Perspective: computer science and the brain

I’m Russell Beale. I’m a professor of human-computer interaction in the School of Computer Science.

One of the things that we’re interested in from a computer science point-of-view in terms of brains is that computer scientists can actually take some of the work that’s being done in brain science and create computer models of these sorts of things. So we can understand which aspects of behaviour are explained by these models and which aspects are still left unexplained. So we can effectively run controlled experiments on models of brain function and cognition and other things like that and, therefore, feedback results to other scientists.

What’s also useful is that, because the brain has a lot of power and flexibility, the things that the brain can do tends to be the sorts of things that computers, if you programme them normally, are very bad at doing. So computers are great at remembering fine details for large amounts of time. But they’re really terrible at recognising pictures and images. Whereas with humans, we’re terrible at remembering fine detail for large amounts of time. But we can recognise objects, and mugs, and cups of tea, and dogs, and other sorts of things with no problem at all.

So what we tend to do is we tend to take these models of the brain, produce computational models, and then give our computers these sorts of capabilities. And what my work involves is taking those sorts of artificially intelligent systems and applying them to the interactions that people have with computers to try and make those interactions more natural and more understandable and so forth. So we’ve got people doing projects on waving your arms about to control systems, and the computer being intelligent enough to recognise those sorts of gestures and understand things.

And this is particularly useful for people who are in wheelchairs and got certain disabilities that they’re actually able to interact with things and control systems in a much more natural and convenient way.

One of the other interesting areas of brain science is to try and get a better understanding on emotion and personality and sometimes how this is impacted by injury and so forth. From an interaction point-of-view, we’re interested in making our interactions with computers much more enjoyable, much more natural, and so forth. And to do that, one of the areas that we’re investigating is to give computers a more emotional response to people so that it feels, rather than you’re talking to a dead machine, that actually you’re talking to something that is interested in what you’re talking about and trying to communicate.

And so a lot of our work builds on the brain science understandings of emotion to try and build these capabilities into systems. One of the difficulties is that emotions are very subtle. They’re very fleeting. And it’s actually quite hard to build these systems, which is why they’re not in commercial use. But in the labs, we’re having some very interesting results. So we found, for instance, if you talk to systems that are emotionally appropriate back to you that you stay a bit more engaged with them. And so if you’re trying to change people’s behaviour, for instance, to make them eat more healthily, that longer term engagement builds into a better habit for them.

And we have better success with that sort of thing. And you tend to trust these sorts of systems a little better. And, again, we’ve run lots of experiments that those sorts of things. And this relationship between the two sciences is one of the reasons that I like doing computer science, because we do a lot of work with a lot of different other disciplines and try and pull these sorts of things together.