 Okay, here's Baby X waking up. Now Baby X can see me. It can see my expression. So we can we can speak to her. Oh, yeah. Who's a cute little baby? You are. Yes, you are. She'll look at whatever's going on so I can wave my hand around. She'll follow that around. I can play peek-a-boo with it. Here you go. Boom! We can even abandon Baby X. We'll start getting stressed. The different muro modulators changing here. We're getting cortisol released and it's starting to cry and get upset and so you can come in and comfort it. It's okay. It's okay. It's okay. I came back to the University of Auckland to do a project which was very experimental. We're trying to make computer system feel unrobotic by giving it sort of emotion, giving it expression, giving it a real emotional rapport with the people interacting with it. There are lots of models that we're using, things like how the ocular motor system works, for example, how the brain drives the eyes to look at things. If I zoom in closer, you'll actually see the ocular motor system. So here are the different nuclei which are reacting. So when visual stimuli comes through, it goes through the optic pathway and it goes into the brain system which fires off a particular nuclear which moves the eyes around to look at what it's paying attention to. There's a whole system which is running live using physiological models to basically create that movement and using biologically plausible information processing systems to create the nerve activity to create that. These are all based on current theoretical models. Now to put all these together, we actually had to develop a special language. It's kind of like a building block system for creating an interconnected brain model. We're building a system which can gauge your reactions and then build in the appropriate response. Now that has an enormous potential over a whole range of fields. We can look at things like any type of simulation technology, like for example say it could be a civil defence emergency. So how would you train people to react with a very stressed out person, for example, with things like psychology research because we can build human stimuli which people react with that we can precisely control which you can't do with video or other means. This work can create a new type of human computer interaction technology which can be used essentially to build a new media. And if we generate that technology here and we have the leading edge in that technology then I think we can revolutionise how people interact with electronic media.