Dr Kathleen Richardson is Departmental Research Associate at the Department of Social Anthropology, Cambridge University. She has written on how humans relate to robots and was a visiting researcher at the Humanoid Robotics Lab, MIT, Boston. Dr. Blay Whitby is a lecturer at Sussex University and the author of several books including Artificial Intelligence: A beginner’s Guide, and he works on the ethics of robotics. A recent talk of his was called Do you want a robot lover?
Kathleen Richardson told us that robots being developed at MIT and in Japan were pretty far from the ideas we have of humanoid workers found in science fiction. Scientists are trying to build companions that might one day help care for the elderly and children. Second, she noted the widespread predictions that robots would be developed into sexual partners and third, she introduced what she called the ontological uncertainty that these kinds of experiments seem to encourage. They promote confusion between the categories of human, animal and robot, by breaking down the boundaries between people and objects and replacing humans in different kinds of relationships.
When she arrived at MIT Kathleen said that there were several machines that looked like they might have stepped from science fiction stories: “They had a robot with a friendly human face that made me feel all my childhood fantasies would come true,” she said.
But after about six months Kathleen realised that they were not making robots that would do useful tasks for people. The robots were designed to interact with people on an emotional level to provide companionship rather than physical assistance.
Illustrating her words with images of the robots projected behind her, Kathleen particularly spotlighted a Japanese robot baby seal called Paro that was used as a sort of comforter for people living in an elderly care institution. The white, furry thing, about the size of a lapdog, looked at people and wiggled slowly and seemed to respond to, and enjoy, being petted. Some in the audience found it a bit creepy. Kathleen said this sort of machine is seen by its developers as almost a new species, designed to play a new role in its interactions with people.
The purpose of these emotionally responsive, companion machines was to in some way replace the relationships that people might otherwise have had with humans. Interpersonal relationships are envisaged as being exchanged for those with machines of a new kind.
People have sex with a variety of objects already and some may feel there is something wrong with that while others may not. But in some ways recent developments in the building of sex dolls have moved them away from being objects that provide simply physical stimulation toward providing an emotional relationship with a machine companion.
Citing an excellent Channel 4 documentary on the relationships people have with “Real Dolls”, Kathleen explained that these very expensive, very lifelike dolls (which are built for people to have sex with) are replacing women in the lives of the men who buy them (although there are male dolls, most are female).
“She’s not a living human being,” said one owner about the advantages of his Real Doll over a girlfriend. When one sees relationships with real people as difficult and problematic, an object that can provide some of the characteristics of companionship without the inconvenience of a real person can seem desirable.
Donna Harroway’s Cyborg Manifesto is a good example of the attempt to break down the boundaries between humans and machines and humans and animals. Feminist theory aside, the creatures that are imagined or built by robot developers also attempt to break down those boundaries. The companion species represent an attempt to both realise the existence of creatures that defy ontological definition and to bring humans into relationships with them.
‘They make some wonderful killing machines…’
Blay Whitby responded first to Kathleen’s presentation, emphasising that the issues raised were very important and prompted ethical questions of all kinds. The robot has not developed in a direction that resembles a human in a metal suit. People have sex with all sorts of things and one writer has even predicted that progressive states will recognise human-robot marriage by the year 2050, said Blay.
The breakdown of the categories of human and robot is part of the history of the philosophy of love and many people see that our definition of love will change. Paro the robot seal shows the very different reactions that people from different cultures have to robots. In Japan the attitude to robots is very positive and Japanese people look forward to robots acting as childminders and nurses.
In the US, “where they have made some wonderful killing machines,” said Blay, there is still great resistance to the idea of robots automatically killing people. One of the ideas that Blay considered as having great ethical significance was the ambition to program the Geneva Convention into the software of robot “soldiers”, armed drones that might have to make life-or-death decisions.
In Europe we are generally more cautious and beneficial applications that are envisaged include the automation of residential homes for the elderly, allowing them to live independently for much longer than would otherwise be possible.
