Return to Ralph Lombreglia's introduction, The Genie in the Machine.
More on Digital Culture in Atlantic Unbound
Examples of Krause's interface designs:
See the online gallery of Krause's imagery maintained by MetaTools.
April 16, 1997
Artist at Play
An e-mail exchange with Kai Krause
Ralph Lombreglia: Your well-known computer graphics tools don't really look or feel like "tools." You and MetaTools are famous for playful, almost physical interfaces that sometimes feel more like an arcade game than a "productivity application." Sometimes they look like visual art themselves. The basic philosophy seems to be that if a tool is fun, game-like, and as tactile as possible, it will provoke artists to create things they would never have created otherwise.
Kai Krause: It's a deliberate method, of course. First off, the interface is not an afterthought -- it is half the process of creating our software, the other parts being deep mathematics, the feature set of the product, the proprietary internal structure and code, and so on. Nor is the interface itself a mere collection of buttons and "dialogs." I see it more as if we are creating a movie or writing a book or carving a statue. The user is coming into our "world" for a time, and that world is geared toward exploration. The software is not merely a tool that executes what you already know; it is capable of much more than you could possibly already know. We want the interface to create a sense of wonder and lighthearted playfulness: everything is just one trial mouse-click away, the user sees results quickly, and everything can be undone. "You can't hurt it and it can't hurt you." Even people who might not describe themselves as graphic artists will want to spend a little time in our "world," and we want them to get at least positive feedback, if not the proverbial "shiver down your spine" payoff of a full creative process. The ultimate success would be to compress that millisecond of adrenaline rush into a product and let people simply experience creativity. To be sure, this has been heavily critiqued by the semi-pros, who find it too playful and poo-poo something like Kai's Power Goo as a mere toy. Then again, we have over 2 million copies of Goo out there worldwide . . . and I make no excuses for its being a diversion rather than brain surgery or tax accounting. Still, we are now following that model and adding real tools to the toys.
I am convinced that the real beauty of computers can lie in the infinite combinatorics [combinatorial mathematics] and chaos: you don't have to approach the screen with an image already formed in your head as you might have to with a canvas. You should be open merely to steering the process and seeing how many branches into unexpected and unknown territory can happen. When we created the Texture Explorer, for instance, it became commonplace that someone would go in there searching for a piece of granite and end up three hours later with glowing magma barnacle larks vomit instead. The production people are truly pissed off at that, yet all I can say is, tough cheese. . . . There is plenty of software out there for rendering a 3-D slab of granite. Or you could get this totally newfangled thing, Actual Reality. I hear it's great: take a picture of a real piece of granite. . . . woof! But when you want to make a button pop off your Web site you may end up back in Kai's Power Tools looking for that unpredictable effect. Power Tools were just plug-ins to other software [for Adobe Photoshop and compatible image editors]. Now the new frontier is KPT Unplugged, which will be a self-contained set of real tools, in 2-D, with depth in 3-D and movement in 4-D. And yes, we have no idea yet what all people will do with these things. Quite lovely, that.
RL: In some cases (such as Bryce 2), you have used a pre-release version of the tool to design the very interface for the final tool itself. This behind-the-scenes production fact seems to capture something essential about creativity in a digital age. The possibilities of infinite regression make one giddy.
KK: Recursive fun, yes. Nice that you caught that. We did indeed often render the entirety of a new interface with our own products, and sometimes in the one we were actually designing at the time. This is not just an intellectually tickling concept, it is a basic part of pushing the envelope, much like machines making machines: what we're doing simply has not been done anywhere else, since we're inventing how to do it as we do it. Some of our powerful forthcoming tools have interface elements executed with our old Kai's Power Tools and even with the low-cost consumer-oriented Kai's Power Goo. Knowing that, it's hard to believe that anyone could still think those earlier tools were mere toys or professionally "useless."
I do love the notion that the line of what is possible will still move out for many, many years. I always say, "we have seen nothing yet." In fact, the way I like to say it on stage is "we have seen exactly 0.3981 percent of all the possibilities" -- and then to watch people write down that number. : )
RL: As an artist, do you still use traditional, physical media? If not, are there things you miss about it?
