CCONVERSATIONSONVERSATIONS WITHWITH CCLEMENTLEMENT MMOKOK ANDAND JJAKOBAKOB NNIELSENIELSEN,, ANDAND WITHWITH BBILLILL BBUXTONUXTON ANDAND CCLIFFORDLIFFORD NNASSASS (INTERVIEWED AND EDITED BY RICHARD ANDERSON) Photo by Steven Pemberton

(top left) Bill Buxton, (bottom left) Clifford Nass (above, left to right) Clement Mok, Jakob Nielson, and Richard Anderson

46 interactions...january + february 2000

article

The 1999 Conference on Human Factors in Computing (CHI 99) had as its theme “The

CHI is the Limit.” At the conference, participants posed the following questions: What

are the limiting factors to the success of interactive systems? How can we enable users Jakob Nielsen Tto overcome those limits? What techniques and methodologies do we have for iden- tifying and transcending limitations? And just how far can we push those limits? This

theme and its questions provided a framework for the first live interview sessions to

be conducted at a CHI conference.

Tapped for these interview sessions were itself as the architects for the new economy. In some of HCI’s most original thinkers. 1988, Clement founded Clement Mok Clement Mok and Jakob Nielsen were asked Designs, later renamed Studio Archetype, to to address the Web and Web design limits to establish clients’ digital presence using interac- human–computer interaction (HCI). Bill tive media and eventually the Internet. (Stu- Richard I. Anderson Buxton and Clifford Nass tackled human lim- dio Archetype was acquired by Sapient Usability/Design/ its to HCI. Wayne Gray and Bill Gaver com- Corporation during 1998.) Before forming Discovery Adventures pared their perspectives on methodological his own agency, Clement spent 5 years as a 63 Woodside Lane limits to HCI. And Don Norman and Janice creative director at Apple Computer. Clement Mill Valley, CA 94941 Rohn squared off to address organizational is the author of Designing Business: Multiple (415) 383-5689 x2 limits to HCI. The interviewer was the Inter- Media, Multiple Disciplines, published by Fax: (415) 383-5187 views Chair for CHI 99—Richard Anderson, Adobe Press. [email protected] whose work is all about setting the stage for Jakob Nielsen is a principal of the Nielsen [email protected] avoiding or overcoming limits to HCI and Norman Group, a user experience consultan- [email protected] who has conducted several interview sessions cy. Until July 1998 he was a Sun Microsys- http://www.well.com/ over recent years on the stage of the BayCHI tems Distinguished Engineer and the user/riander/ program. company’s Web usability guru. Jakob coined Two of those CHI 99 sessions are repro- the term “discount usability engineering” and duced here. At least one of the other sessions has invented several usability techniques for will appear in an upcoming issue. fast and inexpensive improvements to user interfaces. Jakob’s most recent book is titled Web and Web Design Limits Designing Web Usability: The Practice of Sim- to HCI: A Conversation plicity (New Riders Publishing). with Clement Mok and Jakob Nielsen Richard Anderson: Are most Web sites The first session paired individuals who have well designed? come to the Web from very different places. Both have made impressive contributions to Jakob Nielsen: Definitely no. My conclu- the Web and to Web design. sion is that about 90 percent of all Web sites Clement Mok is chief creative officer of have miserable, miserable usability. That does Sapient Corporation, a consultancy that bills not mean that the other 10 percent are great;

interactions...january + february 2000 47

and have been very comfortable with. In looking at the Web and the Internet, each community or discipline needs to fundamen- tally look at their genetic makeup and their DNA and at what they consider “good” to be, and then challenge the assumptions on which they base their views.

Jakob Nielsen: Actually, I think that a lot of the things that we do know are still wait- ing to be applied. This applies to the soft- ware. . .for example, makers of Web browsers still need to implement a lot of the things that we knew in the hypertext field 10 years ago. If we would just do what we Clement Mok and Jakob Nielsen it just means that they are acceptable. That know, we would get better browsers. also does not mean that users spend 90 per- Regarding Web sites, I’m leaning toward cent of their time at bad Web sites. People the view now that Web sites are just funda- spend most of their time at the good sites, mentally never going to really make it. When but if you look across all Web sites, the vast, I say “never,” I mean in the next 10 years or vast majority are extremely difficult to use. so, because any substantial Web site is a con- tribution of a very large number of people. Clement Mok: I’m going to answer very You can have a central design team in a com- much like a graphic designer as well as an pany that is in charge of the home page and art director: it depends. I honestly don’t in charge of the information architecture think that our criteria for good Web site and search and big picture kinds of things. In design have been determined. The very fact many companies, there are decent people that we quickly make judgment calls about working on these things who have a decent what is good reflects a bias; our judgments clue as to how to do them. So, those things are loaded and are colored by a set of lenses. may in fact be done well. But, take five clicks Jakob is from a constituency in this commu- from the home page, and what are you nity that is coming to the Web from soft- going to get? Something horrible done by a ware development, while I and others come product marketing manager in Kentucky. So to the Web from the world of communica- the challenge is not just to get the central tion design. We both have a bias going into Web team back at headquarters to under- this world, and within those limited param- stand how to do good Web design, because eters say things like, “good is about clarity we are achieving that via our speeches and of communication;” or “good is about books and consulting and such. The chal- usability;” or “good is about performance,” lenge is to get the thousands of people who Clement Mok when in fact, a good Web site is all these are contributing their content to understand things with nuances of gray in-between. good Web design; that is a much harder chal- lenge—a challenge that I don’t think will be Richard Anderson: Clement, in your book met over the next 10 years. Designing Business, you talk about how old Intranet design is even worse, because habits die hard. What are some of the old there you have some random person in the habits that you believe are holding us back next department creating Web pages. It’s a or slowing things down in Web design? long time coming before we get all these people to understand how to do Web Clement Mok: Old habits holding us back design well. reflect the things that we don’t know; we very often fall back on things that we know Richard Anderson: Talk about some of

48 interactions...january + february 2000

article

the disciplines you both say need to be interface is your customer interface; the two involved in the design of Web sites. What are the same. Therefore, you need usability are some of them, and why should they be to be there, because if people can’t use the involved? Web site, they are going to become cus- tomers of some other company. Clement Mok: Let me start with just a few But the right business model also has to be from my perspective. The Internet started there. out to be a communication medium, so the On the other hand, we should not be say- discipline and craft of communication ing that you can’t design any Web site unless design is apropos for Web site design. But you have a team of 50 people, including Richard Anderson the Web has evolved into a way of doing three Harvard MBAs. Sometimes just one business, a way to conduct transactions. So person is enough. But in that case, you have skills that have to do with interaction to have all these little hats to put on. And design, software design, application design you’ll need to put on your little usability hat … are needed. once in awhile and say things like, “Let me But those are just the surface disciplines. go and get some other person in and see if You also need to have a good understanding they can use the site.” So even if you’re the of the psychology and the social science only person involved, you still need to get behind how people interact with the world. that outside perspective. Understanding the ergonomics of how something is used is not enough. The Inter- Richard Anderson: Jakob, the working net and how we deal with the Internet is title of your upcoming book is Secrets of an only a part of a bigger picture. We need to Information Architect. What are some of understand how it all fits in with our daily those secrets? lives. That entire spectrum of disciplines is need- Jakob Nielsen: I can’t tell you; you’ll have ed. It might be difficult to apply all disci- to buy the book! (audience laughter) plines on every single project. But when one There are a lot of different levels of infor- looks at the projects out there and the kinds mation architecture, and some of the more of Web sites that get implemented—many fine-grained levels may not be appreciated. fall short because these disciplines are not The bigger level deals with how to structure applied. the entire information space. That is often the focus of the professional information Jakob Nielsen: I agree that you need all architect or is why you need several people these different disciplines. And you can add on the team, because that’s not something more. In the future, we will probably have that your average marketing person or aver- more multimedia on the Web, so you will age writer will be able to do. need people who know about film produc- There is also information architecture in tion or audio design. Right now, the Web is the small, like when you’re constructing a mostly visual design and text and informa- story—let’s say a product description. That tion architecture, but we will keep adding as product description should not just be a it changes. page, it should be a set of pages, and archi- Today the Web is also the main driver for tecting that set of pages is information business strategy. So you also need people architecture in the small. This may not who understand business models. require a professional information architect; I’ve been known to make lots of snide the marketing person might be appropriate remarks about marketing people. Actually, I for that level of information architecture. make snide remarks about clueless market- You can get even smaller to answer ques- ing people. You need clued-in marketing tions like, “When it’s just a few pages, how people who understand that the Web is do you connect them?” An analogy is the your connection to the customer. The user work of an interior architect compared to

interactions...january + february 2000 49

the architect of the building; different peo- article or go into a headline. When writing ple work at different levels. In information headlines for print, you can very often use architecture, it’s specialized information teasers, puns…fun headlines, while that architects for some things, Web site editors doesn’t really work in the Web. In the Web, for others, and also people who are respon- you have to be much more straightforward sible just for a set of pages for a certain and describe clearly, in that one headline, event, a product, or an article. what the page is about, because people are not going to spend the time reading the rest Clement Mok: It’s interesting how infor- of the page. In the Web, remember that the mation architecture has very different inter- default is that you arrive at a bad page. pretations in different communities. At When a user goes to a new page, on aver- Sapient, we define information architecture age it’s bad—on average it’s “Back” button very differently than Jakob just defined it. time; users don’t give a second to decide The practice of information architecture in whether they want to dig into the page. our firm is also very different and is actually Sure, people also take a short amount of more specialized. We look at information time to decide whether to dig into a print architecture within the framework of a con- article, but they can more easily see the pho- cept. Yes, we use the word “information,” to illustrations, because they are immediate- and we also use the word “architecture,” ly there; in the Web, the illustrations are but we look at the information architecture going to come downloading later. In print, from the perspective of the user. We focus it’s easier to read subheadings or additional on user experience and on what users want text in addition to the big headline; on the out of their experience, and we do so with- Web, only a few things dominate. in the framework of a concept as opposed to In a study I did recently with Mark Hurst, the thing being created. we looked at how people could apply for There needs to be more discourse about jobs online, and the result was horrible. In the language of information architecture, one example, the person proceeded through because it is a word that gets misused and a series of steps and finally arrived on a page has very, very different meanings within dif- that had the headline that read, “You’ve ferent communities. now completed the job application.” So what did the user do at this point? The user Richard Anderson: Jakob, you have writ- said, “Ah, good, I’m done,” and closed the ten and also spoken lots about how people thing. Now if the user had bothered to read read on the Web, and you’ve said that too the body text, she would have seen “blah, many sites are designed by print people with blah, blah, blah, blah, blah, and by the way, advertising backgrounds. What is or what we want to make sure that you are really should be the difference between Web sure this is the correct thing to do, so please design and print design? press a confirm button to actually enter your application into the database.” Nobody Jakob Nielsen: I think that there is a very reads that body text. Users don’t click the last large number of differences. There is the dif- button that says, “Not only am I done, but I ference in page layout, where in print you confirm that I’m done.” The person went Clement Mok have a fixed canvas you can work with, and through all the steps, did all the work, and the entire thing is immediately presented to thought she was done because of the head- the reader when the reader turns a page in a line, and she lost all her work. magazine. On the Web, download time dom- inates usability to a great extent, plus you Clement Mok: Jakob, you are drawing don’t really know what the user has on-line, some very broad generalizations about therefore you cannot control every pixel. things that don’t work. In the world of com- Another difference lies in the writing of munication design, we consider appropri- the words that go on a page or go into an ateness and context. What you said might

