registered authors login here For more on PMWiki, see pmwiki.org |
Dourish says, "Imagine what it would be like if any other technology had undergone such rapid advances in price/performance." This seems very rudimentary, but this really let my imagination go wild - disposable cars, space travel, augmented reality TVs, etc. So why not a new I/O scheme with a computer? I'm intrigued by Weiser's concept of "ubiquitous computing" and I feel that as technologists we have the responsibility to provide people with a better way to do things, even if they resist. But often marketing overtakes the ubiquitous computing ideal. If you buy a Mac or iPod, for example, you buy into a lifestyle - just look how good-looking and hip the people in the commercials are. You are buying the ability to sit in Washington Square park and compose a song in GarageBand or burn DVDs for your friends of your road trip to the world's largest donut stand. You have a sense of ownership, whereas an automatic door in a grocery store reacts to your presence and makes your life easier, but you can't own it. But to quote Tom from a Wiki discussion, "For me, I think it comes down to the fact that functionality lasts, fashion doesn't." I think he's right. The functionality of the keyboard and mouse has lasted. But people have to feel that a device fits into their "personality" in some way. I believe people hate to be sold, but they love to buy. Without the desire to use a technology it is destined to fail, and it seems that functionality on its own does not fill that desire. what are folks waiting for!? there's a great idea here that's been gathering dust since the 60's. granted, using the projector/camera combo (from above? easy fix) has fundamental OCR/AI problems. but why is math fundamental to computing? why not simply for graphics/layout? allow text as a rendered image, but no need to read it. imagine photoshop and quark/freehand totally devoid of language dependencies, manipulated by intuitive gestures. squeeze an img to make it smaller. no learning curve or steps to remember. the computer needn't read or understand anything but RGB levels. for those who like text entry, they can still buy a computer. for those who think in pictures, dislexics, non-native speakers, etc. a new Digital Desk would be an ideal tool. i bet there'd be a way to even create a rudimentary pseudo-IDE, in the way Flash isn't at all C++, but does support IF... THEN, FOR decisions. bet it could be combined with MAX. arranging a graphic flowchart from popup choices, no text entry required. (like Flash, it wouldn't need to do things like make specific calls to lower level functions, because that's not something that makes tool useful to the vast majority of these people) instead of naming files, maybe it would just recall the top 10 most frequent colors. you'd choose from a color wheel and it'd show all the files that use that color a lot. i dunno, but perhaps the move from text to graphics stopped only part way. Dourish wonders a lot why computers, though having changed specs drastically, haven't really changed in the way we work with them. (thiough that's maybe arguable. for instance, imagine the swarm of bloggers who did not feel privy to using a BBS, though these things are really only cosmetic differences about 15 years apart. the habits of users hasn't really changed but has shifted. the computer is still on a desk, but the desk is now in the living room, not just the office. however, ubiquitous computing in general is neat but could also easily turn into surveillance and big brother watching you. it puts a lot of emphasis on sharing info and avoids the issue of private info, and treats everything as either an all or nothing ownership. sometimes a tool shouldn't have that aspect, but give the choice over entirely to the user. a screwdriver has no inherant personal/public property issues. it could be used for either prying open a locked door or sealing a box shut, but mostly it is for completely neutral jobs. the super's brother tightening the screws on your front door actually sets aside the personal issues for a moment and simply concentrates on making the door work better. who it keeps in/out is an abstract construct of secondary concern to the user of a screw driver. having your screw driver engraved with your name is rarely considered and does not enhance the use in any way. whereas password protection is used (by osx and windows) to keep novices without admin priveledges from messing up the system. the neat thing about the Digital Desktop is that anyone could walk up and decide this is the tool for my job. the ownership aspect is really just for the IRS or theft, not intrinsic to the tool. The tool could still contain "profiles" but its successful operation is in no way addresses propriety. a refridgerator just stores food that needs to be kept cold. you CAN write your name on the milk carton, but few of us do. if the label is too hard to make out for a color blind person, the refridge keeps the milk cold all the same, is still useful to to them. there's no need to involve the text here. I saw a lot of ITP projects described in here: The feather project reminds me of the one in the main floor area and also Mark Argo's hugms, and the string connected to a motor that moves in response to ethernet traffic is a lot like Tom's own email/clock project. I really liked when Tom mentioned that a good tool should "recede," allowing the user to focus on the action/expression rather than the object being used for the expression. I don't know if I've done a good enough job of following this philosophy when making my own device, instrument, tool project, but I can see that the idea fits