technologies to fade into the background (ubicomp folks make the point: how many motors do you use in a day)
We touched on the “honest agent” issue earlier, but the issues of accessibility of technologies will continue to be a challenge. One answer is a generational one: the new generations are more native techology "speakers." But if we are to really meet the challenge of this gap be we need to address it in other ways, as well.

Shumin Wu – The article made claims on the importance of collaboration to foster creativity. Is it possible that any specific collaborative system, which invariably requires a common interface, will be too rigid that it may stem individual creativity?

Therein lies the challenge! How can we support both? (kipling: the strength of the wolf is in the pack; the strength of the pack is in the wolf). I would say that we see that there may be many ubiquitous interfaces, some common and shared, some distributed, and personal) that would interact together to support individual and social creativity. The EDC is not viewed as an “end-all” interface, just an attempt to move away from the “desktop box”

Rhonda Hoenigman - The article talks about citizens working collaboratively to address traffic conjestion in their neighborhood. What role would city planners and statistics about population centers and frequently traveled roads play in this development process?

Our view is that city planners and other experts are also participants in the process. At times, traditional decision-making processes tends to put a high value on these participants (and they are valuable, informed contributors!). The paper was trying to focus a bit more on those whose voices are often not heard for various reasons–how do we help them to be informed in order that they can bring their valuable perspectives to the process in a meaningful way. Although we didn't focus on the experts as much, we do see them as important.

HCI Issues

Guy Cobb – Is the concept of a 2-dimensional display screen (even a touch display) fundamentally limiting as an interaction device, or is it that such devices haven't been used properly in the past and haven't been living up to their full potential?

I would agree with your point. While there are limitations (hey, wouldn’t we all love to have a holodeck?), the limitation is more in the paradigms of use. (The EDC is, after all, a 2D display screen)

Paul Marshall – The paper mentions that the EDC is attempting to move the focus away from the computer screen and towards creating an integrated environment - is this to say that the previous research in HCI is nearing an end (or peak) and that the primary path forward for HCI is "rethinking computational media?"

I would say that HCI has traditionally focused more on low-level interaction issues, and needs to extend the focus to issues that are higher-level, longer-term interactions (I would say “rethinking computational media” would be a broader characterization of the issue, but understanding how this impacts HCI is an important area of study within that)

Ths table describes a broad range of areas of study that impinge on what one might call "human computer interaction"

With the EDC, whereas there are many issues that need to be addressed related to the lower ends of the spectrum, most of our interest is on the areas related to the mid- to upper-levels.

Kyuhan Koh – The idea for the EDC sounds promising and interesting, but I wonder why there is no experimental data and statistical data for the idea. This paper mentioned about social aspects and human factors a lot, and I think it would be better to cover some experiments dealt with participants' satisfaction or frustration because even a great idea on the desk sometimes might not be a good one in the real world. The description of the idea is quite clear but when it deals with social or human factors, we may need at least small size of experiments to make it sure as I've learned.

Caleb Phillips – My question is basically the same as Kyuhan's. I'm curious if subsequent work made a statistical survey of the effectiveness of the process or satisfaction of the participants. And, if so, what those results were. As a related follow-up question, how does the experimental design here differ from what a Psychology researcher would devise? In other words, is there a pattern of experimental design that is unique to HCI problems?

The types of interactions that we would hope to support with the EDC are more complex and multifaceted than lend themselves to traditional HCI interaction studies. Often the sorts of interactions we are interested span multiple individuals and involve complex tasks and activities (although, there are low-level aspects of the EDC interactions that could benefit from such studies)
There have been some studies: [Hornecker etal; Warr et al]

Saroch Panichsakul - Is it going to take too long to overcome HCI challenging problems? If we cannot do that in the near future, how is it going to affect the EDC research and the development of real-world applications in this area?

There will always be new challenges! But we have already begun to see more wide-spread interest and work in areas related to the EDC (Tabletop conferences, Microsoft’s Surface…)

Lee Becker - The introduction of the paper talks of the summary of challenging problems for the future of HCI. In these items there is a lot of discussion about accessibility, however it seems that for much of the world accessibility is more limited by cost than anything else. Is cost a consideration in this research or is there an assumption that things can be made affordable after revising the methods?

In terms of cost of an environment such as the EDC, we have always worked from the assumption that Moore's Law would work in our favor for making the environments available for citizen participation processes–eventually.

Our concerns of accessiblity focused on other "costs" as well: cognitive costs, openness of the process, costs of participation, time, etc.

Evolution, Other Domains, Future Directions

Yifei Jiang – I am interested in the open system and system evolution mentioned in this paper as a part of collaborative design. This paper also builds simulations within the EDC. I was wondering how the evolution is processed by the systems/user (or both) with a right direction.

Yuli Liang – It's very possible that every stakeholder have much different view of points. When there are many participants, would the system evolving causes the redundancy of the system? How to make a system evolve "efficiently"?

Some models we have been working on include the SER (Seeding-Evolutionary Growth-Reseeding) model and the LCMS (Location-Comprehension-Modification-Sharing) model.
This is still a big question we are working on. We are studying open-source development models, social networking system, and what we call "participative software systems" for insights into how to push this forward. Visual programming has been an effort we have pursued, but we are also looking into mash-up approaches as well as genetic programming and human computation.

Of course there are many views of what "efficiency" might mean. Often, as computer scientists, that view is focused on computational efficiency. However the efficiency of user accessiblity, or effective means of communicaiton and negotiation might also be things that need to be considered under that rubric.