Both these applications raise difficult questions. Even a benign helping house for the elderly opens the way for programmers to decide when people should get up in the morning. The ethics of all these applications are complicated further by the input that must reflect the biases of the people who build them.
The discussion took the form of questions from Dr Robert Clowes, in the chair, and then further questions and points from the audience. Kathleen and Blay responded to as many of them as they could but, as the mere sample below shows, there’s a very great deal in this subject…
What actual advances in robot technology have been made in the last 20 years that have led to the development of robot companions?
The robots exhibited by institutions are not very capable and the video clips one sees of them take hours to set up. They are not really capable of the things they pretend to do. Shouldn’t we consider these things for what they are rather than how they are presented by their builders?
Where’s the big vision of robots helping to take our society forward? What about all the robots used in production?
Aren’t we missing something important by concentrating on the personal relationships roboticists try to build?
A human being can easily create a relationship with an object emotionally and it doesn’t have to be a robot. TE Laurence had a name for his motorbike and even an object as limited as a pebble can be imbued with characteristics it doesn’t possess by a human.
The robots doing real work are rather boring and are basically clever tools. A real, independent robot is perhaps 100 years away.
Before we worry about the ethics of robotics should we not try to build some really good ones?
What is ethical and what is not is a big question itself. We raise concerns about a killing drone but a similar machine [unarmed, presumably] searching for people lost in the snow seems to have no ethical problem.
Even in relation to developing more useful robots, aren’t the emotional elements good in that they help develop machines with some human qualities and raise the robotics game generally?
The internet showed that when people became familiar with the technology they begin to get creative with it and take it in directions never foreseen by the early pioneers. Until people can get to grips with the actual technology of robotics, we can’t see how it might change.
The interesting idea of a house helping the elderly is an example of how useful robots could be. Are developers stepping back from the possibilities of a brave new world? Is it possible the emotional robot might be a bit of a dead end?
People who build robots have a great deal of backing and they are, in the end, trying to sell us something. It’s very important to look at who is funding what and where their money is coming from.
A drone that goes to work for you seems a great idea but one that automatically kills things replaces a soldier that might have made a different decision and that seems a bad idea.
Can decision making processes be duplicated by machine? That seems to open up the possibility that robots could run amok and discourage their development in useful areas.
The relationship that a person can build with an object can go beyond simply representing the person another way. It may be that someone who displays their personality on Facebook can come to see the interface itself as their personality, not merely a representation of it.
MIT can spend billions on developing things that seem to be headed for the entertainment industry, such as a Disney animatronics display, that are not really robots at all. The way that these machines have been presented to people as robots is simply bad science.
The discussion promoted about these robots suggests something else is going on. Theses machines have many potential uses but they are also vehicles that promote odd discussions about our society that do not reflect what is really going on.
Whoever creates a thing and programs it must take responsibility for what it is and for what it does, including the people who funded its development, but there are still going to be major deviations from the outcomes that they had imagined.
Phew!
Vanilla emotions
From the chair Rob recalled some of the discussions in previous salons that had examined people’s recent fears of intimacy. Discussions on singletons, social networking and friendship had thrown up data that suggests far more people live alone than ever before. A surprising number of people will say that they have no friends at all while others would claim to have thousands on Facebook. The broad retreat from intimacy and human contact may explain much about developments in robotics.
Rob pointed out that the emotional companion robot seems to be a means of editing out the inconvenient aspects of real relationships with people. This creates “vanilla emotions”!
A washing machine with a face
Kathleen said that there had not been any great advances toward real robots but that the impetus toward robot companionship had come from the changing attitudes of humans. Drawing an analogy with animal rights, where we have regulated our relationships with animals, Kathleen said it was because we have come to view animals differently than we did in the past. “The animals have not become smarter, nor did they campaign for their rights themselves,” she said.
The robots that she had seen developed had no more advanced technology than one might find in a washing machine. They have stuck a face on a washing machine and they might soon offer such machines as sexual partners.