50 interactions...january + february 2000

article

be true for that job application, but you Clement Mok: It’s really about relevancy imply that it is applicable to every Web site. and context. If you walk into a store where Attributes of tease and the ability to pro- all the clerks have on white lab coats, you voke or to communicate desirability or lust would still have a problem. You might think, or … all these things do and should play an “Wow, yeah, things are really efficient,” important role in the Web. because they all dress alike, they look like sales people, and you know how to find Jakob Nielsen: I think those things should them. But that is just as much of a problem. be put elsewhere, not in headlines. It’s too clinical; it’s a very industrial-strength way of dealing with things. There is a warm- Jakob Nielsen Clement Mok: Put them elsewhere? and-fuzzy component that needs to be Where? there. In short, there are no absolute rules. The Jakob Nielsen: Well, you still have the minute that everyone thinks there are and body text, and you do have things like illus- follows them, the Web will lose its relevancy trations. Note that I don’t say “have no for the user. graphics;” I say, “minimize the graphics.” Jakob Nielsen: I think it’s a matter of Richard Anderson: Along those lines, degree or where the main emphasis is. For Jakob, you say that of “look and feel” on example, with print, the “feel,” so to speak, the Web, ”feel” is more important. comes from leafing through the magazine. So it has some importance, but the “look” Jakob Nielsen: Right. dominates. On the Web, because you are driving and you are moving, the “feel” dom- Richard Anderson: Elaborate on that. inates, which is not to say that the “look” is irrelevant. It’s a matter of the balance Jakob Nielsen: Basically, the Web is a very between the two components. user-driven environment. You go to a Web site to get what you want, when you want Clement Mok: Consider an art gallery. If it, how you want it. That is the feeling of there’s no “look,” it’s going to be hard for going to a Web site. So, if the Web site feels that art gallery to be compelling. Is that not accommodating, feels that you’re in control, true? feels that it’s giving you what you want… that gives you a good, good feeling—a good Jakob Nielsen: On the other hand, con- experience of being there. That feeling is to sider the Web site of The Museum of Mod- a great extent dominated by the simple fact ern Art; that Web site is really unpleasant to of speed. Speed is not the only thing that use. For example, they don’t have the matters, but it’s very important. If you just biggest part of their collection online. They sit there and wait, whatever comes might be do use thumbnails, but I don’t think they use good, but you often get annoyed by the them effectively. At least that was true wait. In the studies, anyway, users say it’s not when I looked at it a month ago. worthwhile waiting. You get this irritating feeling that this Web site is sluggish and Clement Mok: Nothing is absolute. that it’s slowing you down. You think that they think that they are more important Jakob Nielsen: Okay. It is certainly true than you are. They’re not really giving you that if you want to be able to get to a high- service. It’s sort of like entering a physical resolution scan of a painting, that may in fact store and finding sales clerks all dressed in take a minute to download. But before you fancy Armani suits and not wanting to talk get to that, you want something that says to the customers. Now that’s a bad store. there are these 10 or 20, or even better, a Those are not helpful sales clerks. thousand paintings online for you to look at.

interactions...january + february 2000 51

Clement Mok: I agree, but it’s the abso- understand all our special terminology. In lute statements that absolutely frighten me. the various project meetings, designers have (audience applause) to stand up for the poor users because they are not there. Hopefully, every now and Jakob Nielsen: On the other hand, if you then they are brought in for a usability test. don’t say these things, people are going to But, on a day-to-day basis, it’s the designer’s continue creating horrible, sluggish, Web responsibility to stand up for the user. designs. Richard Anderson: Clement, in your Richard Anderson: Clement, is the way book you talk about how some of the disci- that you design for the Web different than plines that you believe need to be involved the way that you design for print? in Web design are not necessarily prone to collaboration. Is creative design such a disci- Clement Mok: Absolutely. I think it goes pline? back to the absolute part and the nonabso- lute part. Print is finite; it is an object that is Clement Mok: Graphic designers as well not temporal. The Web, on the other hand, as software designers are trained and genet- does not have a sense of permanence; it is ically engineered to be solo pilots. They sort quite ephemeral. A Web site is not only a of meet and get the brief; then they go off framework and a structure you plug compo- and do their magic. And I think engineers nents into over time. It evolves and morphs are that way, too. Inherently, and given the in ways that make it hard to get a sense of training that people have had, it is about what this one thing is at any one point. The being stars—about individually creating fundamental difference is that what you are great experiences. The Web is so complicat- designing does not have a permanent state; ed with its many interdependencies that it’s it is fluid and has different dimensions over important to not only nurture the great cre- time. It is, in fact, closer to performance art; ative individuals, but to also create the great it’s hard to describe what performance art is squadrons and the squadron leaders. at any one point, because it is constantly in flux. Richard Anderson: Are usability engi- neers prone to collaboration? Richard Anderson: What do you consid- er to be the role of the designer? What is Jakob Nielsen: That’s an interesting ques- the responsibility of the designer? tion. Usability, at least if you take it serious- ly, is about understanding different Clement Mok: Facilitate, mediate, inspire. perspectives and understanding how other people think. So, theoretically, usability Jakob Nielsen: Also to take the users’ engineers are better at collaboration. They role—to be the users’ advocate on a project. can still argue and say, “No, this is how it has Clement Mok A lot of other people on projects are all to be,” and they can get very stubborn focusing on what the company wants, such sometimes because all experience and 20 as dumping lots of messages on the poor years of research show that the download user. Again, the Web is so user-controlled, time has to be this fast. Indeed, that is some- users are going to go where they want to thing I’m going to stand on also. go. But that fact is still new to most business people. The designers are the ones who Richard Anderson: How do you achieve have to stand up for the user and say that collaboration? What is the secret to getting we have to make it easier for people to get people to collaborate? here, or we have to present information in a way that makes sense to someone who does Clement Mok: Collaboration, I think, not work at this company and does not requires engaging the individual or the

52 interactions...january + february 2000

article

group to take on a change. The minute that Jakob Nielsen: I think that most big cor- the metalevel of understanding within the porations are doomed when it comes to the group that the group is about to do this one Internet, because the Internet is an environ- thing is not there, collaboration is not going ment in which you are extremely close to the to work. When people are in disagreement, customer. It’s about direct contact. Yet these they don’t buy-in at the metalevel that they big companies have lots of layers of insula- are about to alter something fundamentally. tion between their decision makers and the You have to operate at a concept level so actual customers. The decision makers never that people are engaged and ready to see a customer in most of these places. And accept a change. in most of these places, the way to get pro- Another piece of the equation is about moted is not by doing something good for communication and really listening, and the customers, but by being smarter than Jakob Nielsen about creating a forum in which disagree- the other vice president. That kind of orga- ment can happen. Agree to disagree, and nizational structure may have actually create a respectful environment to facilitate worked very well in the past when change that. You can agree to disagree without hav- was very slow, when things moved very ing the respectful environment; that will slowly, and when sheer mass of scale had destroy collaboration. major benefits. The Web is a much faster Last but not least, I think you actually moving thing. have to have a physical space in which you Tight contact with the customer is becom- physically work together. ing much more important. In the old days, you could send out sales people to Jakob Nielsen: I also think that having schmooze the customers on the golf course, some sort of external criterion for how to and that’s how you could overcome the fact proceed can help. Otherwise, you risk every- that the company didn’t really care about its one arguing against each other: should we customers. Nowadays, if you don’t care do this or should we do that? And everyone about your customers, it is blatantly obvious might argue until the cows come home. This from the Web site, because it is not cus- is where the user becomes quite important. tomer oriented, and people just go to some If our main goal is to create the optimal user other place instead. So most of these big experience—to really satisfy customers, then companies are genetically suited to work in that can become the decision criterion. If an old-style, industrial environment; they you have that idea, and we have this idea— are not well suited to work in an interactive well, they are both good ideas. What will be environment where the quality of customer decided will depend on what actually gets service and truly caring about the customer the users their results faster, or on what sells is the important thing. more, which I guess is ultimately often the The 100 or so of them that will survive will goal. Having that external decision criterion be the ones in which the CEO decides that is important so that the decision is not based the Internet is the future. If there is a chan- on who can argue the strongest. nel conflict, too bad—we are going to go for it anyway. We are going to kill our own Richard Anderson: Jakob, you have said internal departments instead of having that you believe that 80 percent of the For- someone else kill them. Most companies are tune 500 companies will fail, because they not willing to do that. won’t make the transition from the visible world to the virtual world. Clement Mok: I agree with you on many of your points. But when you talk about fail- Jakob Nielsen: Right. ure, are you saying these companies will just go away, or are they going to have to deal Richard Anderson: Say some more about with how they fundamentally operate their that. business and with how to interact with the

interactions...january + february 2000 53

customer through another intermediary? Jakob Nielsen: Well, maybe you’re right. Are they going to have to look at who they It’s probably true in both directions. People are, what their brand is about, and building like Clement are the ones who actually customer confidence? understand what they are doing to the Web. So they can be responsible for managing Jakob Nielsen: There are probably sever- their clients or getting them to understand al different transition strategies. One strate- what to do. gy is to start up another brand and sort of I put it the other way around, saying that feed it, but basically it will grow up to be a the company has to closely manage the new thing. And then you can argue, wasn’t agencies, because you also have this phe- it really the old company that transversed nomenon where some people just go off into a new company? Maybe it was, and if and build a Web site—sometimes a very nice companies do that, and if you look at it that Web site, but one that is disconnected from way, the survival rate might be greater. But the company. That does not affect the men- the old company is still going to shiver and tal change needed for transformation into a become much smaller than it is now. I’m not Net-centric corporate culture. You can have so sure that 80 percent of these Fortune 500 a Web site that looks wonderful and may companies will die, in the sense that they actually do a few things very wonderfully, will have no revenue in 10 years or 20 years. but if it’s disconnected from the base com- But I think they will disappear from the list; pany, the base company is never going to be they will be so much smaller. very good. You can see this in a lot of the air- What are the new companies that are line reservation sites. They are doing nice going to grow up instead? Well, some might things on the front end, but the back end is be spin-offs of some of these old companies. not intended for consumers. So it is never In many cases, they will just be completely going to be a really good travel agency new companies. Those who are really doing online. That is why I say that the company well on the Web today are almost all Inter- must decide it is going to be an Internet net-only companies—companies that are company and that it is going to be in control pure, pure Internet. of its projects. People often ask me for the secret of Ama- zon.com: how come they are so much better Clement Mok: The way that Jakob is talk- than anyone else? I don’t think they have ing about these relationships is very inter- any secret other than that user experience is esting. In the world that I come from, it’s their number one criterion: does this make it called, “managing brand—managing and easier for people to buy a book or more dif- extending the brand experience.” Really ficult for people to buy a book? That is how talking about what a company’s brand they decide what to do. All the older compa- attributes are and what they might be nies have all these impediments in place that allows a company to move into these new make it difficult for them to have this one spaces. Let’s say that McDonald’s wants to criterion of “are we going to make this easi- create a personal information management er for our customers or not?” They have five Web site; McDonald’s brand is not going to other criteria to think about as well. allow them to do that with credibility even Clement Mok if they are able to develop the best user Richard Anderson: Jakob, you have experience. It is about the connections to argued that very few companies really and relevancy with the brand that existed in understand how to use the Web or how to the analog world; it is about what those compete on the Web. Yet, you have argued relationships are and how they’re managed that companies must keep Web agencies on and about how well brand attributes trans- a tight leash. Shouldn’t it actually be the fer. other way around, to the extent that that is At the same time, interaction with the cus- possible? tomer online creates new brand attributes

54 interactions...january + february 2000

article

that influence the dirt-world brand. Take zon.com. We may have had great trust in UPS as an example, which took the effort to the previous management—maybe they’re go after Federal Express by developing a still in place, maybe they are very nice peo- robust Web site. Now perhaps as competi- ple. I have nothing against them knowing tive as Federal Express, they have developed what Web sites I visit. However, brand attributes of being a major technolo- Amazon.com might decide that was not a gy player that have transferred to their dirt- good investment next month and will world presence. I don’t think too many decide to sell them to some evil empire. The people are really looking at how significant point is that users are being traded like cat- such changes can be. tle, and as long as that’s the attitude on the Jakob Nielsen Web, that we’re eyeballs, that we’re not Jakob Nielsen: I think you’re right, that really customers, that we’re not really indi- doing a good job on your online press can viduals, that we’re not really treasure, that reflect back on your old-world press and say, we’re like a Third World colony set to be “Yes, this is an efficient company, and, for strip mined—as long as that is the attitude example, we can get your package out, and on the Web, users’ trust will drop and drop you can track your package …” On the oth- and drop, and people will become more and er hand, it might play the other way around, more distrustful of any kind of offer or pro- which for some other companies is often the motion that is attempted. case. Some companies try to develop a Banner ads are a great example of that, brand of being very welcoming, very serving because users have this notion that if they to their customers. But, go to their Web site, select a banner ad they are going to get and you basically can’t do a thing. I think junk—they’re not going to get what was that is hurting these companies’ brands in a being offered. This is despite the fact that lot of cases. there are some ethical advertisers, including the CHI conference which advertised this Richard Anderson: Jakob, you’ve session via a banner ad, and sure enough, described the Web as a “low trust society.” it’s happening. But so many other advertis- What do you mean by that? ers are not ethical; for example, they adver- tise, “Click here to speed up your Internet Jakob Nielsen: The Web is a low trust connection” or whatever, and you click society, because you click in something and there, and it’s just an ad for some stupid you get something in response, but how do thing you would never want. You get a few you have any clue whether it is any good? of those, and at the end of the day, you say Not only that—the current atmosphere of you will never click on an ad again. the Web is that of utter disrespect for peo- ple’s rights. It’s like, “let’s trick people into Clement Mok: I think there is another giving us their e-mail addresses, then we’ll dimension to trust. Trust is lost very quickly, spam them forever after.” There are all but trust is also built up very quickly. Ama- these little boxes that say “yes, I want a zon.com and eBay have made explicit disclo- good offer,” but do you have any idea what sures about what they are going to do with you will get if you check those boxes. It’s too the information users provide. This has affect- much about conning users caught up in the ed the amount of users’ trust; this affected Web; it’s not about true disclosure about beliefs about how they conduct business and what’s going on, such as saying “every two what they are as companies. The building of weeks we’re going to send you a certain trust in those brands was accelerated. But, at type of message.” the same time, technology can accelerate a Plus, there’s a lot of horse trading with loss of trust and confidence in a brand if various Web sites. For example, Alexa, which many key principles are violated. is sitting on a vast database of personal information, was recently bought by Ama- Jakob Nielsen: I agree that you can build

