In sci-fi films, when anyone gives a computer emotions, it all goes horribly wrong. The computer becomes vain, doubtful and irrational and Armageddon by wayward technology is only narrowly avoided. Yet machines that can recognise our emotions and respond to them are a hot new area in computer design.
We're not talking about handing a global missile defence system over to a PC with a lot of feelings, but we are talking about improving the interface between humans and computers by making computers easier and more intuitive to interact with. For example, computers that sense when a user is getting frustrated could try a different way of explaining how to troubleshoot the internet connection. Cars that can tell when a driver is about to fall asleep could sound an alarm.
Though computers have become more attuned to how humans behave, there are still glitches in communication. Your PC may have the processing power of a small space station from the Seventies, but it can't tell if you're upset. Building an awareness of human emotions into computers would be very beneficial, whether it's through facial and smile recognition, through analysing the words you say or type, or monitoring the pitch of your voice or the style of your gestures.
The laboratory for Affective Computing at The Massachusetts Institute of Technology puts it like this: "Emotion is fundamental to human experience, influencing cognition, perception, and everyday tasks such as learning, communication, and even rational decision-making. However, technologists have largely ignored emotion and created an often-frustrating experience for people, in part because affect has been misunderstood and is hard to measure."
A computer doesn't know that you're frustrated; it just knows you're doing it wrong. Currently, you need to be a human to detect that sort of thing – or at least a dog. But there are some distinct commercial advantages in making emotional computers.
At the Consumer Electronics Show last month, Asus (creator of the Eee PC) revealed a range of concept computers called Waveface. It features a watch, a laptop and a TV, and as well as touchscreens and high-definition images these futuristic candy-coloured gadgets promised emotional intelligence.
"Every element in the user's environment, even down to the user's own physiology and emotional state, can be sources of data to help deliver the right information and services at the right time," is how Asus described its new wunder products. Jonney Shih, CEO of Asus, explains what that means. But it is not, it turns out, sessions where you tell the TV about your feelings.
Shih anticipates a future where the office, the home and the public space are all connected by cloud computing. The internet will be accessible everywhere and the amount of information on the internet will only grow.
"So much information now is available from the internet, from the cloud," he says, "so the challenge now is how to screen or filter that information so that you can get it when you really need it." Asus's Waveface products will use data about you (from where you are to your emotional state) to help tailor the information and services it offers to what's relevant to your particular situation.
To this end, Asus has a concept of an intelligent agent, a sort of invisible avatar for the user, that is built into the computer interface and acts on the user's behalf.
"We have what we sometimes call the intelligent agent," explains Shih, "which tries to help you get what you need at the right time and the right place. The goal is to achieve real-time, real-place and real-human information. Making sure that you always know where you are, know where your friends are, know your parking space and points of interest to you: restaurants, meeting places."
The details of the Waveface products are still shrouded in manufacturer secrecy, but we can expect them on the market in three to five years. But it is not just computer manufacturers who are intrigued by the new interactivity.
Other companies putting money into this area include car manufacturers. Toyota hit the headlines in 2005 with the Pod, an experimental car which signalled the mood of its driver to other road users by brightening, dimming or colouring the headlights. The car was embedded with voice recognition software by Scottish company Affective Media which uses the tone and the pitch of the voice as indicators of the driver's emotion.
As well as altering the headlights, if the car judged the driver to be stressed it would release soothing perfume, flick on some mellow music and perhaps suggest a less-congested route. If it sensed a driver was about to doze off, it would wake them with an alarm.
"Stressed drivers are bad drivers," says Ray Warde, CEO of Affective Media, "and so is a person about to fall asleep.
"The voice of a driver gives a lot of information about the way they are actually behaving at the time," he explains. "At the moment only certain models of cars will talk to the driver but that will become more common and drivers will start to control cars through speech recognition."
There are innumerable ways car or mobile-phone makers could incorporate this software into their devices.
While manufacturers and money men consider the commercial uses of emotionally-intelligent computers, academics such as Dr Russell Beale, a senior lecturer in Computer Science at Birmingham University, take a broader view on why it's useful for computers to be able to detect our emotions and then respond as if they too were emotional beings. We already treat computers like they have feelings. Beale mentions a seminal book by two scientists called Reeves and Nass, The Media Equation: how people treat computers, TV and New Media like Real People and Real Places, which was published in 1998. The book describes a psychological experiment showing that people treated computers (at that time, standard Nineties terminals with green lights on a screen) as if they were social beings – as if the machines had emotions and personality.
"Without these systems trying to be sensitive or 'affective' in any way" Beale says, "people were still viewing the computer as if it had responses and they didn't want to upset it.
"Since then," he explains, "people have realised that when you're dealing with an interactive system the more interactive it gets – the more you start thinking, 'Oh, this is how I interact with people.' That makes you think about the aspects of interaction that are missing and emotion is obviously one of them."
The second reason is engagement. "Research so far has shown that people are more engaged with a system that seems to demonstrate emotional responses," Beale says. "It makes people more caught up in the interaction that's happening, whether you're talking about games or learning or healthcare systems."
There are social uses for emotional technology. Beale is working on projects that use emotional computers to help people: from getting them to eat healthily to giving up drugs.
However, there are limits to where we want emotional computing. It's a question of context. "People often say 'it's going to be much better dealing with interactive computers because it's going to understand me better', but frequently we use computers simply as tools," Beale says. "If you're writing an urgent report you don't want Microsoft Word saying to you, 'You seem a bit stressed, would you like to take a break?' You'd be like, 'No, I've got a deadline.' Sometimes you just want systems to respond as tools."
He cites the Microsoft paperclip as an example of an inappropriate intervention. "It would ask you sensible stuff, but for most people, it kept popping up at the wrong time and getting in the way."
Yet there other cases such as teaching and sports coaching when emotional response, whether encouraging or slightly negative, will boost performance. Beale gets to the heart of the emotional computer: "In the same way that it's easier talking to someone who has empathy with what you're saying, who gives you extra responses apart from just a straightforward verbal reply, it makes our everyday interactions much easier.
"In the right cases, putting empathy into computer systems is really going to help."