in very well with what Douris is saying, both about tangible bits and ubiquitous computing. I think an important take-away is that we should be making devices that take advantage of the benefits of virtual bits, but that also are well-designed (and specialized) enough that we are able to use them intuitively for the expression/action intended. I stole the receding line from Dourish, who says he took it from Heidegger. I remember the promise that computers (macs) held for me in 1987, tortured through my Art School days of inking velum, using plastic ellipses and flexible curves to ink a line only to smudge in because I am cursed to be a lefty. Then someone brought in a Mac SE30 and I discovered Illustrator (88). Suddenly my curves were bezier and quite impossible to smudge (if not to delete). I was quick to embrace, but not questions the "mouse" I just accepted that I had to draw with, let's face it a bar of soap. So, to me that is how we were all convinced... learn this way and life will be better. Alas, I still cannot draw better with a mouse than a No.2 pencil. I very much enjoyed the writing about Virtual reality vs. "Physical Virtuality". As cool as VR seems, I like much better the idea of ubiquitous computing. The multimedia computer is a needless block of inertia, we sit in front of it and it seems more as if we are serving it than the other way round. So what is the answer? I think that we would be happier if we could interact with the computer all over our home and out into the world, not by sitting down and staring at a screen but by dictating and making choices verbally and visually in the context of what we are doing. Each piece of the puzzle would fit with the other and the computations power would be specialized. The idea of these separate devices coming together to a "unified effect" is very attractive because there are so many distractions and small amount of wasted time because activities in our lives are not coordinated or remembered by any kind of helper device or application. The desktop metaphor has survived not because of it brilliance, but because we are not questioning it as much as we should. To be happy with the statis quo is not an indicator of a good design that has stood the test of time, but of a business model that supports staying in one place. As it's been said, "good is the enemy of great" I think the DataDesk idea is interesting, but there's two things that make it problematic: First, that the whole idea is predicated on the notion that we want to own our data, not our interface. DataDesk and so many ubicomp ideas, going all the way back to pads, tabs, and boards, rely on the notion that people will share interfaces, and data will migrate with them. As Gilad points out, that runs counter to the idea of marketing. Second, in order for systems like DataDesk to become practical reality, projectors need to cost less and LCDs and be as bright. They were nowhere near that at the time. It's conceivable that they will be in the very near future (they're getting there), but it ain't there yet. Sometimes I wonder if all this longing for super-interactive tools all around us is maybe too much. Can there ever exist a 'paperless office'? The technology is there already, but there still are some activities where we cannot substitute paper with anything else. Paper has many personal properties, many relating to our memory. Personally, when I look at a piece of paper, I tend to subconciously memorize layouts and position of text on the sheet. For me, this trigger of photographic memory happens is not as strong on the computer screen. I think that the trend of ubiquitous computing is a logical step forward, and am greatful to be a part of this generation which will probably go through many stages of rapidly evolving technology, until the technology itself will become 'invisible', as stated. But I do think that we have to be very cautious as to how we approach all this new and **** technology. Like David, I also am very interested in these two approaches to the relationship between computers, people and the world - Virtual Reality VS. Ubiquitous Computing. Virtual Reality is a great tool but at the same time extremely dangerous. I've recently been to a lecture about a fairly new software - Secondlife. Just from their title - 'Your World, Your Imagination' - its obvious to grasp the danger in this type of software. Seeing the demo of this software, I can easily imagine people sinking into these imaginary worlds because they are, in a way, a cooler place to live. The fact is that there are already people living in this 'Seconlife' world, and making their yearly salary through their virtual characters online (the most successful currently making 300K yearly). Just insane. As Frank Lantz points out, there's more people in World of Warcraft in any given hour than have ever logged into SecondLife in its entire existence. The road to SecondLife is littered with the corpses of other systems that have attempted to do the same thing: MUDs and MOOs, AlphaWorlds, Time-Warner's Palace, and many others that have attempted multi-user online worlds, 2D and 3D. I think they're interesting, have spent some time in them in the past, but frankly, the ones that succeed are the ones that are based on structured activities with simple rewards: in other words, games. So if you fear people sinking into virtual worlds, fear WoW. I've never heard of that game. I'll definitely check it out. gilad In response to David's experience of drawing with a mouse - I've been thinking about the image of computer use that I remember (hazily) from the film "Minority Report". I remember people interacting with a large glass stand-alone screen. Kind