 So I'd like to ask how we can apply what we're learning from decentralized systems without central control in which simple local interactions among the parts combined create the behavior of the whole and how that functions differently in different environments. So I study ants and ants smell with their antennae. And when one ant touches another with its antennae, it can tell if the other one belongs to the same colony and whether it's been doing the same task. And we've learned that ants use just the rate at which they interact. There's no message. There are no instructions. And we know this because when ants contact little glass beads coated with an odor, they react the same way. So we try to understand the networks that are created through interactions among ants. You can see ants touching each other with their antennae and how these networks change in different conditions and are regulated differently in different environments. So we can look at a network and, first of all, see which ant is contacting which ant. In these diagrams, the red circles are the hubs, the ants that made a lot of contacts. And we've learned that ants can be hubs in different environments. That is, it's the situation, not the ant, that determine who gets to be the hub. So we look at this in extreme conditions. For example, an experiment we did last week at the International Space Station. That's the astronaut Rick Mastracchio looking at ants solving a searching problem in the very extreme environment of no gravity. But normally, of course, I study ants on Earth and ask how different environments have led evolution to shape collective behavior differently. For example, in the desert where the important constraint is the scarcity of water. And in the tropics, we're operating costs are low. But there's so much diversity that competition is really important for regulating networks. So in the desert, a colony has to spend water to get water. An ant loses water just being outside foraging. And they get their water by metabolizing the fats out of the seeds that they eat. So basically, they have to spend water to get water. Operating costs are really high. And in this kind of situation, it seems that networks are regulated using positive feedback. A forager doesn't go out unless it gets positive feedback from interactions with returning foragers with food. The system is inactive unless something good happens. Now, we can ask how systems like this are evolving. We see similar systems also in human engineered systems. The internet works the same way that a data packet doesn't go out unless it gets positive feedback in the form of an acknowledgment from the router saying that there was enough bandwidth for the last data packet to go out. We call that antternet. But we can ask in natural situations how this is evolving by asking how the regulation of colony behavior leads to reproductive success in numbers of offspring. And that's what's shown there is a map of my long-term study site with parent and offspring relations. So it turns out that in the current drought situation in the Southwestern US, natural selection is favoring colonies with really, really stringent positive feedback so that they favor conserving water rather than going out to forage. They're sacrificing food intake in order to save water. And those are the colonies having the most offspring. Those are the colonies doing the best. In the tropics, it's a completely different situation. It's so diverse there are many different species competing for the same resources. And so in this situation, networks are regulated using negative feedback. They just keep going unless something bad happens. So each colony creates a circuit, an ongoing circuit, of foraging ants round and round in the trees. And those circuits just continue like a fiber optic network unless something bad happens like an encounter with a neighboring colony. So it's negative feedback that stops the system only if something bad happens. So I think that we can use these kinds of what we're learning about networks in different environments to ask how even in human societies where local interactions in the aggregate allow groups to accomplish certain tasks. And I want to end by showing you inside an ant nest the interactions between returning and outgoing foragers to let you think for a minute about how very messy and noisy these interactions can look at the local level and yet be very effective in the aggregate for colonies to adjust to changing environments.