KK: I used to love airbrushing and, even more, oil marbling. I came up with hundreds of techniques for mixing inks and glue and dyes and crystal solutions. I painted with crushed mirrors and I stuck butter in glass plates to insert into slide projectors (with time the heat would move bubbles around in an incredibly organic, chaotic way). Little did I know that it was all just fractals done with a manual method, creating what would otherwise have taken one hell of a parallel supercomputer.
I do miss that total real-time and real-depth dimension. Electron Microscopy is another fascination of mine, which I am now finally going to get back to. I also would like to deal with blown glass at some point. The simple fact is that there ought not to be a real distinction between electronic versus analog media. A true artist will never blame his tools.
RL: With all these marvelous tools on every computer desktop, is there a danger that much digital art will begin to look somewhat the same? When the magic comes from algorithms, does it carry a kind of confinement with it?
KK: Perfectly reasonable argument, and I totally agree. There is a dilemma here for me. On one hand, I am responsible for a lot of mediocre creations that do not deserve the moniker "art" in the first place. When I get sent endless floppies and zips and CDs with glass balls on fractals and page curls over noise textures [examples of the easiest and most obvious imagery to create with Kai's Power Tools], I am on one hand disappointed with the lack of determination to search deeper than the presets we provide. On the other hand, I also have to see that everyone, no matter how cool and professional they now are, started with a bunch of silly yellow triangles and ugly-as-dung pixelly lines.
In other words, we all have to go through phases of silliness and sameness to find our unique space. My hope is that at least some of our amateur users find enough fun to stick with it and discover their graphic heart. The analogy I always use is that of Sony's inventing the camcorder: all the movies of weddings and babies are ludicrous if watched under the guideline of cinematography, but the camcorder is not a shortcut to Citizen Kane, it is an immensely beautiful and important tool to preserve memories. In that sense I do not like my tools to be approached as one-click-art short cuts to Mona Lisa, but as beautiful aids for playing with your own brain. As such, I'll compare them to computer games any day.
The most unimaginative misuses are visible from ten yards away, while the subtle little glow and soft texture may be almost completely invisible. With Bryce [the 3-D tool orginally designed for generating synthetic landscapes] we've had a chance to observe users for more than three years, and it has not led to "more of the same." In fact, we were amazed: at first nearly everyone used the basic textures and primitives to make all the standard Bryce clichés -- chrome balls on mountains, pyramids in water, and so on. Yet soon users began sending us the most amazing contraptions: photo-realistic guitars and trucks built out of many little shapes that were technically "mountains" but that produced audio knobs or rubber tires. Surely the tools have a certain look and that can lead to sameness, but that's actually more of a social problem than one of technology: conformity in style and flat emulations of the existing stuff are the common practice in 95 percent of all books or movies or records. Sad, yes. But I am always happy to see the best escape over the top.
RL: What do you see as the next likely steps in the evolution of graphical computing?
KK: We cannot even grasp what will happen here. It used to be that you could look ahead in military machines to see the curve. When I had a 128 x 128 Compucolor with 8k of RAM in 1979 I saw a 1024 x 768 ultra-killer dream of a machine at a military convention. Now I am sitting here with Sony's 1920 x 1280 HDTV tube and I cannot easily look over that horizon. It always seems to go so slow in one sense and yet so fast in another. I would love to think that within twenty years we will have paper-thin displays with a billion elements looking every bit as real as paper, ink, or wood. The CPU may be spread over a million parallel centers, acting in true real time on the images. Who knows. . . . Then again, for me the real machine of the future may simply be one that boots in one second and is dead quiet. One gets to be grateful for the little things. ; )
RL: What's wrong with the interfaces of today's "productivity applications"? Are such things as pop-up Help windows and automated file-formatting "Wizards" a good idea, or just a band-aid on bad interface design?