of like the screens we see actors use in films when they're pretending to be generals directing combat operations in a 'war room'. What I like so much about this interface is that the user is on her feet, and can interact with the computer by moving her body through space. It seems to me that this would be a great interface for designers, artists, etc. I tend to get stuck in my head when using a computer. The static position of the body can encourage tunnel vision in terms of creative thinking. There is no variety in spatial attitude - it becomes overfocussed and all direct effort (to use Laban language), no indirect effort. I feel like I could improvise with speed, surprise myself, shift point of view with this kind of interface. I could literally dance with my software application. In response to the digital desk, although I like the fact that this means that I would have a more varied and dynamic interface than screen/keyboard/mouse, it does still mean that I would be constantly staring at digital projection (probably in a darkened environment, so I can see the projection clearly), which might get to me after a while... I guess the ultimate in ubiquitous comuting is when the computers become invisible, and the environment is actually enhanced, instead of just replicated (in a more expensive and error-prone way). Perhaps this vision of ubiquitous computing needs to come hand in hand with progress in renewable sources of local energy? Those batteries have to be disposed of somehow, and after all, it's all going to be useless if the power goes down. In response to Gilad's comment about Second Life, I attended the same talk, and it kind of blew my mind, also... but I'm really not sure whether this is actually a social reverse (or advance). I've never been into gaming or VR, and I have no facts to base this on, but I'll hazard a guess that the people who like to spend a lot of time in Second Life are the same people who would have been playing a lot of Dungeons and Dragons 20-30 years ago, or spending hours reading Tolkein and Isaac Asimov. I'm really not sure if spending hours in Second Life is so much worse than spending hours in either of this other activities? They're all fantasy lands, and share the same degree of physical inactivity. In fact Second Life is more creative in a way, because you can generate the world as you go, and the only limit is your imagination. And the inhabitants of Second Life are interacting with each other. In fact, it looks like an amazing visual representation of all the desires, expectations and assumptions of contemporary culture. True, it's not pretty... but it'll make a good psychology thesis for someone. The Minority report interface is a popular one. Several versions of it have been made, both before the movie and after. At a workshop I attended this summer, I saw the first commercial version that I thought might last, basically a shared virtual white board with a GUI that was a cross between a normal windowed environment and DataDesk. One of the differences between that interface and the Minority Report one, though, was tangibility. You touched the wall to write on it, move objects, etc. I think tangibility is very important to reliable interfaces. Gestural capture is difficult at best, as Dan and Mushon have shown in their Google world project. There are many ways for a computer system to interpret a gesture. More important, perhaps, when we make a gesture in the air, the feedback from the system can never be as definite as the firm reliability of feeling a button resist the pressure of our finger, or the satisfying friction of a pen across a whiteboard. That said, I find gestural interfaces interesting, and hope to see more attempts to make them feasible. The most interesting point I took away from this article is that we are now 1/2 way from the "here" described in this piece and the "there" they suggest we'll get to. We have, often, several channels for data - nearly ubiquitous. We also have many means for ambient info displays and tangible computing. So what has worked and what isn't working, and how do we address the implementations of new technologies using the successes? I would argue that htere's no "there" to get to. It's a continual process of re-evaluating "here" and deciding whether or not it works for us. This might make me unpopular to more technically-minded ubicomp enthusiasts. Here are my notes from the piece: "Interaction with screen and keyboard, for instance, tends to demand our direct attention; we have to look at the screen to see what we're doing, which involves looking away from whatever other elements are in our environment, including other people." Ubiquitous computing stands to radically reinforce or alter this. How / why? I'm curious about what the ramifications are and how the incorporation of computing "all around us" can allow us NOT to spend more time swapping time interacting with people for time interacting with machines. A large part of the ubicomp ideal is that computing isn't a process that people engage in like writing, but that computation is something that aids us in other actions. For example, take designing a car, a process which (in the past) was done by laying the car plan and section out in real space, at full size, with masking tape. Car designers learn so much from direct modeling that drafting, whether it be CAD or not, cannot show them. The ubiquitous computing approach to car design would have us design a sensing system to record the car designer in the studio as she tapes out, moves things around, etc, rather than having her sit down at a screen. So when Weiser, Dourish, et al talk about spending less time interacting with