Video sponsored by

Filmed at FITC Toronto 2010
The Design & Technology Festival
More info at

Presentation: It has to be this
Presenter: John Underkoffler

Over the past three decades, we've ruinously forgotten the singular importance of interface: we invented one in 1984 (it was called the Macintosh at first), and then we stopped. Since then there has been almost unimaginable progress in everything else -- graphics, networking, processor speeds, memory capacities -- but fundamentally none in the interface. In relative terms, then, the interface has gone backwards. This wouldn't seem to be the case, from a casual glance, because everything looks great: the visual sophistication of what sits on the screen these days is astounding. But the utter paucity of our facilities for interacting with that stuff is an embarrassment: we squat in the middle of the world's most richly stocked art gallery and grunt incoherently at the walls. In response the art does very little.

That's not interaction.

The way forward -- and we have to get there, because we're today ramping up to a catastrophic digital implosion in which we, the human half of the human-machine relation, slip soundlessly into a kind of mute illiteracy, robbed of all agency -- is to teach the machine about space. A computing environment built on an understanding of real-world geometry, in which pixels are understood to be present not on isolated Platonic graph paper but rather in the room, among other architecture, objects, and people, is an environment that can offer interactions of a new richness and depth: implicitly like our interactions with non-digital parts of the real world. We call such a system a Spatial Operating Environment. Crucially, an SOE deploys input modalities (what we humans do) that are as nuanced, expressive, and sophisticated as its output modalities (what the machine shows us). That means hands. Human hands -- not clutching a crude mouse but free to point, shape, gather, sweep, and otherwise express -- are the true agents of our spatial intent. In response, the SOE must abandon traditional on-screen elements (the windows and pull-down menus and buttons designed and optimized, remember, to be driven by a mouse) and instead provide something irresistibly new.

We'll look at all this in actual practice: Oblong's g-speak SOE, long in quiet development, is now in use by early adopter clients and in universities; but its arc will bring it to ubiquity within a few years.

John Underkoffler is founder and Chief Scientist of Oblong Industries, developer of the g-speak Spatial Operating Environment. Earlier, during fifteen years at MIT's Media Laboratory, he was responsible for innovations in optical and electronic holography, novel animation systems, several large-scale visualization techniques, and the I/O Bulb and Luminous Room systems. His applications and installations are in use at commercial and educational facilities, have been shown in galleries and museums and at festivals, and have received a highly small number of awards; the human-machine interfaces he's devised over the past two decades remain widely influential. Interstitially, John has been science & technology advisor to films including Minority Report, The Hulk (A.Lee), Aeon Flux, Stranger Than Fiction, and Iron Man. He currently also serves as adjunct professor in the USC School of Cinematic Arts and on the board of the University Art Museum at California State University, Long Beach.

Loading more stuff…

Hmm…it looks like things are taking a while to load. Try again?

Loading videos…