Science Interviews

Interview

Sun, 16th Jan 2005

Humanising Machines

Professor William Clocksin, Oxford Brookes University Chris - we've been talking about making man more like a machine, but you work on trying to make machines more humanised.

Part of the show Cybernetics and Computer Vision

William - That's right. I work on vision, which is our preferred sense. We get a lot of information from it, so we want to find out how it works and how we can enable computer systems to see like we do. We use vision to make judgements that are intuitive and subjective: judgements that we don't really know how to explain. This means that vision isn't just something simple we do. It is tied in with our emotions, desires and everything else that makes us human.

Helen - Is it right that one of the things that is most difficult to do is to get machines to recognise individual faces?

William - Yes. Faces are interesting. Humans are very good at recognising the faces of people we are well acquainted with, but we're not very good at recognising the faces of people we've never met but only seen in photos.

Chris - Isn't there a special area of the brain that's just to do with faces? Some stroke victims can lose the part of their brain that enables them to remember faces; even of people they have known for years.

William - Yes, this does show that faces are important to us.

Chris - Is one of the main goals for people working on computer vision to make computers that fall for optical illusions like humans do?

William - This is already happening and we are arriving at a much better solution for how these illusions work. In my group, we are more interested in practical applications such as the detection of genetic disease by looking at chromosomes. At the moment, this is done manually and is very labour intensive. In the future we hope to pass this task over to computers who will be able to spot defects.

Chris - When a woman is pregnant, the only tests she can have done to see if her foetus has any birth defects are invasive. These may cause the foetus to spontaneously abort. How does your work improve on this?

William - It has been known for some time that some of the cells in the mother's blood stream come from the unborn baby. We need a few drops of the blood for us to be able to find enough of foetal cells to look at. We then subject the cells to a number of tests we have developed to see if there are any genetic defects. These cells are viable and are a good representation of the cells in the unborn child. Our method is currently in the very early stages of development and there are other people around the world who are also working on it. We're looking forward to some very exciting new tests that will be able to give us accurate information.

Chris - How does the computer spot the foetal cell? Surely cells look very alike.

William - They do, but imagine this situation: if you spot a cell in the mother's blood stream that has a male chromosome in it, it can't possibly be hers. We use tests like this to spot that they don't belong to the mother.

Chris - There has been lots in the news recently about microwaves from mobile phones harming the brain. You think that this is unlikely to happen. Why do you think that?

William - I'm no expert on this, but if you look at the wavelength of microwave radiation and compare it with the wavelengths that brain tissue is able to pick up, they are in completely different regions. Overall, I'm not sure. There are many complex processes involved in disease and I wouldn't like to hazard a guess at whether mobile phones are a cause.

Chris - I do worry that mobile phones will affect me. Kevin has already said that he saw his brain patterns changing in response to a text message coming in. How do we know what the long term effects will be? How do we know that our brain cells aren't having extra genes switched on in response to long-term exposure? If extra genes are switched on, it might also switch on growth factors that may cause cells in our brains to grow uncontrollably.

Kevin - I think that anyone who says that there is no effect at all is wrong. I don't think that's the case at all as it obviously has some effect. The problem is knowing what will happen long-term. I think it will be similar to what happened with smoking, which was once advertised as a good thing. Only time will tell whether mobile phones affect us or not.

Multimedia

Subscribe Free

Related Content

Not working please enable javascript
Wellcome Trust
EPSRC
Powered by UKfast
STFC
Genetics Society
ipDTL