Interaction

My primary interest in interaction has been in sketch-based modeling, but I've long felt that interaction is the poor stepchild of graphics, and will, I hope, someday asie like Cinderella to take its righful place.

Think about this: in the early 1960s, computers cost millions of dollars; a single computer might have 100 people using it (each getting his or her little time slice during the day, since only one person could use it at a time). A good engineer might have earned $15,000 dollars a year. So the raio "computer cost/human cost" was, say, $10million/$1.5million ~ 6.7, where I'm counting up the salaries of all the people using the machine. Now look at the current decade. An engineer can be paid $50,000 per year; the computer that s/he uses might well cost no more than $1000. So the ratio is now 0.02.

In the early 60s, people worked long and hard to get their programs ready for the computer, and in some cases "programmed" it by hooking up wires. The goal was to save time for the computer, at the cost of using more human time. That made sense economically. But now things are completely different, and yet we still have amazingly primitive interfaces, and spend time memorizing clever codes ("cd", "ls", "grep", ...) to make things simple for the computer.

I want to move beyond this nonsense. To make that happen, we need to open up whole new ways of interacting. cwTomer Moscovich's multi-touch interaction work gives an example of the sorts of things that are possible. So does Olga Karpenko's work on 3D sketching; the interaction is relatively simple (drawing curves) -- the shock comes when you see what's done with it: at the cost of a lot of computing, we manage to make 3D shape from those curves.

There's an opportunity in interaction research to step beyond the limitations of technology. Most interfaces to computers make us feel clumsy at first. Every time I walk into a virtual environment lab, I'm reminded of how I feel after dental work, when there's still novocaine numbing me. I have to concentrate just to avoid drooling; analogously, in most VR setups, I have to concentrate to avoid other problems, almost all of them interface problems, usually created by the devices: trackers have jitter and drift, for instance, so one has to be aware just to hold one's virtual hand in one place! In trade for this newfound clumsiness, we get so little back! We ought, upon entering such an environment, to be granted superpowers, albeit with slightly shaky hands. The right superpowers could help us compensate for that shakiness.

In the meantime, I'll continue trying to catalogue my frustrations with present-day interfaces, and try to see how to advance them for the future.