interactions...january + february 2000 55

trust in the Web if you faithfully follow up chopped up in little unpleasant page views on all your commitments. But the Web in to maximize the number of advertising general is really suffering from this kind of exposures. This will get worse and worse as low trust; it is a distrustful environment the click-through rates drop and drop and where people are very skeptical about any drop. My prediction is that the click-through promises that are made. However, Amazon rate will be cut in half every year and will is a company that continues to follow up; reach 0.1 percent in 2001. And as the click- people have this notion that if they order through rate drops and drops, there will be something from it, two days later it will be more and more devious things done; adver- there. tisers will try to squeeze a little bit of rev- I’ve gotten a lot of e-mail about a certain enue out of all those eyeballs. other bookstore about its claims that books are in stock. People have said that they Clement Mok: Two weeks ago, I was sit- looked for a book on Amazon.com, which ting on a panel with representatives of ad didn’t have it or said it would take three agencies, and they were asking exactly the weeks to deliver. So, they went to this other same questions. “What do we do with ban- bookstore instead, which said, “Yes, we ner advertising? Is there going to be new have it in stock and will ship it in two days.” kinds of banner advertising?” It was fasci- So they bought it at this other bookstore. nating to hear that they could not think out- What happens? Three weeks went by before of-the-box and were instead thinking there they got the book. In effect, the bookstore will need to be different forms of direct lied to them. Whether the bookstore lied to marketing and more advertising formats for them because the book was in stock until a and different ways to treat banner advertis- half a second before they ordered it we will ing. Their solution is to not look to banner never know. But those particular customers advertising, but to look at what they need will never buy at that bookstore again. to do to develop an ongoing conversation and relationship with their customers. Until Clement Mok: The whole issue of fulfill- advertisers get burned so bad and spend so ment and of reliability of information is an many advertising dollars without getting important piece of what I think is a very any return on their investment, we will con- complex puzzle about trust. tinue to suffer from this whole notion of having banner ads and interstitials and all Richard Anderson: How might we get sorts of other funny words; say, what the away from banner ads? Is that a possibility? hell are those things?

Jakob Nielsen: My answer is micro-pay- Jakob Nielsen: They are just making ments. Those who pay are the ultimate cus- worse ads—more and more intrusive, more tomers and will ultimately be the ones and more annoying ads. I think people will things will be designed for. I don’t think we react more and more against that—buying Clement Mok can develop what is going to be the most software that screens out the ads, for exam- important communication medium in the ple. That should be seen as the final cry for world if it is purely advertising based. That help. If it’s so bad that people actually pay medium is not going to serve society—it is money for software to get rid of these ads, not going to serve people. Unfortunately, that’s a true sign that this entire strategy has people will have to pay, and they just have failed. to recognize that there is no such thing as a And, Clement, your comment about rela- free lunch. If you want to have a service that tionships with customers is such a key point. is paid for by advertising, it is going to be Good relationships are not based on taking designed for advertising. Articles are not out your club and banging a customer over going to be written in a way to give you the the head. If you want to have a good rela- best information; instead, they will be tionship with a customer, you don’t start out

56 interactions...january + february 2000

article

by cheating them, which is what many of without merit.” On the other hand, these current ads are doing. Your approach Microsoft wouldn’t like to have me there, needs to be more trustful, more relationship because I would also say that this so-called based. Think of it more as marketing and integration is just a little Band-Aid™. We less as advertising. Build up Web sites that want real integration. I can give them a list actually give customers good service. That, I of 10 additional things they need in the next think, is a very, very viable strategy. version. However, that means that content Web sites are going to lose their revenue sources. Richard Anderson: Studio Archetype They had better think of other revenue conducted a study with Cheskin Research on streams. the trustworthiness of e-commerce. Say some more, Clement, about that study. Jakob Nielsen Richard Anderson: You made a quick reference to micro-payments. Can you say a Clement Mok: A client had come to us little bit more about those? and said, “Design a very secure site for us, and design it so it looks like it is secure.” Jakob Nielsen: The main problem with What??? “Looks like it is secure?” That con- micro-payments is that there are too many founded us. Isn’t it enough to make a Web schemes right now. It’s not that there are site secure? The answer is “no.” any technical problems. Too many different We decided we would conduct surveys, do schemes have been invented. I think that focus groups, etc., to try to understand what Microsoft has abandoned their responsibility people identify as trustworthy. We talked to to lead the industry here. They should a variety of people—from people within the decide on one of them, put it into the oper- industry to total novices—about what they ating system, and just ship it from now on. felt was important to building trust. We That is the only way it is going to work. Peo- found that many of the things we had ple are not going to download something to thought were important, like security plug into stuff—that just doesn’t work. You encryption, are important, but we also can say good things about Bill Gates or bad found that building and destroying trust things about Bill Gates, but the fact is that happens in a short amount of time in this he rules the world. So, it is his job to just pick Internet space. Technology performance one of them. I don’t even care which one. plays a role, security marks play a role, brand My favorite is MilliCent™, but I don’t really plays a role … but navigation plays the most care. Just pick one of them, and stick it into significant role in shaping trust. You can the OS. Then the OS can become the have a no-name brand, a teeny tiny compa- enabling mechanism for the network econo- ny, but if you make your Web site very clear my. If it’s not done properly, it’s not going to and easy to navigate, you can build up trust work. very quickly. Actually, the Microsoft antitrust suit that People’s response to several well-known is going on is interesting, because it hinges brands was miserable. For example, people on a user interface question: should the couldn’t find any product in the SharperIm- Internet really be part of the operating sys- age.com Web site, and they rated the com- tem? Should the browser be integrated or pany very low as to the extent to which it not? Despite it being a user interface law- was trustworthy to transact business with. suit, none of the parties have dared call a That navigation was of greatest significance user interface witness. I think it is because was the most surprising finding. both of them know they would be in deep trouble. If the government called me as a Jakob Nielsen: To me, that says that if witness, I would say, “Sure you need to inte- you make it usable, then people will feel grate the browser with the operating sys- more welcome and will trust it more. The tem; kick out this stupid lawsuit—it’s analogy with the physical, old-fashioned

interactions...january + february 2000 57

shop is that if you have salespeople who are long term. Let’s take an airline as an exam- helpful, that is better, and if you have sales- ple. The airline can run a million ads that say people who refuse to talk with you, and you their planes are always on time, but if there get this feeling that they have something to is an hour delay every time you fly with the hide, this is not a site that really will work airline, at the end of a very small number of and where you can get things done. trips you would say that this is not a reliable Although Clement didn’t say this, their airline. It’s similar with the Web; every single report about the e-commerce trust study is click is delivering some service to the cus- available on the Sapient Web site, and I tomer. Therefore, the feeling of the delivery think it is a very good report. of that service determines the brand much more than that which some people who talk Richard Anderson: Studio Archetype has about brand get more preoccupied with. a discipline called “brand strategy.” What They’re preoccupied with the superficial does a brand strategist do? aspects of branding, such as how big our logo on the home page is. They ignore the Clement Mok: Brand strategy is about substantial aspects of branding, which are trying to reconcile the brand attributes in about promising the user or customer a cer- the dirt world with the brand attributes in tain thing, and then delivering it. If you do this new digital world. What is the tone of not deliver it, people will very quickly learn the conversation between the two? It can be to ignore you, because they get bombarded a friendly conversation or a very hostile con- with so many messages in modern society. versation. What it is can shape what a brand does and determine whether the attributes Clement Mok: The great digital brands are relevant or have credibility. Brand strate- are going to be about companies who actu- gists within our firm look at this space for ally manage reliability—reliability of provid- companies that are like deer standing in a ing valuable information. There will be vehicle’s headlights and being told to go variations, but as we move forward, the change—to do something on the Internet or great brands will be about reliability rather they won’t continue to exist as a business. than just quality. We are helping companies find and articu- late their brand attributes in the dirt world Richard Anderson: Jakob, one of your and in the Internet space, then figure out a columns is entitled “Personalization is Over- path forward as to how to change or build rated.” Is it really? customer perceptions. Brand strategists mind those relationships with clients. Jakob Nielsen: Personalization is overrat- ed, because it is the bad excuse of the per- Richard Anderson: Jakob, I think I’ve son who can’t design a good Web site. It’s seen you quoted as saying that companies like, I can’t figure out what to show to the are preoccupied with branding. user, so we can just leave it up to them; since I don’t know to design it, I’m sure the user Jakob Nielsen: Well, I’m not sure I said can do it. That is basically the attitude you that exactly, but they probably are. The find in some places. It is the responsibility of Clement Mok potential downside to branding is that some the Web designer to design that experience, people think of branding as image. What I that environment, for the user, so that the claim branding really is is the customers’ user can then take or enter that environ- impression of you and what they actually ment and get what they want. A navigable get from you—in other words, the actual environment in which you can actually fig- experience. This goes back to the question ure out what to do and get what you want of whether look or feel is more important. now is the more important goal. The feel, or what you actually get, is in many That does not mean that there should be ways what determines the true brand in the no customization features. It is very nice for

58 interactions...january + february 2000

article

users to get in and tweak the design: let me but the Macintosh team. Being part of that see more of this or less of that, or let me team shaped the way that I look at the always get the weather forecast from my world of designing the entire user experi- home town, and so forth. I do believe in that ence. It is not just about the screen or the type of customization. What I do not believe graphics; it is about looking at how the in are the cases where the system is trying to machine, the box of that machine, and its model or second-guess the user and stereo- contents fit within the space. It is about the type them. Do you know the infamous thing supporting tutorial that goes along with the on Amazon.com that says, “Welcome back, machine, and about all the demos. It is my Jakob, except if you’re not Jakob, click focus on the totality of the experience and Jakob Nielsen here”? It goes on to say that you really the need to work together to create great should buy these crime novels. I don’t really experiences that Steve and the entire Mac- care about crime novels. Yes, I bought quite intosh team influenced. a lot last week, but they were for a birthday Others who influenced me were the hus- present for somebody else. It’s trying to be band-and-wife team of Charles and Ray too smart about it. What is very good is that Eames. They are designers. They try to con- for every book they list, they also list five nect the most mundane day-to-day things other relevant books. Every day you are with technology that is useful and desirable. interested in some other thing: today I am They have designed films. In [19]62, they did interested in this book, so here are five oth- an amazing multi-image show with 30 er books that are relevant. That is a way of some-odd projectors at the U.S. Pavilion in being driven by the users’ immediate needs New York City, which in a way was the start and requirements. When you study these of the multimedia movement. And they did things with users, you will quite often get the powers of ten. They have looked at tech- feedback that says, “Don’t box me in; I’m nology not as the end, but as the means— not just that person; I have broader perspec- the means to connect. Making tives and views.” connections—making meaningful connec- So, I’m not saying there should be no cus- tions—is so important. tomization features; they can be useful. I’m just saying to not rely on customization as Richard Anderson: Jakob, same ques- the main design strategy to overcome defi- tion. ciencies in simple navigation. Jakob Nielsen: Back when I was a stu- Clement Mok: Another way to cut what dent, I read a lot of Ted Nelson’s books: Jakob is talking about is, in my world, called Dream Machines, Literary Machines, and “point of view—editorial voice.” People still those books. They influenced me a lot and like to be led through information and actually got me into hypertext. He was resources that have a very distinct point of extremely visionary, even though he never view. That piece of the equation is often actually shipped anything, which doesn’t missing. matter because he wrote some very, very good books. Richard Anderson: Clement, who has The other person I’ll mention is John influenced your work and your perspective a Gould at the IBM Watson Research Center. lot? I’m actually very different from John, Who are the people you have greatly because I don’t go around measuring every- admired—people who have influenced the thing very carefully. He did. He believed that perspective that you are sharing with us anything you could have a discussion about today. could be measured. He did some extremely careful studies on how people read from Clement Mok: A lot of people. One is computer screens versus paper, and varied Steve Jobs—actually, not Steve Jobs himself, every possible parameter of all that, to final-