machines, it's their dream that, though we are still doing so, we are not as focused on the machines. Whether that ideal is ever met is arguable case-by-case. "The rise of the personal computer - and, more broadly, of personal computing - was an attempt to break away from the ten-dominant paradigm of mainframe computing." The network is the computer, as they say. W/ web services are we returing to this, except on a much broader scale? "Why not put computation wherever it might be needed?" THAT IS THE QUESTION! What is "need"? Today's teenagers "need" a cell phone at all times; need is contextualized. "A single user might have, at his or her disposal, tens or more of the inch-sied devices, just as we might have many post-it notes dotterd around, stuck to computer screens, walls, books" No offense, but this line of thinking is horrible. It works w/ post--its because they're disposable, have high resolution, and are easily accessed and thrown away - things a PDA-like unit are not! Moreover, a major trend in mobile devices (convergence) indicates that users want *less* devices, not more. You're forgetting that the pads, tabs, boards, idea developed in the late 80's. Their thinking was that presumably PDAs (which were still in the future then) would be disposable, or at least so common as to be shared like umbrellas, rather than kept by one person. They were thinking longer term than perhaps you are. After all, cell phones are getting close to disposable in places like New York; who's to say that in another few years PDAs won't be as common as umbrellas, and that we won't care about them being borrowed, taken, or lost, as long as our data remains accessible by us and only us? Storage media are that way now. How many times have you given away floppies, CDs, and now I'm seeing the same with thumb drives? Is it so farfetched to imagine cheap viewers being used in a similar way? For me, the thing that makes pads a bad idea is not the technical limitations you mention, but the idea Gilad mentions above: "I believe people hate to be sold, but they love to buy. Without the desire to use a technology it is destined to fail, and it seems that functionality on its own does not fill that desire." Point is, people like to own things, so technology companies make things that are desirable. What if we started designing computers that were as cheap, generic, identity-less and single-functional yet necessary as umbrellas? "Information is expected to be able to move around between different devices." This makes sense for displays, but issues of security/privacy and personalization mean that people will want their *environment* to move along with their data. This is a wholly different challenge. not really. I can carry my environment settings with me on a thumb drive, as long as the operating system I plug into can play it back. Interestingly, people seem to prefer phone mtg's, perhaps so they don't have to look decent for their boss (and vice versa) at a moment's notice? "Similarly, if the room 'knows' that there is a meeting in progress, then it can take that information itno account to generate an apprpriiate configuration." This seems to ignore that we're really talking about sensors, data, and code - the "meaning" we derive or assign, and the associated actions taken as a result, are contingent on *us* and how we design the applications. It's not ignoring it so much as it is assuming it as a given. That said, context is still the biggest problem in ubiquitous computing research. "The other, perhaps most important, piece of context it made use of was the fact that it was the Reactive Room." Please note that this also means people changed their behavior to fit their understandings of the expectations of the room - whether those be in accordance with meetings or not. Sure. Matthew Chalmers refers to this as "seamfulness", that there is no such thing as a seamless interface, and that therefore the job of the interaction designer is to design what he calls "beautiful seams". I'd describe it as designing so as to change the participants' behaviors expectations minimally rather than maximally. Just because we know they're going to change to adapt to the technology doesn't give us license to do whatever the hell we want. "A related issue is how tangible interaction transforms the sequential nature of interaction at the interface." See multitasking - how many of us have accidentally copied something into the wrong IM window before? I have no criticisms, but a thought kept running through my head as I was reading this: Which of these concepts, if any, have potential? How can we judge if they have potential? When the perfect solution comes along, will we instantly know that it's the perfect solution? Or will we begrudgingly accept it at first and eventually forget how we ever lived without it? I ask because none of these stood out as ideal solutions. (Imagine having to stare straight downward onto a desk to see a projection coming from the ceiling. How is this an improvement upon a wall projection? I think it's a big step backward, because it still confines us to the desk. I think the whole point of renewing the human-computer interface is to free us from chairs and desks.) Concerning the VR vs Ubiquitous Computing argument, I think I happened upon the answer (at least, I realized my own preference) while reading this. I read it with another book, "The Social Life of Information," at my side. I thumbed through the book to find themes similar to those in Dourish, trying to pull quotes from it so that I could cite them here and sound smart. The book's index was insufficient, so I went to Amazon and searched for text that I knew to be in the book. THEN, once