KK: I love them as they are -- they are so nicely ridiculous that we look quite happy by comparison. Let them continue with Wizard dialogs stacked upon dialogs. . . . I will just smile. Surely such things are meant well, but they're not addressing the underlying fallacies. Can it be done differently without dumbing down or becoming condescending? I believe so, and I will try to earn my laurels the hard way in the coming years. This one will have to be debated on the issues and won with results. None of my previous designs should be dissected as representing my style per se. They are all mere history by now. Every one of them was diluted by a long list of pragmatic realities to get things out onto a store shelf by a certain date. I have my best work yet ahead of me. As I go forward this year I am more and more independent of the watering-down effects of deadline pressures, and I can retain more control. As I come out with basics like word processing, I guess we can revisit this theme and see if I was able to do anything any different or not.
RL: A debate rages in the literary world about the impact of computers on writing and reading. Some thoughtful people feel that digital tools and experiences are incompatible with the consciousness required for genuine literature. Others feel that computer tools and networks could give rise to a new golden age of literary production and publishing. Do you have thoughts on the fate of literature in a digital age? Does a similar controversy exist in the fine arts, and if so what are your feelings about it?
KK: I would guess that never in the history of this little planet has there been this much writing going on. Even the analog paper output is at an all-time high. The percentage of time spent by ordinary people on communications is also higher than ever. Surely quantity is not breeding quality -- and there may be no correlation at all -- but the idea that people are not reading any more or that paper is fading away is misguided, I believe. As I get quoted a lot: 97.3 percent of all statistics are useless! The Golden Age concept is also usually prone to tricky statistics. There usually is no such thing as Golden Anything upon closer inspection. The person living next door to Shakespeare is likely to have been not only completely illiterate but also given to bathing only on even leap days. For every Goethe there were countless lumps of more common human clay. Conversely, looking at how many young people now have totally new kinds of jobs creating Web pages and content and 'zines and who knows what all, I venture to say that there has never been a higher percentage or number of people involved in generally creative work. Of course, I am shooting from the hip with these numbers -- they are not exactly in Popular Stats Daily magazine.
Whether all this breeds better chances for Mozarts and Bachs to emanate is not clear; statistics are elusive here as well. For the few major forces and famous names there were immense percentiles of general "Duh!" out there. I still think that the tools and technologies themselves are not to blame here, nor do I think we should look to them for salvation. We should rather have basic IQ pills to dole out if we want to lessen the percentage of romance novels versus "literature." Appreciation of detail and subtlety, learning how to discern qualities and judge a work in the context of its time and creator, lengthening the attention span -- all that is equally necessary to support an "age of literature." We live in an age of Burger Kings at the ends of the earth and infomercials clogging the airwaves. It's hard to understand why some people would feel that those same old analog signals turned into digital numbers would represent some evil empire, as if the digital numbers were all "6"s . . . hmmm.
RL: Can you envision creativity software that might do for text artists what your products have done for graphic artists?
KK: Give a guy a chance . . . it's only spring yet!
RL: Are you working on audio/video-enhanced interfaces? If computers could really talk to us, would that make them more approachable, friendly, and useful?
KK: I am not actually a proponent of speech recognition as the next great interface idea. Maybe it's because when using a computer I like to watch some old English movie on the side, or listen to music, but I expect that I would feel quite silly and annoyed to have to speak into the machine, no matter how fast or accurate it was. Or it may be that having played piano, I trust my fingers to be an incredibly versatile, agile, and efficient control system.
Clearly speech input for dictation of plain raw text is a neat thing. Why not? But I have a hard time seeing the point of employing speech for such subtle processes as fixing the detail in an image by cloning small portions of it. Making all the elements of an interface move -- yes, that will be a natural extension. The performance cost of such animation for now is too high, and I don't want to trivialize it with a few rotating icons. Rather than waiting for computers to talk to us, I wish the damn things would just simply work right and not crash. We have had ten years of eight-character filenames and are even now only an inch away from that era. It seems unrealistic to jump over so much basic interface territory when it is still in need of improvement. Clearly I can only attack a small portion of these things myself with my merry band of MetaTools specialists. But we have set our sights pretty high on making a dent out there. As we always say: We are the dentists. . . .
Copyright © 1997 by The Atlantic Monthly Company. All rights reserved.