interactions...january + february 2000 59

ly discover the need to have better resolu- Richard Anderson: How has humility tion. We still don’t have that needed resolu- played a role in your work, Jakob? tion, even though the need has been known since his studies were presented at one of Jakob Nielsen: Well, I don’t know. our early CHI conferences. It is a disgrace Clement is certainly right that there is a lot that we still have such bad screens. of stuff we don’t know, particularly about John was really a great…I was a great the Web, because it is so understudied. It is admirer of him. I learned a lot from John amazing that you have hundreds of millions Gould even though I’m not doing things in of people using the same system, and we his exact style. I am more into the “let’s do don’t really know why. some quick-and-dirty kind of thing, and get On the other hand, there are a lot of the insight,” rather than get all the numbers things we do know, and I think it is more my John would have wanted. responsibility to go out and say, “People, this People like Jack Carroll, Tom Landauer, is wrong; this is what you should do.” Even if and Tog [Bruce Tognazzini] —how can you it’s only 90 percent sure to be true, it still has not say Tog?—also influenced me. Lots of a much better chance of moving people in good people. the right direction, rather than overlooking these things. If in 99 percent of the cases that Richard Anderson: In a very recent inter- you make download times faster you also view, Clement, you talked about how you make user experience better, you might as learned about humility as a student. What well tell people that. Otherwise, they are not role should humility play in the wild world going to pay attention to that issue. of Web design? The concept of humility does play a big role in Web design. You have to remember Clement Mok: None of us know where that when you’re designing a Web site, we are going. You don’t know what you you’re designing this little speck of dust in don’t know. We are colored by our past the universe of the entire Internet. You’re experiences, and we can go out and claim contributing a tiny little minuscule incre- that we know everything and that this is the ment to this bigger user experience. I call way it should be. I have done that; we have this Jakob’s Law. Jakob’s Law of User Experi- all done that. It is a part of human nature. ence is that users spend most of their time It’s easy. on other Web sites. No matter who you are, It’s much harder to ask, “Why not?” When people spend most of their time on other we take on a project, we ask our clients the Web sites. That says to me that you cannot appropriate list of questions, such as “Why drive the user; they spend most of their time are you doing this?” and “What’s the busi- elsewhere. ness objective?” When someone off the cuff throws in something that appears to be Clement Mok: Jakob, what if your “’99 audacious and bizarre, we totally discount it. percent” is wrong? What if it is only 85 per- Clement Mok In some projects, that something was a bril- cent? liant idea, but we just threw it out the door. We should have looked at it and asked, “Why Jakob Nielsen: Well, it’s still the big not?” That is how the bold and bright and majority, right? It is true that there is no rule extraordinary ideas come about. Ted Nelson that doesn’t have an exception. Good and Marc Andressen [inventor of Mosaic and designers know the rule, and they also know co-founder of Netscape] are two people who when to break it. However, most of the mil- asked, “Why not?” The Web is the result of lions of Web sites and billions of pages are asking, “Why not?” Humility comes from fail- designed, as I said before, by some product ing to and then seeing that you didn’t ask the marketing manager in Kentucky. You have “Why not?” question. That is when you say, to tell these people the basic rules, and then “Yeah, I messed up big time.” once they get to be brilliant…

60 interactions...january + february 2000

article

Clement Mok: But broadband and many Audience Member: Many leaders of the other things are coming down the pipe very information economy seem to bemoan the soon. Hence, I think that dictum of 99 per- fact that the Web and the Internet have not cent with regard to download time is irrele- become the pervasive, ubiquitous appliances vant. that they expected, or that it’s taking longer than they expected. Do you think the obsta- Jakob Nielsen: But the download time cle is financial—that people can’t yet afford issue will always be with us, because it is the technology for it to become ubiquitous? based on basic human characteristics. The Or is it the bandwidth? Or is it the lack of one-second response time rule says that any- really useful stuff for people to do online? thing slower than one second is going to feel unpleasant, is going to feel sluggish. Clement Mok: All of those things, but Jakob Nielsen Even if you can say for sure that at some patience is virtue. Adoption sometimes point in time we will have enough band- takes 10 years. Consider the penpoint oper- width to send a current Web page in a sec- ating system; it actually started in 1990 or ond, I’m sure some designers will find a way 1991. Things take a bit of time. of making it a fancier page that is slower to But, indeed, a lack of useful, usable, and download. desirable things on the Net is preventing it Eventually, you’ll be right, but that will be from being ubiquitous. more than 10 years from now. Jakob Nielsen: I agree. But I think that Audience Member: I wonder if the window the cost is also an issue, because it is still too of opportunity to fix the Web is closing. I expensive—particularly overseas. On the wonder if (a) people will so totally mistrust other hand, that is the one thing we know it, that all will abandon it, or (b) we’ll just will go away. If all other problems in society accept it, with the poor performance it has. were like that, we would be happy. Ten After all, we all still shop in stores where years from now, the computer will cost 1 people are unhelpful and either wear fancy percent of what it costs today. So, that prob- suits or white lab coats, and those stores lem will go away. seem to be doing quite well. A much more serious issue is not being addressed very well right now. Some of the Jakob Nielsen: I don’t think it’s too late. I recent data from Denmark says that person- really believe we are still in the infancy of al computers are now in 52 percent of all the Web, and it’s going to become much Danish households. That is only 3 percent bigger and much better. I really believe that more households than 2 years ago. So, there the Web is the ultimate consumer empower- has only been a minuscule number of addi- ment medium, because every single mouse tional households getting a PC. Sales are still click is a vote for what’s good and what’s strong, but that is because the same house- bad. That drives the traffic to a better Web holds that already had a PC 2 years ago are site, and the worse sites fade away a little buying more. So, basically, the other half of bit. I am very optimistic. As new things come the population are just not buying. It’s a mix out, the better Web sites will bubble to the of lack of motivation, because there is noth- surface, and the poor ones will fade away. ing for them there, along with a matter of That is also why I think, by the way, that things being too difficult for them. If you the current stock market valuations are compare the user interfaces of Windows 95 problematic. There is this notion that and Windows 98, the incremental increase whomever is sitting on a good market share in usability might in fact be 3 percent. To get now will inevitably have good market share the other half of the population to go on- in the future. But, there is very little loyalty line, these systems need to be more than on the Web, and if something better comes just a little bit better—they have to be enor- along, people are going to go there instead. mously better. That is what we are still wait-

interactions...january + february 2000 61

ing for. There are certainly projects under only a couple of years ago, there was no evi- way to make systems enormously better. dence that our 20 years of experience had But, right now, they are just too difficult to any impact at all on Web design. I’m won- use. dering if you have any hypothesis about why no impact was made when the Web first Audience Member: You’ve talked about started to take off, and about how we can how to get people to agree when they are do better in the next 10 years. collaborating, and about the trust issue, and about banner ads and how we might get Jakob Nielsen: Usability has never been a away from them. I think it all comes down to success in computer companies, because getting better data. In the trenches, when they haven’t cared about customers. They you’re working with the ad sales people and have been able to get away with that, the producers and the designers and the because the customers let them get away engineers, and there is disagreement about with it. Because the people who buy the what should be on the page and about software don’t really know or understand where the ads should go, what do you do? how it is being used, and for lots of other There is such a paucity of information out reasons, computer companies don’t need to there right now. It’s very hard for me to make usability their guiding criterion. argue to the sales people why we don’t The Web is a little bit different. On the need the banner at the top or why we don’t Web, usability is becoming a guiding criteri- need the sponsorships on the side. Because I on for quite a number of companies. In the have no data to support my claims, they beginning of the Web, this was not the case. don’t believe me. This is a call to action to At the beginning of the Web, everyone was the community to do more studies, to iden- so excited: “Wow, I can put stuff up, and tify the things that will enable users to get people in New Zealand can see it.” They what they want and advertisers to get what never thought about how it was going to be they want. used. There was a revolution just in the fact that you could put stuff up. That is probably Jakob Nielsen: I agree. There is incredibly why usability was downplayed in the very little research being done on how people early years of the Web, and why wild exper- use the Web and why people use the Web. imentation, which was probably appropriate There is some amount of usability testing for that stage of the Web, was up-played. being done that affects individual designs, People tried all kinds of odd things, most of but there is not much generalization hap- which didn’t work, but that was fine for the pening. Yes, we need more. And I guess we pioneering stage. We have now entered the do need more data about what makes stage of attending to the business critical advertising work well, even though I think aspects of things, like whether you keep banner advertising on the Web is a doomed your customers and whether people can find endeavor. products. Usability is critical on e-commerce sites, because if people can’t find the prod- Audience Member: Though our field is a lit- ucts, they are not going to buy them. You tle over 20 years old at this point, I have can now tell that a large number of compa- Clement Mok been a little bit disappointed in how much nies are, in fact, making usability a very progress we have made—in how much major, major issue in the management of impact we’ve had on the computer industry. the Web. Internet-only companies certainly I would have hoped that by this time we are, but so are some old-world companies. would have institutionalized usability as part of software development, but I don’t Clement Mok: I want to add that the CHI think we are anywhere near that. community was not part of that initial com- My bigger disappointment is that when munity that was really developing and the Web really began to take off, which was doing things on the Net. So, much of what

62 interactions...january + february 2000

article

you had learned was never transferred. from being some sort of nerdy specialty that Being a player, and being involved as a prac- a few people would gather to discuss at this titioner within that space, is important to kind of conference, to being what makes an achieving that transfer of knowledge. There entire billion-dollar company live or die. is an amazing amount of catching up that That change in the importance of usability is needs to happen. 100 percent due to the Web. The fact that the Web now makes your customer interface Richard Anderson: My final question is into a user interface has so changed the related to that. Has the onset of the Web importance of this field that we can now go advanced the field of human–computer pick up all the research off the shelf and say, Jakob Nielsen interaction or set it back? “Okay, we know how to do this; let’s go do it.” The people who do that will be worth Jakob Nielsen: Very narrowly construed, billions of dollars more than the people who it has set it back, because the design is now don’t. like design was for IBM mainframe termi- nals. That’s the type of design you get for a Clement Mok: I think the Web has funda- lot of applications people are putting on the mentally altered the fabric of how we even Web. So, very narrowly construed, the Web look at interaction. We have been siloed and has set the field back dramatically. have been largely looking at applications. But I think the Web has extended it, But the Web is a communication medium, because it is now business-critical. That, I and it’s an application, and it’s software, and think, is the ultimate. We know a huge it’s a distribution channel for your business. amount of stuff about human–computer The Web has fundamentally altered our pro- interaction, but it’s just been sitting there in cess of thinking about what we are all as a conference proceedings and not being used. community trying to shape and create. So, I Because of the Web, usability has changed think it is all for the better.

Human Limits to HCI: A Conversation with Bill Buxton and Clifford Nass

Cliff Nass is an associate professor of commu- fessor in the Department of Computer Sci- nication at Stanford University, with appoint- ence at the University of Toronto. Bill is well ments in Symbolic Systems, Sociology, and known to CHI conference attendees, having Science, Technology, and Society. In addition, presented some wonderful work at almost all he is codirector of the Social Responses to past CHI conferences, but a couple of CHI Communication Technology Project. His conferences ago, he threatened to never areas of specialization are experimental studies return. He appeared at CHI 99 from Japan Richard Anderson of social-psychological aspects of human– through the power of technology. computer interaction and statistics. He has published extensively, including The Media Richard Anderson: Bill, why did you tell Equation: How People Treat Computers, Televi- a CHI conference audience a couple of years sion, and New Media Like Real People and ago that you would probably not return? Places (Cambridge University Press), which he authored with Byron Reeves. Bill Buxton: While I didn’t necessarily Bill Buxton is a designer and researcher mean it literally, I wanted to express my con- specializing in human aspects of technology, cern and disappointment with our discipline human–computer interaction, and technolo- and the state of HCI. When I look at the gy mediated collaborative work. He is chief design of computers, for example, I see noth- scientist at Alias | Wavefront Inc., as well as its ing since the introduction of the Xerox Star parent company, SGI, and is an associate pro- which shows much progress. And I question

interactions...january + february 2000 63

the value of the research we’d been doing in and really look at that product and under- the community. If I piled up all the literature stand how many important and great ideas that has been published by the CHI commu- appeared in it, and, sadly, how few have nity since the Star’s introduction, it would be entered the interface community more gen- taller than I am by far. Yet it has brought us erally, and how many have entered without nothing that’s gone significantly beyond the giving credit to Bob. Star, and, significantly, the innovations repre- sented by the Star were accomplished with- Richard Anderson: Let’s talk about some out the benefit of such a body of research! of the underpinnings of that work. (I’d like So what does that say about the value of both of you to talk about the kinds of things what has followed? It is not that the research that you want people to understand and was a waste of time; it has told us a lot. But attend to regarding human limitations, but in terms of helping consumers and affecting let me start with Cliff.) What is the SRCT the design of these things called computers model, Cliff? What is the “media equation?” that we foist upon them…I would say that we’ve had very, very little effect. I find this Cliff Nass: Fortunately, it’s one of those lack of impact an interesting challenge, and I things that can be summarized in a sen- also find it a disappointment. tence. In essence it says, “People’s interac- tion with computers, television, and other Richard Anderson: Cliff, as I recall, the media is fundamentally social…and we CHI community didn’t exactly warmly mean “fundamentally” in the strongest pos- embrace you and your research not too long sible sense. One can take the theories of ago…perhaps partly because of the associa- how people deal with other people and tion you had with the development of apply them to human–computer interaction, Microsoft Bob. I wonder if you could tell us and also take the experimental methods and some of the reaction that you received early experiments that are used to understand on? these theories and apply those, as well, to understanding and testing human–comput- Cliff Nass: I think one should separate the er interaction. By doing that, we discover reception of the research program, which many, many new things we never would certainly wasn’t warm, from the reception have guessed were true—and also change of Microsoft Bob, which also wasn’t warm! our focus to variables that might be impor- In the research program, we were trying to tant to regular people—variables such as do something radically different—and try- fun and liking...things that tend to be ing to challenge a lot of the fundamental absent from CHI conference presentations. ideas lurking around CHI and the interface community (more generally). So, it wasn’t Richard Anderson: Is it always true…do surprising that it took a long time and study people respond to technology socially in all after study before people began to say, cases? “Hey, there may be something here.” And happily now the work is more warmly Cliff Nass: I think it’s most useful to think received. With Microsoft Bob, what hap- of the answer as “yes.” One of the reasons Bill Buxton pened was really a hysteria towards that that the reception at CHI may have been a product, which I think had more important little less warm than it otherwise might have and unique interface insights than any oth- been was that I believe that people should er product I’d seen in a long time…in fact, try to state their theories in the strongest since the Xerox Star. Now there were prob- and most forceful possible terms. This does- lems with it, as there often are with first- n’t mean you believe them that strongly, but time products. But I think that now we it means that you voice them that strongly should calm down. It’s been over 5 years so that others can know when they agree or since Bob was released. It’s time to go back disagree. So, I think the answer is “yes.” Fur-