I got the page number, I picked up the real book and read it (even though the whole page was already on my monitor). People love the power that computers give them. However, we still prefer physical things over intangible objects. Having the power of a computer injected into a clothbound book would be far better than "virtual reading." Coincidentally, "The Social Life of Information" was written by Marc Weiser and John Seely Brown, who Dourish worked with at PARC. You'd find a lot of similar themes. Weiser died very young, not yet 50, I believe, of cancer. By the time he died, ubicomp had become a full research area, with a few major academic/industrial conferences spun off of it. I attend one of them every year, and assign many of the readings in Networked Objects (which is probably why Hugms and other projects seem to echo some of these themes). Before he died, he said he felt that the computer scientists and engineers working on ubicomp were mostly getting it all wrong. They were so caught up in trying to algorithmically recognize context, for example, they failed to recognize that computing is one element in an ecology of work, play, etc. Dourish, Brown, Nardi, O'Day, and others have picked up this theme, and tried to drive home a more technologically humble approach to ubicomp that stresses deep user research, ethnography, re-design in the wake of actual use of systems, and so forth. The point you're making about the tangible pleasure of reading is part of that, for sure. I think I'd take it even further and say that being able to lift material from that book and use it in a computer system would be even better, because it would let both systems do what they did best, and take best advantage of my ability to make that judgement. Maybe all I need is a tiny scanner that takes the text from one to the other. "The critical move here was to see ubiquitous computing as a technology of context; where traditional interactive systems focus on what the user does, ubiquitous computing technologies allow the system to explore who the user is, when and where they are acting and so on." It's an alluring and challenging statement. Interactive systems shouldn't just react to you and they shouldn't just allow you to play with them. They should explore you. To do this properly they'll need to reach out and touch us. They'll need to sniff around our necks and prod at our bellies. They'll need to take actions and record how we react. This is the first step to knowing, and we should make our creations attempt it even if we're unsure how to execute the knowledge computations that follow. Probably the second step is to train objects to use the products of their initial exploration--that first sniff--to drive actions that improve the information flow--repositioning the sensors to obtain a better whiff. With rich information from an exploratory process, the task of determining contextual action would be significantly eased. Bilateral exploration is key to true interaction. That means we need to make interactive systems curious. How do we do that? After working in print production for a couple years - I cringe at the thought of a paperless society. There are some things that can’t be duplicated onscreen. When you design a brochure or annual report (assuming these will still exist) – onscreen color never matches printed color (and often varies greatly from screen to screen) Even using pantone or toyo matching systems doesn’t alleviate the need for a printout. Also, it’s been my experience than mistakes (in text for example – mistakes that spellcheck won’t cure) are more frequently caught on paper than onscreen. And finally.... we’re all going to go blind at a very young age if we are only able to communicate and retrieve information by constantly staring at screens.... “What these things have in common is a triumph of the virtual over the physical.” Why must one triumph over the other? It seems like they should enhance each other . (I think Tom made this point earlier). I like Matt’s point about a human-computer interface that would free us from the desk. That made me think- what is it that chains us to the desk? Is it really the PC – or is it the task we have to perform at the PC? Take data entry as an example – you can’t really do it on a bicycle or running around the room (or if you can...an itp’er will invent it I suppose). My point is – some tasks will always require us to sit, be still and focus on... something – be it some projected Digital Desk or a typical computer screen. People who hate this type of work should consider other fields... like lion taming or sky diving instruction. I'm with with you on the fear of the paperless society. I like books. It's been interesting watching the change at ITP as laptops have become prevalent, particularly for physical computing. Used to be that there were a bunch of desktop PCs for phys comp, stuck in the shop. Then we went to rolling carts with the desktop machines, which was unweildy, but mobile. Then we went to laptops, which made for more table space. I'd love them to be floating, but that doesn't always work, as Chiat/Day learned in the 90's. People like having their own environment, as Josh pointed out. Increasingly, I'm seeing people work work on all of their own machines, near where they're working on the physical construction of their work, which is encouraging to me. That's the kind of balance I think is near ideal. Sometimes I think "And then, TABLETS!", but in all honesty, I don't know for sure whether I'd use a stylus for my main text input, because I dislike the feel of a stylus on a screen. Not enough friction for me. When a stylus feels like a fountain pen, I'm sold. |