Ashok Basawapatna– The EDC convergence system idea interested me. Do you ever foresee a point in time wherein people attack any problem they percieve to be interdisciplinary, collaborative, or multi-system in nature through the construction of something resembling an EDC convergence system and the active recruitment of people outside their discipline or end-users who the problem solution will effect?

An environment, where participants from various domains could express, share, teach, and learn from other participants, encourage participation from new perspectives and do so with many forms of media and in many modes ranging form face-to-face to virtual to distributed to... would be wonderful. If the EDC gives even the slightest forshadowing of such an environment, then it will have served its main purpose.

Jinho Choi - I think HCI system should be designed in a way that provides different services for distinct individuals; however, there doesn't appear to be an obvious solution to the matter for EDC. Clearly individuals weigh each variable differently, so the question is how we decide the weights that satisfy everyone. If we go by statistics, there is a possibility to satisfy the major group (which must be still hard because not all variables are independent), but the minor group will never be treated well since it's not likely that they will change their normal routes for transportations. How do you get around with the problem?

Yes, this is the point that I tried to make in class–that the EDC table is not the be-all and end-all of the types of environments we envision. It simply provides one face-to-face social-interaction component of things that would also be combinations of web-based social interaction, repositories, individual information gathering and reflection spaces, etc

Nwanua Elumeze - My idea of a perfect school would be one where children sew electronic circuits and build automata for physics class; design handbags and shirts with graphs of their favorite equations for maths class, as well as write and act in their own plays for history class. Teachers would be "guides on the side" who help children independently and collaboratively construct their own knowledge, rather than "sages on the stage" who need to transfer a huge body of knowledge. How could some of the conceptual principles behind the EDC be applied successfully (and non-trivially) to a school environment where it may not even make sense to talk of "collaborative design"?

I think the EDC is more applicable to domains where the "material" goes beyond what is accessible directly in settings such as what you describe. To me, the "hands on" nature of what you describe are essential to certain components of learning. (This is a lot of the reason behind having "tangible" pieces in the EDC). At some level, the EDC tries to make virtual environments tangibly accessible. This could bring into the classroom the ability to design cities as part of the math and history classes and learn both the domain of study (math or history) and urban planning concepts (this is the thrust of the project we have with the Univ of Costa Rica).

Mohammad Al-Mutawa - As mentioned in the paper the EDC framework is applicable to many domains, not just urban planning, Do you think that in the very near future, with the advancement in technology, it would be possible to build an EDC system that is highly flexible/customizable and can service any domain?

We have always hoped to push the EDC beyond location-based domains, such as urban planning and architecture. One area we thought would be interesting to pursue would be something like organizational management, where organizational designers might use the EDC to redesign an organization using various organizational representations such as org charts, etc.

At some level, having a general purpose system would be ideal. However, our perspective has always been that creating such general-purpose systems suffer from the fact that the general tool might be userful conceptually, but adapting it to a given situation (domain) is difficult, yet that is what is need to be useful for that domain. This led us to the concept of "Domain-Oriented Design Environments" where specific domain support is provided in order to make the environment useful for a given domain.

Certainly this might be accomplished by having a more general core system with domain-specific "seeds" that provide sufficient support for the specific domains for users to take and evolve. We're not there yet.

Keith Maull - Conflict and competing alternatives often arise in the design problem, especially collaborative design. It seems that even in collaborative situations, humans may not have access to or explore alternatives. Considering the collabortive design of the transporation planning problem in the paper, there are many issues that may create tension between the balance of cost, maintability, neighborhood impact, environmental impact and scheduling constraints. How can the EDC be extended to provide deeper support for handling these issues? For example, when the cost of the bus began to exceed the users' threshold, the system might suggest alternatives such as light rail, or alternately become a "stakeholder-player" and suggest a change to the route that would accomodate viewpoints that are not represented.

Yes, these are the sorts of ideas we hoped to provide support for. Some of Gerhard's earlier "AI" work was on "critiquing systems"–tools that would point out breakdowns and allow people to explore alternatives in the context of that breakdown. Ernie Arias had the idea of extending this to "Virtual Stakeholders" where certain perspectives could be "represented" by an agent that would watch for certain cues and provide feedback from that perspective when appropriate (e.g., so that I could represent the fact that I don't want a bus stop by my house at a meeting that I can't attend). There are many challenges to creating such an environment, but these sorts of things would be ideal.

Jane Meyers - Does the EDC allow participants to visualize/model the problem before allowing them to articulate/describe possible solutions? To what extent would including or excluding this hinder the creativity of the participants? For example, because a certain bus route is underutilized will the first, not necessarily most creative or best, instinct be to modify this component?

This is one challenge in designing scenarios. We have tried various approaches ranging from presenting lots of information before moving to the decision-making process to throwing them into a decision-making scenario right away to see where various decisions succeed/fail (particularly trying to point out where they didn't have enough information initially), then allow them to take what they learned to try again. The overall view is that the EDC could provide a sort of "flight simulator" enviornment, where people could try out as many ideas as possibile to see what the strengths/weaknesses of various decisions would be. We are far from realizing that, but ways to increase opportunities for individual and social creativity have always been important goals for us.

Joel Pfeiffer - Has any work been done for networking between cities? For instance, Boulder and Denver would each have an EDC and, using the information in each EDC, improve the bus system between them to maximize efficiency and lower cost?

Ernie Arias (our colleague from architecure and planning) envisioned what he called "citizens corners" which would set up EDC-like environments in local community centers to allow greater participation from, for example, outlying areas. (The funding to try this out never came through, so we haven't realized that yet).

Your idea would certainly be an interesting extension to allow the interaction among various municipalities impacted by a decision, or by each others decisions

Last modified 19 September 2007 at 10:23 am by haleden