64 interactions...january + february 2000

article

thermore, I think that we learn a lot more to be understood in order to execute social when we make believe the answer is always aspects of user interface correctly? “yes” and focus on those one or two times in everyone’s life when we are able to think Cliff Nass: Just about all aspects of human past it. behavior…simple things such as politeness and flattery, and richer social rules such as Richard Anderson: What characteristics being part of a team, feeling cooperative, all of technology prompt social response? the ways we induce cooperation, and all the ways we induce trust…all the things that Cliff Nass: That’s a great question, and make us feel like we’re working together and that’s actually something that we’re really we’re being polite and we’re being pre- Clifford Nass trying to nail down. One of the most likely dictable (predictability is a fundamental things to do it is the use of language. I’m not social response)…anything that makes us feel talking only about speech here, though like this is a natural and social experience. speech clearly ups the social ante. Even a text-based system that uses language is Richard Anderson: Give us an example enough to generate social responses. Inter- of how tending to politeness could affect action, cognizance of what happened user interface design? before, and reacting to that cognizance…doing something social that’s Cliff Nass: That was actually the first study normally done by a person…that seems to we ever did—way, way back. It simply cue social responses. Things like voices, faces, involved people working with the computer, and the other anthropomorphic features are and then either the computer they worked also relevant. It doesn’t take much at all. with or a different computer…asking, “How well did that first computer do?” With Richard Anderson: Why do these charac- humans, we know what would happen. If teristics cue social responses? Richard asked me, “How do you like the interview so far?” I’d obviously say, “great.” Cliff Nass: We think that the answer is an You in the audience would likely say “great” evolutionary one. The greatest opportuni- as well, but you might say less nice things ties and the greatest problems a human than I would. Because of the social demands faces are other people. Though not true of when people ask questions about them- all species, all our primary predators and our selves, they get more positive answers than primary prey and our primary mating part- when others ask. When we did the same ners are all in the same species. Therefore, things with computers, sure enough, people humans have evolved so that when they see gave significantly more positive responses to anything that even hints at being human, the computer that asked about itself. their brains immediately kick in and say, “Ah Another important point is that they ha, this is a person! And I will act as such denied this vehemently, even after being because there are so many advantages to showed the data. They said, “Well, it obvi- doing so.” Our ancestors that didn’t do that ously applied to all those other computer really lost out. So we are living with an old scientists who didn’t know better; it couldn’t brain that’s now confronted with new possibly apply to me!” So that was a clue media that comes along and fools us and that we were on to something, and we’ve tricks us. But our brains are so powerfully probably done 40 or 50 experiments since. directed to those evolutionary fundamental The way that plays out in interfaces…most responses of socialness that it’s almost interfaces are remarkably impolite, not just impossible to overcome. in the Miss Manners sense, but in the sense of following the Gricean rules of polite Richard Anderson: What are some of the interaction and all the other rules of polite important social rules that you believe need interaction. We have interfaces that con-

interactions...january + february 2000 65

stantly violate politeness norms. Yet, we’re to have lower status; and there are some cas- surprised when people don’t enjoy them, es where we want it to seem to have equal get angry at them, get frustrated, etc…. status. For example, if you had a very com- plicated medical monitor, you wouldn’t want Richard Anderson: Why do so many peo- it to sound highly submissive and tentative, ple despise the dancing paper clip? saying, “Well, perhaps you need a little more insulin right now—maybe we should consid- Cliff Nass: I was involved in the work on er that.” Instead, you would want one that the dancing paper clip, but I think they’re seemed confident, dominant, and basically in right to despise it for many reasons. The sin- charge. And if we sweared at it, it should say, gle best reason is a fundamentally social “Don’t do that to me! I can swear at you, you one. When you ask people why they hate can’t swear at me.” There are other devices, that paper clip so much, one of the first other contexts, other roles, other situa- things they say is, “Well, every time I write tions…in which we want a submissive inter- ‘Dear Fred,’ the thing pops down and says, face. So, the answer is that the interface ‘Oh, I see you’re writing a letter,’ and they should be consistent with the role that is dismiss it. The first time, it was okay, it was has—exactly as we admire those people most helpful. The second time, it was at least try- whose behavior and personality match the ing. But by the forty-seventh time, it was roles that they have. clearly being at best passive aggressive and at worst down right hostile, implying that I Richard Anderson: MIT and IBM recently couldn’t make a decision about the right announced plans to develop a computer’s thing to do.” Well, we know what we do ability to detect users’ emotional states so with people like that—we hate them. In that it can respond more appropriately to fact, the first rule in the Dale Carnegie users. Do you think that these plans have course of “How To Win Friends and Influ- promise? ence People” is—remember things about people. The paper clip doesn’t do that. Also, Cliff Nass: Yes. I think that there are real- it manifests a particular personality style ly two problems—both are important, but that is not terribly popular; it’s a rather dom- only one of which is interesting to myself inant, unfriendly personality type. Those are and my lab. One is a very hard problem. some of the reasons. How do you detect emotion? How do you Now with that said, there are characters in identify what a person’s emotion is? The there that lots and lots of people like— other problem is—what do you do about it? unfortunately, the interface was designed so We recently proposed an interface that did that you couldn’t discover them. But if you one of three things. When it measured your go in and look for the other characters, the emotions, it either said, “This computer has research shows that almost everyone finds a detected your emotion. It is such and such, character that they like. Its lack of memory is and it will adapt to that emotion to make still annoying, and that’s a problem, but the experience go better.” Or it said, “This there’s a range of characters in there; unfor- computer has detected your emotion.” Or it tunately, that was hidden in the interface. said, “This computer has detected your emo- Bill Buxton tion. It’s going to really impede your perfor- Richard Anderson: When we swear at a mance, so get with it and fix that!” Now, all computer, should it swear back? three of those responses can be really appro- priate in various contexts. Hence, the hard Cliff Nass: It depends on what you’re using question (I think) is not, “How do you iden- the computer for. This is important. There tify people’s emotions?” but “What do you are cases where we want the computer to do with them once you know them?” Again, seem to have high status relative to us; there I think the psychological literature has are some cases in which we want it to seem answers to those questions.

66 interactions...january + february 2000

article

Richard Anderson: Bill, let me shift over that is, what’s changed, they’re probably to you. A lot, although certainly not all, of going to list the following: that they’re your work has been focused on input smaller, they’re faster, they’re cheaper, devices. Give us a brief overview of some of there’s more of them, they’re networked, that work—but particularly its motivation. they can sense location and motion, and so on, and finally, the input/output devices are Bill Buxton: This will be a bit of revisionist changing. history, probably, since what I say today is Now, I’m pretty sure that the answer that I probably different than the story I would would get from almost everybody in the have told you 20 years ago. audience, or anybody they know, would be I believe in natural language systems, but about the same for the first few items on the I also believe that body language is part of list—smaller, faster, cheaper, and more of Clifford Nass that natural language. In our use of manual them. However, when you get to the bottom input, primarily in interacting with comput- of the list, things are going to be fuzzier. This ers on the input side, we were basically is significant, since I believe that in terms of handicapped—literally with one hand tied design and where we’re going, my list was behind our back for the most part, and with- given in inverse order of importance! That is out the use of all of our fingers except for to say, for anybody working in HCI, the most one. Our ability to articulate using body lan- important things are the input/output guage was just horribly restricted. devices. The second most important things I’m a musician. I started out just trying to deal with location and motion sensing. (And make better musical instruments and trying I mean “location” in a social sense as well as to make art with computers; that’s really a spatial sense, so I know my location relative where my interest in input devices—and a to this other device or in terms of this person lot of the insights—originated. or in terms of other things around me.) The As my experience grew, what I found was third most important: the networking. Every- that it wasn’t just music systems that thing else is actually completely boring as a neglected input. The same was true whether research topic to anybody in HCI, because you’re doing a drawing, doing a spread- there are millions of people working on sheet, or anything else. Graphics got all of smaller, faster, cheaper, etc. That stuff is the attention—this is funny, because I work going to happen anyhow, so why waste your for a computer graphics company. There was time on it? a discrepancy between the bandwidth that As I said, of greatest importance are the the computer could use getting to me and input/output devices. Coming back a little what I could use to get to it, as well as the bit to Clifford’s way of speaking, it’s about richness of the vocabulary that I could apply first impressions. It’s what you see, feel, and over the same bandwidth. hear that shapes the core of your mental It always has struck me that if you wanted model of the nature of the beast. When I to make substantial change in user inter- give lectures, I go around and ask people to face, first of all, change the devices, and sec- draw quickly, in 15 seconds, a picture of a ond of all, focus on the input devices. Six computer. I have over 3,000 of these things months of research and a couple thousand in my filing cabinet, and over 80 percent of dollars focused on input will get you way them don’t include a computer; instead, further ahead than 10 times that investment there is a CRT, a keyboard, and a mouse— on graphical output—just because of the the input/output devices. Those are the relative stages of maturity. So, in my own things that shape the mental model. If I had work, I went where the weak spots were. asked people to do the same thing in the Today, I’d give you a slightly different ‘60s, they would have drawn key-punches answer, and this ties in with some of what and some other things; again, the computer Clifford was talking about. Today, if you ask would have gotten left out. people what’s going on with computers, That leads to a most powerful realization:

interactions...january + february 2000 67

the most powerful thing shaping our mental pencil effectively. But what happens when model is an accident of history and is, there- she sits down in front of a computer? She fore, a candidate for change. If I have a gets something that is unworthy of a grade- good sense of the mental model I want to school child—because it’s a general-purpose create, in terms of the right model for the device—it’s a 15-dollar device. You need right person at the right time, by designing something to capture the skill of somebody the affordances of the I/O transducers, I can who has the skill of drawing. And the trans- conjure that model up in the user’s mind. ducer to capture the skill of an artist is very, In my view, the only good computer is an very different than the transducer needed invisible computer. So, if a user is conscious to capture the skill of a court reporter or of of interacting with a computer, it represents somebody operating a car. a failure of design. So, a lot of the issues that My view is that very strong, specific Clifford is talking about start to not only devices are how you make computers disap- challenge our thinking, but also to provoke pear, and how you change the computer questions as we move into the family rela- from something that gets in the way into a tionships that we would see in ubiquitous prosthesis that helps empower us and actu- computing and embedded computing— ally makes the world simpler. especially when you start thinking about There’s a very simple litmus test for me in foreground and background types of inter- design. Does the design make the world eas- action. Most of the social interaction that ier to navigate and function in, or more Clifford and Byron speak about is around complicated? If it makes the world more the foreground and not the background— complicated, it’s bad design. which is actually one of the main areas for future change. Richard Anderson: Does the CHI commu- nity understand design? Richard Anderson: Let me follow up with a related input device question. You Bill Buxton: The CHI community isn’t a have argued for a need for a collection of design community. That’s not an insult, by strong, specific devices, each best suited for the way;. There are lots of communities that limited tasks. What do you mean by that? aren’t design communities. I don’t believe that the CHI community is primarily con- Bill Buxton: In evolution, whether we’re cerned with design. This is manifest in many talking about biology or almost anything ways, such as in the very small interaction else, strong, specific systems have typically between the CHI community and, for exam- won out over weak, general ones. I believe ple, IDSA [the Industrial Designers Society of the same is true in user interfaces and inter- America], which is concerned with one acting with computers. The notion of a gen- aspect of design. It’s not a criticism. It’s an eral-purpose computer and general-purpose issue of not having found a way to work workstation is simply an obsolete notion. together. To sort of paraphrase some of what Clif- Let’s look at priorities. If you talk to com- ford says—it’s extremely rude. The skills that puter manufacturers, they’ll tell you that you and I (and everyone else) have acquired, they can’t afford to spend more than a few Bill Buxton and on which we base our sense of self and dollars on a mouse or a tablet or to manu- our identity, are completely ignored. Our facture a keyboard. Yet, a good watercolor systems, by their design and by their physical artist has a set of sable watercolor brushes, manifestation, are contemptuous of what the collection of which costs more than the we’ve spent our lives learning how to do. total cost of a Macintosh computer. The cost People keep saying, “Why can’t computers of a single violin bow of a professional vio- be as easy to use as a pencil?” Well, my wife linist—the bow alone, not the violin, just the went to art college. She spent something bow—costs more than a Silicon Graphics like 18 or 20 years learning how to use a workstation. Somebody who’s highly skilled

