Our guest speaker for "Ecomimicry", Nina Fefferman, told us about her research into the World of Warcraft "Corrupted Blood" 'virus'. Specifically, as an epidemiologist, the studied the insights that a digital outbreak of a contagious disease could shed onto how social systems (like thousands of gamers) respond to an epidemic. She backed observations about how the real world worked out of a crude approximation of the real world with thousands of players controlled by real people.
In our discussions of Ecomimicry, we talked a lot about how these large social systems online can emulate real social systems. We also talked about how one could intentionally design a system to take advantage of these parallels and behavioral mechanisms that emerge. In the case of online communities, people coming together creates an emulation of the real world. But the opposite is possible - we can create a system that emulates the way people work, without the people.
I did some quick Google searches for demos of particularly neat swam robot behaviors. I came across this site, which seems to be a slightly abandoned page by a research group investigating using the idea of "pheromones" to let the swarm robots communicate and work together toward goals. Insects use pheromones to signal to each other, so this team has taken the idea from the biological space and used it to establish a set of rules about how these robots should all behave. It is this emergent behavior from simple rules that is going to keep pushing what large groups of autonomous agents are going to be able to accomplish.
So here's my idea for a mash-up: Groups of people from all over the world are cool, albeit somewhat limited in their ability to physically affect the same place. Swarms of robots are cool, albeit somewhat dumb. What if we created a system where individual people were able to take control of a swarm robot to accomplish some task? It could be something as useless as "spell out a word" or as complex as "sort these boxes on the factory floor." If you could drop 10,000 robots on a disaster scene and volunteers could drive them around and flag distressed people or dangerous sites, how useful would that be? Nothing can improvise solutions to complex problems as well as the human brain, but it's hard (and expensive!) to get a lot of human brains together in the same place. Amazon's Mechanical Turk does the same things for purely-digital mundane tasks (Is this a picture of 79 State Street?), but imagine harnessing the same mechanism to do things in the real world.
Thousands of people, thousands of robots, useful telepresence work.