As an anthropologist, Kathleen has seen many different relationships with objects in different cultures. In one society shells and axes would be gathered together for display, for example, but the objects themselves are representative of the relationships and position their owner has with other people in that culture. The gift and act of gift-giving, for example, displays a very different attitude to an object than if it has simply been bought. A robot lover is an object that shows an emphasis on the personal sphere that represents a changing relationship between people and objects.
Narcissus fell in love with his own reflection in the story that ends either with him wasting away, unable to tear his gaze from his image while Echo calls him hopelessly, or he commits suicide. This idea of creating a copy of yourself and becoming involved with it can reduce actual relationships to nothing.
Kathleen said the term robot was first used in a 1920s play to refer to people who were built in factories to work. The play dealt with the issues of what it is to be human and how humanity can be defined. By the 1930s the term stuck itself to the idea of a mechanical person and spread into the rest of our culture. The robot has been closely associated with annihilation since this time, in both physical and psychological ways.
Kathleen said we all want to be able to just tell machines what to do. As a robotics expert, she cannot get her video recorder to do what she wants! Robots might help people with commitment problems but this too could also isolate people from the experience of real things. For example, a dating manual will always tell you not to show your real self at first and to improve your image.
Blay agreed that human attitudes drive these developments. In one way the first time we meet artificial intelligences in our culture it is usually in a game where we are trying to kill them. The technologies used in the robots we have been looking at are dressed up to encourage human responses. “We know if you put big eyes and a pair of floppy ears on any machine people will respond to it – even if they know they being fooled by appearances, they cannot help responding!” said Blay.
An automatic killing machine raises many difficulties. For example, it does not take the decision to go to war in the first place but its actions distance those who deployed it from what it does. They can blame the machine if, say, it knocks down a Palestinian school in Gaza. Whatever the uses of the machine and however independent it may be, its programming always reflects the prejudices of those who designed its software. Booking a plane seat by credit card can upset a system that is programmed to behave as if everyone with the title doctor must be male.
Blay couldn’t say to what uses future robots would be put. But the tendency in our society for people to be much more ready talk very personally about their problems and reveal more about themselves in public is strong influence on the thinking behind the robot companions. Blay could see that sex with robots might help some people even if it was unlikely we would all have robot partners one day.
Another problem Kathleen had with machine soldiers was that by removing humans from the experience of war it means that, however bad the experience may be, there would be chance for those fighting to deepen their understanding of war and of their understanding of human emotions and, perhaps, change their attitudes to it.
Blay said that he had found a copy of a book he had written 20 years earlier on the use of artificial intelligence in wars that had been in sold second-hand after resting on the shelves of the library in the US military’s Missile Command. He knew therefore that military strategists were very interested in how machines would be deployed in war.
Asymmetric war describes the usual kind of conflict one sees these days where one side is vastly superior to the other technologically speaking. The technological superiority gives rise to the use of suicidal tactics because conventional military or guerrilla methods would be instantly overwhelmed.
Military drones are not actually that good yet. A battlefield where massive robot forces faced each other would indeed be a very uncomfortable place for a human to be.
Yes, there are decision-making programmes that are very good but, unlike the anthropomorphised robots, they aren’t very glamorous. Blay warned us not to be blinded by the packaging that comes with trendy research areas such as artificially intelligence and androids.
Kathleen said scientists are always trying to do things differently and consciously try to change things. They believe that technology is not stopping them and that it is the tool by which they can come to control anything. If the androids of science fiction are unlikely, the efforts to improve human machine interaction can still be very beneficial in many ways such as better user satisfaction.
The Brighton Salon would like to thank Dr Kathleen Richardson and Dr Blay Whitby for a fascinating discussion. We would also like to thank those who joined in from the Informatics Department at Sussex University and Simon Belt, who runs The Manchester Salon. Finally we would like to thank Bellebys College for hosting the event and for Mr Peter Travis’s help in particular.
This is personal report of the proceedings at the Brighton Salon, January 2009 the salon’s secretary, Sean Bell. I’m all too human and any mistakes, unforgivable omissions or misattributions may be addressed to jo.e.bell@btinternet.comThis e-mail address is being protected from spambots. You need JavaScript enabled to view it .