68 interactions...january + february 2000

article

is willing to do whatever it takes to acquire There are lots more examples like that the technology that will be able to capture that are going to emerge. Things will evolve, the subtleties of nuance of gesture or of but there’ll also be a lot of casualties along whatever, just to accommodate the skill that the way. they’ve spent years and years acquiring. It’s the contempt for that skill that’s manifest in Richard Anderson: Cliff, I wonder if you general-purpose devices, which are basically would like to respond to Bill’s comment jacks-of-all-trades, masters-of-nothing. That about Microsoft versus Apple. frustrates the hell out of me. Cliff Nass: Yeah. That was something that Richard Anderson: Will the day come was, in fact, reported in the press. A couple Clifford Nass when collections of strong, specific devices of reviewers thought it was an Apple prod- will be easy to obtain? uct and said, “This is great, another break- through in design by Apple,” but when they Bill Buxton: Absolutely. But, ironically, found out it was by Microsoft, they sort of good design is a really thankless task, reamed it. because if it’s really good design, you prob- A related notion is that it was a foray that ably don’t notice it in many cases…it just dis- said, for one of the first times in computing, appears into the ecology of your everyday “This is not for people who love computers. world. The things that don’t work are the This is for people who have tasks to get ones that shout out at you. So, failures dom- done.” I think that the incredible reaction of inate your perception, because successes just people in the computer community was become invisible. We are starting to see partly due to their feeling that they were strong, specific devices coming out and being left out. In fact, we went back and working very well. Though when I say very looked at the literature that came out when well, it’s relatively speaking. automatic transmissions first appeared, and I felt sorry for Clifford and this whole you read things like this: “Automatic trans- thing with “Bob,” and I’d like to use that as missions will allow anyone to drive. Instead an analogy to talk about introducing new of needing to go to courses and understand ideas. First of all, I’m curious how Bob would how the car really works, any ignorant per- have been received in the press and other son can get into a car and use it. In fact, even places had it come from Apple rather than women will be able to drive this new auto- Microsoft. I think a lot of the negativity had mobile.” The extent to which that was nothing to do with the design at all—it had exactly much of the press on Bob was to do with where it came from. Regardless, remarkable. “Why, they think you can use a why would we expect a new paradigm to computer without taking a course!” have been gotten right the first time or ear- Bill is exactly right—there is incredible val- ly on? One of the fairly successful new ue in designs that meet a particular need for approaches out there right now is the Palm the expert. But there is also incredible value Pilot. But what’s interesting about that is in designs that meet a need by working for that everyone in the audience probably can people who want to know nothing about make a list of the companies that have gone it…who don’t want to study for one minute. bankrupt and failed in doing pen-based Both communities need to be helped, and PDAs. Yet, despite a litany of failures (one Bill’s idea of specialist devices actually helps after the other), one came out that in many both communities. But the primary ox that ways is indistinguishable from things avail- was gored with products like Bob was say- able 10 years ago. But Palm got a few little ing, “This isn’t a game that you guys who subtle details right that made all the differ- build computers can play in.” I think Apple ence. It’s still a terrible device, but it meets could have pulled that off better than threshold…it actually does what it’s sup- Microsoft, but it was a startling and upset- posed to do. And it will still get better. ting notion.

interactions...january + february 2000 69

Richard Anderson: Bill, some of your about the same level once they’re complete- other work has involved the use of 3-D inter- ly obsolete, and at the same time, very few faces. Jakob Nielsen started one of his fairly of us will have spent time saying, “I’m recent Alert Box columns with the words, assuming that the graphical user interface “2-D is better than 3-D, because people are and the Netscape browser and the CRT are not frogs.” Is he correct? obsolete. They’re not going to be around much longer. What’s next, and how does 3- Bill Buxton: Maybe if he lives in South D (and audio and gesture) fit into that new Park. (audience laughter) world?” Regarding some of Clifford’s notions about Richard Anderson: Is there a large role our relationships with computers—what for 3-D on the Web? happens when the computer is actually just the large whiteboard that I walk up to and Bill Buxton: There’s an assumption that looks like any other whiteboard? Will embedded in your question as to what the we have the same kinds of relationships that Web is. Do I think 3-D has a large place in he and Byron talk about in their book when the Web, as we know it today? No, probably I have a very strong, specific thing for teach- not. The number of people that work with ing up on the wall that looks and feels just VRML [virtual reality markup language] and like a whiteboard? I’m taking notes right the success there have not been particularly now on my Cross pad—it’s just a pad of great. But the notion of Web browsers paper and my pen. How is my relationship today is totally boring and just not interest- with that machine different than the rela- ing to me at all. tionship my great-grandfather had with a When I think about 3-D and networks and pad of paper? When you break out of the sharing and so on, I think 3-D has a huge box and consider these other forms, the per- place. You know, it’s the old cliché: if the ceptions of the machines change, and the only tool you have is a hammer, then you see devices change. And with these new percep- the world as a bunch of nails. If the only tool tions and the new tasks and contexts that you have is a Netscape or an Explorer brows- come with them come opportunities to er, both of which are identical conceptually exploit things like 3-D and audio. So, it’s hard (going through a CRT and a keyboard and a to talk about the representation without mouse), then you see things very differently talking about the language or the messages than if you changed the nature of what the and the types of models that you’re trying to transducer is. What you are doing would communicate with those representations. change, the content would change, the loca- tion would change, the adjusters would Cliff Nass: I agree with Bill. But a warning: change, and the values of different external the problem with radically new things is that representations (visual representation of 3- the first ones are usually atrocious. The first D) would change as well. It’s the lack of movies were atrocious. The first books were breaking out of those constraints and those boring and atrocious, though I suppose the assumptions which is inhibiting to CHI. Bible doesn’t count. For the most part, when Jakob writes provocative things, but this we try new things, our first attempts are Bill Buxton provokes me in the wrong direction. The much worse. Part of the lesson in Jakob’s “2- thing that worries me to death is that if peo- D versus 3-D” claim is that it’s a lot easier to ple listen to something like that, we’re do rotten 3-D than it is to do rotten 2-D. And going to end up with a situation similar in it’s a lot harder to do great 3-D. nature to the situation where we didn’t The problem for a community that’s in a understand the line-oriented text editor world where things change rapidly is that it until after it was obsolete and we were all tends to become a very reactive community. into graphical user interfaces. We’re going When bold new interfaces appear, we imme- to understand graphical user interfaces to diately say, “Oh, it’s terrible because of this…I

70 interactions...january + february 2000

article

don’t have to study it any more. I don’t have or that Stravinsky had with his pens and the to try to draw understanding out of it.” So, kind of paper and ink he used… clearly he the depressing notion is that as these bold had an emotional relationship with them, ideas appear, the first ones aren’t going to be which actually was consistent with the con- as convenient to use as a Netscape or Internet tent…so, form and content had consistency. Explorer browser. Therefore, the community Look at the relationship one has with a real- tends to ignore them as opposed to looking ly great car. These are clearly emotional for the big ideas. things that go beyond what the whiteboard As far as what happens when whiteboards did. Which begs a question: can we make a become computerized…it’s a lot like what whiteboard that just begs you to come up happens with people in low-level service and do algebra? (audience laughter) industries in the United States. They become Clifford Nass relatively invisible, because the jobs they do Richard Anderson: Cliff, are we looking don’t require us to have a rich social interac- at Bill’s picture as we would look at Bill if he tion with them. A whiteboard that just takes was actually up on the stage with us? information is like someone to whom you say, “please pick up this book for me.” The Cliff Nass: Well, I was thinking about that. person goes and gets it and brings it back, Bill has the advantage of being much more and you may say, “thank you” thoughtlessly, handsome than I am, and he has chosen a but there’s no social interaction there. It’s particularly good picture of himself, and I’m when the thing has speech or some other stuck with what I have! On the other hand, modality that cues social responses—then and on the cheerier side for me, a study on it’s not just a whiteboard. So whether things synchronized speech done by Byron Reeves are psychologically present or invisible is and Dave Voelker says that the fact that my something that we navigate in the social-to- lips move while I talk gives me an edge over social realm as well as in the social-to-tech- Bill in that you think that I’m more honest nology realm. and more intelligent. His words sort of win the day for him anyway! (audience laughter) Bill Buxton: In your book, you talked about arousal. The measure of maturity Bill Buxton: I’m not even going to touch might be when the thing that arouses me that one! when I’m working with a whiteboard is what I wrote on the whiteboard rather than Richard Anderson: Bill, who has influ- the whiteboard itself. enced your work and your perspective a great deal? Who are the people you have Cliff Nass: Yeah, I think that when tech- greatly admired? nologies mature, they’re either radically social or radically invisible. That’s very much Bill Buxton: There have been a few peo- like the roles we have in society. I think the ple. My interest in user interface design was middle ground of where technologies go greatly influenced by Ron Baecker, one of from foreground to background just isn’t a my first mentors; I probably learned more sustainable model in the long run. There will from him about user interface design than be highly foreground technologies, and there anybody I’ve ever met. And there are some will be highly background technologies. great people I worked with at Xerox PARC. But I’m going to actually step out of the CHI Bill Buxton: Actually, I’ll buy into that. community—not to ignore those people, Let’s take the case where we push things to but because they’re obvious, since I cite the extreme. Whiteboards aren’t finely craft- them in the literature. ed musical instruments. If you look at the One of the most influential people in relationships that watercolor artists have terms of input devices was a guy named with their favorite sable watercolor brushes Hugh Lacain, who, with Bob Moog and Don

interactions...january + february 2000 71

Buchla (both of whom I respect hugely), was every user?” I think the answer is “yes”— building electronic musical instruments in realizing that your theories may be the late ‘60s that I had the privilege to use revised… that you may find the third diary and to work with. He may have been the or the new experiment. I think that was a first person I met who was a luthier—an very important lesson. instrument builder. As for the second… When I started this A lot of the people who have impressed research program, not only the CHI commu- me are crafts people who have an incredible nity, but also my own colleagues, thought I ability to execute a detail in making beauti- was a crackpot and urged me to do some- ful things that are functional as well. I keep thing else (I was going up for tenure at the coming back to this notion of tools for time) and said, “This area is insane.” But I artists, partly because I started out making persisted, and experiment after experiment musical instruments—electronic musical didn’t work, and my graduate students all instruments. So influential has been the fact left me, except for one loyal student, that with a beautiful violin or watercolor Jonathan Steuer. I was so miserable that I brush or similar device…you want to touch went off on a hike with my wife, Barbara it, it begs to be touched. And the moment Levitt, and said, “Well, I’ll just clear my head you touch it, you don’t even have to play a and hike away.” And so I hiked, and all of a note—it reflects the respect for your skill and sudden, it started to rain, in June, which is your hard work that the builder…the luthi- impossible in California. So, it was clearly er…embedded in the device itself. It’s people only raining on me. And this was clearly the like that who have most stimulated me. bathetic metaphor for my life, and I was Also, a number of musicians who I worked about to throw in the towel when my wife with early on stimulated me and triggered said, “Why don’t you just do an experiment me. When you design something…when you’re sure is going to work?” And that one you’re the propellerhead who designed the worked and led me to lots of others that silly thing, and you know everything about worked.. So, she was a very big influence. it, supposedly, and then you hand the thing And the third… When I first encountered off to a musician, and they do things with the literature in computer science about your instrument or device that you never computer–human interaction, there was this imagined was possible…in terms of stimu- remarkable arrogance of, “This is what nor- lating respect and challenging you to go on mal people do, and this is what we comput- further…that type of experience had a er scientists, who are deep and thoughtful in strong influence on me. I was very lucky, ear- understanding the distinctions between ly on, to be exposed to that. computers and humans, do.” And, “Oh, those pathetic people—one day we’ll cure Richard Anderson: Cliff, same question. them and train them and make them take classes, and they’ll be okay.” This rubbed me Cliff Nass: There were three who I can the wrong way. This was a type of arrogance identify. The first was my dissertation advis- that seemed really suspicious. I was trained er, Jim Vandiger, who once said to me, “Do as a sociologist, so comments like that, that the best you can with the data you got. we’re different than them, always raised Bill Buxton Don’t try to really answer the question.” hackles. So, I really wanted to set out to find And he said, “Remember that the history of similarities among people, and things that the 13th century is based on two diaries, and were education- and arrogance-indepen- when they find the third one, there’s going dent. So, much of my work tends to look at to be a very different history of the 13th things that are the same whether you know century, but we shouldn’t stop historians.” what you’re doing or not. That really made me think, “Could you do That was the third influence, but I will say these experiments—try to capture basic that that has really changed. To the credit of notions, not test every product, not ask this community, this standing apart between

