Anthony LaMarca
The multiple research paths that Anthony LaMarca has traversed during his career all intersect at Place Lab. Prior to joining Intel Research, LaMarca developed Internet-scale distributed systems for Yahoo! and built bridges between humans and machines at the Xerox Palo Alto Research Center (PARC). A Ph.D. graduate from the University of Washington, LaMarca returned to Seattle in 2001 to help Intel realize the vision of Proactive Computing. He's now managing the Place Lab project and serving as one of the chief hackers.

Q1: Is there an overall theme to your research?
A1: There are two. The gear head in me is interested in parallel and distributed systems--a large number of different components all trying in tandem to solve some particular task. They're extremely complicated, and making complicated things work is fun. For me, it has the same appeal as a hot-rodder dialing in a four-barrel carburetor.

Q2: What is your other research thrust?
A2: I've also always tried to understand how research in complex systems can be influenced by human-centered design. Computer systems are generally really lousy. As clichéd as it sounds, it's true that if our cars acted the way our computers do, we'd be disgusted. People are constantly feeling dumb because they can't get their computers to work. I always tell them that it's not their fault, and that it's just bad software. Computers are way less usable then they should be.

Q3: Why do you suppose this is true?
A3: One important reason is that the traditional approach in Computer Science is to try to hide a complex technology and expose just a simple model to the user. This works great in some cases, but it can create serious problems because when something goes wrong, the user has no idea why. If users are able to build a mental model that somewhat closely approximates the actual behavior of a system, they can negotiate through problems themselves. The Web is an example of that. Ubiquitous computing seemed like another domain where the same philosophy could apply.

Q4: In what way?
A4: Think of the smart homes that proactive computing will someday enable. There will be hundreds or thousands of computational elements all acting on my behalf. If the system doesn't work, or even if it does, I'd like it to be easily understandable on some level to the users. I'm interested in how these systems will manage themselves and also how users will interact with them.

Q5: How does this tie into Place Lab?
A5: Context is really important to ubiquitous computing. And one of the key pieces of context is location. Unfortunately, very few real location-enhanced applications have been deployed on a large scale. It's one thing to run an application in a little research laboratory; it's something entirely different to take it to Madison Square Garden or a busy shopping center. Place Lab is a way for us to test the location component of ubiquitous computing in real world situations using today's hardware.

Q6: What will be the killer app of location-enhanced computing?
A6: Our sneaking suspicion is that there is no killer app. Rather, I suspect the key to location-enhanced computing is making it easy to develop and use, thereby producing a large number of low-to-medium value applications. The idea is that a daily interaction with tens or hundreds of these small applications can potentially result in a qualitative shift in the way we live.

Q7: What location-enhanced application would you find personally most valuable?
A7: This is dorky, but I'd like more data about my life. I've been a heavy Quicken user for the last 15 years. I'm a closet statistician. I'd like to have an Intel Stargate sensor node in my pocket all the time so that later I could pull up data about how far I've walked, where I've been, and how many times I'd visited a place. For me, location-enhanced computing allows me to capture my interactions with the physical world and bridge my virtual and real interactions.