72 interactions...january + february 2000

article

people who are computer literate and com- Bill Buxton: I’d like to give a concrete puter knowledgeable, and those who are example that’s consistent with this, an exam- not, is no longer one of contempt or patron- ple that comes from some of the research ization. So, I suppose that if I entered the that we did at the University of Toronto with field today, I wouldn’t have that third inspi- the Ontario Telepresence Project. The under- ration. lying principle is that if we know the biases or the type of behavior that we want to Richard Anderson: Bill has stressed the evoke, then we can design the biases importance of making technology disappear through the affordances of the devices so in order to minimize its interference on that that’s the behavior you get. And one of human-to-human communication. Can the the strongest biases on our behaviors is the Clifford Nass technology be designed so that the interfer- bias of the path of least resistance. ing role that it plays be played in accordance We used to have these desktop video con- with social rules—perhaps making it a bene- ferencing units that you could use to enter ficial participant in the interaction? each other’s room to talk (this is stuff that grew out of work at Xerox PARC and Xerox Cliff Nass: Yeah. I always look for human EUROPARC). When Hiroshi Ishii spent time examples that are similar, and a human anal- with us, he never used the system. He would ogy is the superb translator. When you go to always walk down to my office, which was a foreign country to give a speech, and you two floors downstairs. I finally said, “Why?” have a translator, somehow the translator and he said, “Well, because it feels rude to (though doing something quite disturbing just jump onto your desk. There’s no way to and traumatic . . . changing your words and approach.” (Marilyn Mantei has talked a lot in some sense owning those words) can about this notion of “approach.”) So, I hung become invisible. The best ones become a small monitor, camera, and speaker above invisible; they become a mere “conduit.” the door, far away, which I could hardly see And as Bill said, the unfortunate part is that (but I could make things out), and the setup they lose all credit for the remarkable work gave a very wide-angled distant view of just they are doing, because good design does me and my desk. And I could also hear peo- get invisible. ple when they would appear there. The So, there are certainly ways to do it, and video circuit was placed in a location appro- much to learn from humans. How do butlers priate to the social function of standing at become invisible? How do messengers the door to see if they could come in, and become invisible? The issue is to think about, because of its distance from me, it degraded “What would we want a person to do to my ability to see the person. It also degrad- facilitate this interaction?” Sometimes it’s, ed their ability to see me in detail, but it “go away,” such as when you get fixed up on enhanced their ability to see global context. a blind date, where instead of just being told If it was OK for them to move in, then they to appear at a place, you’re brought to could move up to the desk where they got a someone’s house for dinner. How does that very different type of view. So now there person gracefully start vanishing into the was a transition, and each location had a dif- background to enable successful interaction? ferent set of affordances that were appro- The person doesn’t keep on popping up and priate for the type of social interactions that saying, “Hi! Remember I’m going to facili- would take place if there was a live person tate this social interaction.” Instead, the per- there. son starts things off, then smoothes the way By adding more technology to the room, through various clever tricks. Those are the but the right technology in the right place types of skills that technologies can have for the right function, the apparent technol- that will facilitate interaction. But again, as ogy and intrusiveness of the technology Bill noted, the designers of those systems were greatly reduced; it didn’t interfere, but won’t get the credit they deserve. there was much more technology there. And

interactions...january + february 2000 73

each of the pieces was bad at some things, meaningful. just as each was good at other things. The By the way, when I talk about the novice- design grew straight out of the social or to-expert transition, I don’t mean novice to anthropological analysis of the architectural expert as an artist or as an animator or as an space and the social functions that appear at accountant—I’m assuming you’re an expert different locations in that space. there. I’m talking about your ability to uti- I got really excited when I read The Media lize the tools. Equation, because I wasn’t familiar with that The term “user friendly” is a marketing work. Clifford and Byron were taking a slogan which I stay away from—just because descriptive approach to things rather than it’s become meaningless. Things in our lives, saying, “This might be relevant to design.” whether they be buttons, shoelaces, saxo- We had sort of stumbled onto this stuff. We phones, pianos, guitars, or automobiles… didn’t have any of the theory behind it, but none of those things are “easy to we were applying it to design. So, my read- use”…none of those things meet Ted Nel- ing of Clifford’s book sort of said, “Yes, son’s “10-minute rule,” which claims that that’s why it worked; that can explain what you should be able to be functioning and we did.” It was like finding the other half to working in 10 minutes. Ted is a charming something we’d been struggling with for guy and is entertaining, but that’s a stupid more than 6 years. statement to make about tools (if, in fact, he ever really made it). If you believed that user Richard Anderson: Bill, you have often friendly was important, all the things I just argued that it is appropriate to design listed, like the violin, the piano, and the car, things to accelerate and encourage the tran- would be obsolete, because they aren’t user sition from being a novice to being an friendly in the sense that you can just walk expert, instead of designing something to up and use them. Yet, that’s what most peo- be user friendly. ple mean when they talk about user-friend- ly computers. Bill Buxton: Yeah. When making the transition between novice and expert…should there be a Richard Anderson: Might Cliff’s work smooth curve? Should there be a gradual suggest that things need to be user friendly transition? Should what I do in 30 stages of while guiding the growth of expertise? problem solving be developed and rein- forced to help me later in the expert stage? Bill Buxton: This is funny, because we just Absolutely! And should I be swearing that had a big argument about this today. Were whole time? Absolutely not! So, in that you eavesdropping? sense, yes, they should have a degree of My concerns about the notion of user politeness and civility. But I would never use friendly have to do with the question of, the term “user friendly” any time, any place, “What the hell does ‘user friendly’ mean?” If except to rant about how much I don’t like you take a sociologist’s perspective or a the term. behavioral psychologist’s perspective, you might be able to quantify it and measure it. Cliff Nass: I agree with Bill. In fact, there Bill Buxton I talk about enhancing the ability to make are things for which your friend is not the the transition from problem solving behav- right person to teach you. Some of my great- ior to skilled behavior—novice to expert, est teachers were not and never could be my because I know what those things mean friends; in fact, there were domains in which quantitatively…I know how to measure their harshness and ruthlessness really those things (well, actually, I don’t, but I’m helped me to achieve my goal, which was to lucky to work with smart people who do). become expert in those domains. But we know how to measure this stuff, and So, is the goal to instantly make people we can actually make specifications. And it’s happy (which is a perfectly fine goal)? Or is

74 interactions...january + february 2000

article

the goal to get people to use a particular than when I was at Xerox PARC or at the tool successfully so that in the long run they University of Toronto. The strategies and the will be happy? Long-run happiness and approaches are very, very different. One of short-run happiness are often competitive, the reasons I made the transition to the and neither one is always the right thing to company I am now at is because I wanted to do. I don’t know of any expert in any area understand both sides. Technology trans- who says, “Gee, the reason I’m an expert is fer—for things to get out from the time because everything was so easy and every- they’re known to the time they’re in the one was just so supportive.” Sometimes you public—continues to take 15 years. I’d like to need pushing. My graduate students tell me reduce that to about 7 years. I err perhaps too much that way. Nonethe- I also want to find out how to have the less, it’s not always the right thing to try to skill of user interface design understood so Clifford Nass be nice. The nicest thing you can do for that people will respect it in the same way someone is to help them meet their goals. that they respect the skill of hacking an Different people do that in different ways, operating system or designing a micropro- because of individual differences and cessor. Since the skill of design is not well because different situations require differ- understood, everybody is an expert, and ent techniques. they all have an equal vote. There’s no oth- And learning things in 10 minutes… I er discipline that I’m aware of where every- don’t actually think Ted meant that. Ted says body has an equal vote, regardless of their many things he doesn’t mean, and I think skill or expertise. So, one of the ingredients that was one of them. that will bring us to better design within companies and other organizations is get- Bill Buxton: Unlike us, right? (laughing) ting to the essence of what the core ele- ments of design are that we have to put into Cliff Nass: Right! We stand by what we place…in terms of the organizational struc- say, Bill! (laughing) ture of our teams and so on. Graphics design is really important, and a Richard Anderson: Yesterday, Clement lot of people work on the visuals, but very, Mok and Jacob Nielsen argued for a need very few people work on interaction. The for information designers, brand strategists, feel…the actual aesthetic of the gestures user researchers, software engineers, and all (the connectivity and the body language in sorts of other disciplines on a Web design that), and the dynamics of interaction, and project. I’m interested in learning what dis- the timing and the tempo, and the fluidity, ciplines the two of you emphasize the need … this is the most neglected part…this is the for. area in which we have the fewest people who are literate. Therefore, we have these Bill Buxton: I can’t answer that question jerky, disconnected sentences in the body without qualifying it. Am I at a university? language of our interactions with comput- Am I at a research lab? Or am I actually ers. building and designing products for people who are trying to do serious things (and art Cliff Nass: I work very closely with a com- is serious, by the way)? Am I trying to make pany, Netsage, that says, “To do design, you tools that ensure that the people who work need social scientists, artists, and technolo- for my company will still have a job tomor- gists.” I think that’s a pretty good formula. row and ensure that the artists who use the I want to emphasize a point that Bill made. technology (in our case, the animation sys- There’s been this idea that somehow design tems) have the best tools available? Where I is this richly democratic process where we all work today, the reality of trying to do good listen to each other, we’re all expert, we all design and get innovation and new con- take each other seriously. I don’t think that cepts into products is very, very different is fully true. I think a better model is that we

interactions...january + february 2000 75

figure out who knows the most about an problem. How do you do good design and area and listen to them most closely. The get great user interfaces with economical artist should listen to the social scientists teams? That seems to be a really important about social science and to the technologists challenge. And it’s a challenge which is all about technology…be informed about the the more pronounced when we’re still in an true constraints and the opportunities. But era where 95-plus percent of the computer the artists have to be trusted to do art; social science graduates in the world today still scientists and technologists shouldn’t be graduate without ever having written a pro- doing art. Nor should artists be doing tech- gram that was used by another human nology. Since it’s really hard to be good at being, much less being marked on their abil- one of these things…let alone two or even ity to do so. three…I think that division of labor makes sense. This idea of participatory design is a Richard Anderson: Why is it difficult for good one, but not when it means the abdi- people to trust the expertise of another? cation of expertise. I was once teaching a class at Stanford, Cliff Nass: The problem has been that which often has these democratic impulses, problems in computer science have histori- and I said, “I can’t remember who said this cally been defined as technology problems. particular quotation: A or B.” One of the stu- As the technologies become simpler and dents raised their hand and said; “Well, lets more accessible…so that you can use com- take a vote on it!” That idea about voting plicated programs without being a pro- about facts…about voting about what’s grammer, trust becomes equalized. In the true, while it is charming and feels good, days when the only people who could read doesn’t result in the best designs. You have were the priests (who were literate), nobody to trust people at what they’re really good at else had much say about anything. They and interject stuff that you’re really good at. could claim a monopoly on knowledge that was hard to challenge, because they claimed Bill Buxton: I’ve spent quite a bit of time a particular skill that was so compelling and in my current job around filmmakers. Film- so unique, that it implied knowledge of all making is clearly a highly collaborate skill other things. In some sense, that’s been the involving tons of people with lots of differ- history of the field of human–computer ent skills, but the fact is that there is one interaction. director…one who has final say. The way we As computers get cheaper, easier, and do a lot of user interface design is that faster, and software gets easier to use, no everybody on the executive team has an group can claim the core competency from equal vote. If we made films that way, films which all power derives. As that happens, would be like our user interfaces, and peo- the social things will work themselves ple wouldn’t line up for a month to watch out…namely if no one can claim a core and them! critical position over all others, then they I wish I could just wave a magic wand and can’t claim core competence in all areas. say, “Here’s how to do it.” But when you’re That’s a very common sociological phe- actually trying to build a commercial system, nomenon. So, as more people can use this Bill Buxton it’s rare that you have the luxury of having stuff and the priesthood breaks down. you’ll that full suite of talents that Clifford and I start seeing a lot more trust and coopera- sort of enumerated. And we left out tion. some…we left out the usability testing peo- The early days of film weren’t as cheery ple and the QA people and even others. All and remarkable as now. The cameraman, of a sudden you have such a huge team that especially the person who could build and you will never get a product to market, fix the camera, had an enormous amount of because either the team is too big to func- power and would often co-opt that into tion or you simply can’t afford it. That’s the areas, such as directing, in which they might

76 interactions...january + february 2000

article

not have any competence. Eventually, cam- that are wrong—that are too simplistic. For eras got so reliable, so easy to use, that example, at a workshop of yours that I those guys eventually lost their status or the attended some time ago, you argued that best ones retained their status for that par- people should be focusing on interactive ticular role. So, I think it’s a matter of the perceptualization instead of scientific visual- technology becoming stable enough for the ization. Also on multisensory instead of mul- priesthoods to vanish. timedia. What concepts are people focusing on now that are too simplistic, and what Bill Buxton: I think there’s an additional should they be focusing on instead? issue. You need to know something about something before you know what you don’t Bill Buxton: I think that there are a num- Clifford Nass know. When you have a product to design ber of things. We’re overly simplistic about or something you’re trying to build and get our view of how the Internet and putting into the hands of the ultimate user, you computers in the classroom are going to rev- want to move…you want to be efficient. olutionize education. And I still question, in What you’re trying to do is the equivalent of terms of epistemology, whether we know trying to do mathematics, and it’s difficult to anything more than we did before that can do mathematics if you have to explain arith- be applied to the technology. For example, metic all the way along. It’s the people who do we understand the difference between a aren’t even literate with the basic opera- whiteboard and a slate? tions of arithmetic who believe that they The fundamental problems are not tech- know as much about math as you do. nological in almost any domain in which I don’t want to sound like an elitist, but we’re trying to apply technology. The tech- education helps you build discrimination. no-centric view, as opposed to a human-cen- You start to see differences that weren’t tric view, is most simplistic thing. I think there before. We know this about the histo- always of the technology as a prosthesis— ry of photography in terms of resolution. whether it be a physical prosthesis or a cog- We know that when Fisher played the first nitive prosthesis or a sensory prosthesis. To stereo, everybody said it sounded exactly build a prosthesis, you need to know and like an orchestra, and we said the same have a theory of the function of the task it is thing about CDs. As we got more expertise intended to support. Then you can design its and granularity, we started to realize that affordances in accordance with that model. they don’t sound exactly like an orchestra. We often cannot articulate the model that’s Think of the paintings you saw in an art his- underlying our designs, because we just do tory course that made you say, “I could have the design without the model. If I want to done that.” As you start to get some litera- make a system for drawing, I had better cy, you realize, “No, I couldn’t have.” The understand drawing, not computers. point is that there is a basic literacy that has to be achieved before you start to see some Richard Anderson: You’ve also argued of the finer granularity to how to make deci- that we lack respect for human capability; sions and to why to think this way or that indeed, you’ve stressed this already today. way. The basic assumptions that we need to What human capability do we lack respect all share so we can get on with the job are for that results in our lack of respect for not there right now…they are there in more human capability? Or to put it a little differ- mature areas. But even art still suffers from ently, what human capability do we lack this problem, so we shouldn’t feel picked on. respect for that makes it difficult to correct We should just recognize the phenomenon our lack of respect for human capability? and start to figure what to do about it. Bill Buxton: It’s interesting…you can go to Richard Anderson: Bill, you have talked any bookstore and see walls of self-help about how people often focus on concepts books about finding your inner self. I some-

interactions...january + february 2000 77

times wonder whether this is rooted in the systems. The more specific the device, the fact that a lot of us just aren’t aware how more I know about what it is, who’s going to we derive our sense of self, and what we’re use it, context, and so on; therefore, the proud of, and our identity. In many cases, it more appropriate the response can be. So, I has to do with the things that differentiate would argue that your father’s computer is us from other people. Those are the things going to be a different computer than your that we work hard at, because usually those child’s computer versus your boss’s comput- differentiators are things we have had to er, because they will be distinguishable by work hard at. We probably have had some the physical and social context in which they aptitude that we worked hard at, whether it appear and where they are used and the be the kid who’s the best math, the kid behaviors that will be embodied. who’s the best at music, the one who has the The biggest problem with computers best calligraphy, the fastest runner, and so today is that they all have personality disor- forth. If we start to understand that this is ders—they would all have to be on Prozac if what’s important to me… this is my sense of they were human beings. One minute I’m a self and identity… and that everyone else word processor, the next minute I’m killing has something like that, then the best thing people in a game, the next minute I’m in e- that I can do as a designer is design things mail, and the next minute I’m sitting there that amplify that, rather than ignore it and for erotic pleasure—what am I? If you look denigrate it—and as a result we’re going to at strong, specific devices such as a personal get great design. If many of us don’t see computer which we commonly call a cam- what it is in ourselves that’s important, then era…it knows that it is a camera; it knows how are we going to take into consideration it’s not a word processor; it behaves like a what is important in others in our camera; it knows that its function is to have designs…so that our designs amplifies what light in and chemistry out, or light in and is important in other peoples. If we don’t see pixels out; therefore, it behaves appropri- it in ourselves, how are we going to see it in ately for its form and function and location. other people? Again, it’s not a technological It doesn’t have so much ambiguity, and the issue; it’s just a question of what makes us more I can let it know what it is. what we are. (audience applause) Look at a flatbed scanner. A flatbed scan- ner has a personality disorder—it doesn’t Audience Member: A social perspective of know what the hell it is. But if you look at interaction would suggest that people the physical affordances, what the hell is it respond as they respond to people. I for if it’s not for scanning flat documents? respond and interact with my mother very Yet, it doesn’t know that it’s for that, and it differently than how I interact with my doesn’t know anything about flat docu- daughter. With my daughter, I “emote” a lot ments. And when you put a document on it, more; I’m very accentuating/emotional. it can’t say, “Hey, you’re a color image—I’ll With my boss, I’m very reserved. I differenti- do you in 24 bits,” or “You’re a grayscale—I’ll ate, and I treat people as individuals, and do you in 8 bits,” or “You’re line art, so I’ll do I’m successful, because I can do that. I’m you in 1 bit,” and things like that. It can have trusted, because I can do that. Do interfaces this kind of knowledge and take advantage Bill Buxton need to be able to differentiate, know the of it. Xerox did some really interesting work individual, and adapt accordingly or do they on this for a product that shipped for awhile, need to continue to just seek the successful but it hasn’t come into the mainstream. As central tendencies and serve the mass popu- soon as we start having devices that know lations? what they are, you will behave with them differently—in exactly the same way, I would Bill Buxton: You played beautifully into argue, that you differently behave with the the argument I’ve been trying to make family and the other people in your social about strong specific versus weak general network that you enumerated.

78 interactions...january + february 2000

article

Cliff Nass: The irony of that is that those sole purpose is to take a measure of your clearly defined devices that have a clear role, arousal. Again, it is exactly what Clifford and a clear point of view, clear orientation, … Byron talk about in their book. are in one sense easier to design for, because And they’re useless for word processing! of the positive feedback that it ensues. If I (audience laughter) have a clearly defined device, such as a cam- era that works like a camera in all the ways Audience Member: Within industrial design, cameras should work and does all the logical design is a process whereby function and things, not only will you as a user find it eas- form get defined. I wonder to what extent ier to use, but I’ll be better able to predict you people feel that we are taking advan- what you’re going to do. Therefore, I can tage of this model. To what extent should further optimize, because I will not only we be addressing the functionality and Clifford Nass know what the initial state is, I’ll know the interface issues at the same time…in other progression of that interaction. words, dealing with usefulness at the same This is true in voice systems as well. We’ve time and usability? been doing studies on things like using a particular voice type to elicit certain voice Bill Buxton: I think that both the CHI com- characteristics. If you have, let’s say, a highly munity and the industrial design community accented voice speaking to you in a speech need to extend their design briefs in each recognizer, we know the types of speech other’s directions. The industrial design usu- patterns you’ll use. We know you’ll hyperar- ally stops when you get to the glass; indus- ticulate. We know you’ll speak louder, trial designers work in front of the glass, not because we all speak louder when we speak behind it. They don’t typically design the with people who we don’t think understand fonts or the icons—they’re working on us. Plus, I can leverage that to tune the rec- things like the box. Yes, I know there are ognizer to those particular characteristics. exceptions, so don’t beat me up. By knowing what’s going on, you not only Fifteen years ago, a lot of people doing can make the user happier, you also get to, user interface design and designing comput- in some sense, positively manipulate the ers actually took it upon themselves to user into doing things that you can then fur- change the physical nature of the box, ther optimize to and therefore further because they understood that it had an effect manipulate the user. It’s why we can carry on on performance. That is, to a large extent, conversations without thinking very hard. one of the lost arts. A limitation that I now It’s because there are a lot of basic patterns see in a lot of CHI research is that we take a that we execute quite well. priori and accept the fact that there is a key- board, a mouse, a CRT or an LCD, and a GUI Bill Buxton: Let me give you an example type of interface, and use that as the point of that goes even deeper than that. BMW and departure and just try to refine that. Now, a couple of trucking companies have been any industrial designer would say, “No, no, developing what is essentially a video cam- no, let’s truly take this notion of form and era with a microprocessor that fits in the function and have consistency from the dashboard of your car and watches you. The design of the fonts right through to the sole purpose of this computer is to get a design of the box (for the function of what- sense of how tired you are, because it actu- ever it might be), and let’s not have any of ally understands the behavior of your eyes, those things a priori. Both sides could try to eyelids, and your glancing in order to alert flow a bit into those opposites and just work you if you start to fall asleep at the wheel. a lot more together, but typically they’re edu- Here’s a case where a very strong, special- cated at completely different institutions. ized computational device will be embed- ded in automobiles that basically can be the Cliff Nass: Well, I think that there are difference between life and death. Their times when you want to accentuate form

interactions...january + february 2000 79

over function. I think Bill is right…that tion.” Fifty percent of what we do should be thinking literally out of the box, or thinking to go back and look at and learn from his- with the box—not just accepting the screen torical precedents and at what’s gone on in “as is,” is a great idea. But, even if you were the past. The problem is that we think we to do that, there are trade-offs between have to be original—that’s how we show form and function that become interesting. how clever we are. We have to invent things Frank Lloyd Wright houses, for example, are from scratch rather than do scholarship and great for form and function, but Frank Lloyd do research and synthesize. Wright furniture is impossible to sit in. The Right now, for example, everybody is try- furniture is quite beautiful, and that’s ing to do the uni-stroke alphabets on pen because he figured people wouldn’t be sit- based PDAs. In the process of writing a ting very much in his houses. So, we can book that I’ve almost finished, I was work- think about when should we worry about ing on the chapter on pen-based stuff. I just the form aspects as opposed to the function went to the library, did my homework, and aspects, or put better, think of form as mak- lo and behold, it turns out that the basics of ing beautiful things function—form being in Graffiti—about 30 percent of the alpha- service of a particular function, namely, bet—were invented in 63 BC by a slave of making people feel good. Cicero named Marcus Tullius. His shorthand, notee Tironianae, was used for 1,000 years PERMISSION TO MAKE DIGITAL OR Bill Buxton: The Macintosh came out in and was used to record the minutes of the HARD COPIES OF ALL OR PART OF THIS 1983; it’s now 1999. Don’t you find it Roman Senate. WORK FOR PERSONAL OR CLASSROOM

USE IS GRANTED WITHOUT FEE remarkable that at the end of the millenni- What I’m really trying to say is that a lot PROVIDED THAT COPIES ARE NOT um, the hottest, most progressive, imagina- of the ideas…a lot of the concepts…are out MADE OR DISTRIBUTED FOR PROFIT OR tive thing that’s happened is that iMac has there, and we tend to think that there’s COMMERCIAL ADVANTAGE AND THAT colored boxes? some magic bullet as opposed to the stan- COPIES BEAR THIS NOTICE AND THE

FULL CITATION ON THE FIRST PAGE. dard, old, traditional things of a combina- TO COPY OTHERWISE, TO REPUBLISH, Cliff Nass: But they’re translucent, Bill! tion of research, scholarship, and invention. TO POST ON SERVERS OR TO REDIS- And that’s us, that’s not the technology. The TRIBUTE TO LISTS, REQUIRES PRIOR

SPECIFIC PERMISSION AND/OR A FEE. Bill Buxton: …and they’re translucent… tools are there to do what needs to be © ACM 1072-5220/00/0100 $5.00 but otherwise, from 10 feet away, they’re done, and there’s no excuse except our- completely indistinguishable from anything selves. that existed in 1982. And that’s considered progress and ground-breaking industrial Cliff Nass: I’d add patience to that, which design—give me a break! (audience is something I alluded to earlier…the toler- applause) ance for designs that are worse overall and don’t work as well, but might have an Cliff Nass: You’re getting wild applause, insight embedded in them. We’re so con- Bill! cerned with just making things better over- all, whatever better means, that we tend to Richard Anderson: My final question to ignore those things that do one thing well both of you is the final question listed in the and lots of other things lousy. Learning from description of this session that appeared in things that seem to be failures, learning Bill Buxton the advance program. What is keeping us from things that in many ways aren’t any from achieving greater benefit from tech- better and are in some ways worse, would nology? be a good lesson.

Bill Buxton: Ourselves. It’s not the tech- Bill Buxton: We should have a quota at nology; it’s our own imagination. I think the CHI that 25 percent of the papers have to be closer you get to Silicon Valley, the more you describing failures and why they failed. confuse the word “invention” with the word “research” and the word “innova- Cliff Nass: Absolutely.

80 interactions...january + february 2000