Science, technology & design

Watch where you look

Imagine if you were playing a video game and you started to feel tired and irritable -then the game started to change before your eyes. Or you looked carefully at an ad, another one appeared, then a voice said, “sorry to bother you”. This is the future of facial monitoring. It simply means using a webcam to watch how you react to whatever you are doing, and responding to you.

Advertisers are keen to use this kind of technology, particularly as it can reveal even brief moments of looking at an ad. At the moment, you have to click on it for anyone to know. But companies like Realeyes are becoming very adept at gauging moods by plotting the position of facial features and creating algorithms to interpret them. For example, you may partially raise an eyebrow to indicate surprise.

Other areas can use this kind of surveillance: security, gaming, education and health care. Realeyes is working with Kaplan in Hungary to see how children respond to games that teach them English so they can be made more effective. Philips is developing a similar technology to keep a more subtle eye on hospital patients, such as newborns. Its vital-signs camera can measure heart and respiration rates very accurately by looking for changes in skin colour and the rise and fall of the chest while breathing. Philips plans to launch an app for the iPad2 so you can measure your own heart and breathing rates using two webcams.

There seems to be a theme of monitoring and surveillance in our articles this quarter. It ranges from medical applications to social networks, from airports to canoes. It all seems to be for our own good. Does anyone have the feeling that Big Brother is watching?

The machine sounds human

For some reason, we like hearing a machine trying to sound human. Though we don’t like it much when a human sounds like a machine. So as early as 1939, Bell Labs was demonstrating the first artificial voice. In the 1970s, we used canned audio responses made by splitting up spoken sentences artificially – but they sounded as if they came out of a can. Today’s AT&T’s Natural Voices still cannot capture distinctive emotion in the human voice.

The CEO of VivoText, Gershon Silbert, is determined to remedy this with a text-to-speech engine. It uses a combination of a proprietary voice-sample database that includes emotion, with software that generates virtual music with expressiveness. VivoText interprets italicised and capped words and analyses words for context, such as where to accentuate the “you” in “what can we do for you?”.

VivoText will target electronic publishing, especially the small but growing field of audio books. The CEO argues that his speech engine will work well for informational or technical books, rather than competing with an actor reading Shakespeare. This begs the question: why the search for ‘emotion’ in the machine voice? It will be interesting to see whether listeners to audio books, who may be older or sight impaired, want to hear a machine read to them.

Knowledge and ignorance of climate change

If you are sick of hearing about climate change, skip this article. (See: Climate change matters – or does it?) Michael LePage of New Scientist recognises there is much we know about this topic, and much we don’t. Given this tricky combination of knowledge and ignorance, he still makes the case that the planet is warming and we are largely responsible for it.

First, we know greenhouse gases are warming the planet. This could be because more heat is reaching Earth (it isn’t) or less is escaping. There are several reasons why less is escaping, one being because of carbon dioxide in the air. Carbon dioxide absorbs infrared radiation that would otherwise burst into space and reradiates it towards Earth, as shown by Satellite measurements. LePage does not explain the other reasons why less is escaping.But we don’t know how far greenhouse gas levels will rise because it depends on how much humans cut their emissions. It is almost impossible to predict the behaviour of a globe full of humans. We also don’t know how the Earth will respond, because, so far, the oceans have removed a third of CO2. Also huge amounts of greenhouse gas are contained in peat bogs and permafrost. Apparently, the IPCC ignores these possibilities because they are too difficult to quantify.

Second, we know other pollutants are cooling the planet. One is sulphur dioxide from volcanoes and, since 2000, from coal-fired power stations being built in China. The Chinese are installing sulphur-scrubbing equipment to reduce this effect. But we don’t know how great cooling will be because pollutants have complex effects in the atmosphere. IPCC models assume more modest warming is partly counteracted by modest cooling of this nature.

Third, the planet is going to get much hotter because of ‘positive feedback’ effects. For examples, water vapour is a greenhouse gas and, as snow and ice melt, they stop reflecting sunlight back into space. We can’t say how much hotter it will be but doubling CO2 warms the planet by 2-3 degrees. Climate models tend to include fast feedbacks, but not long term feedbacks, such as ice sheet coverage. This means we may underestimate the heat. We also don’t know how seriously global warming will threaten life. Up to a third of animals may disappear if they can’t adapt to their environment and how well we fare depends on how well we prepare.

Four, there will be more extreme floods because warm air holds moisture, and more extreme droughts because the sun starts to dry out the soil rather than just evaporating water. We are less able to predict more hurricanes because they only form when conditions are just right, say, high sea surface temperatures with not too strong winds. But they are less likely in temperate regions. Last, we don’t know if or when tipping points will be reached, for example, if the Arctic suddenly cooled, or Antarctica lost massive amounts of ice, or the Amazon turned to grassland. These effects would be irreversible.

Having read this article, you may conclude there is much we don’t know about global warming. Given such an ambivalent picture, you may find it hard to make up your mind. Or you may simply adhere to what you believed all along, which is human nature anyway.

Science loses its clout

Over the next 40 years, science could create more knowledge than was created in all of recorded history. Yet American policymakers increasingly reject it, citizens don’t want to read about it and Australian schoolchildren, among others, don’t want to study it. What’s going on?

In the US Congress, less than 2% have professional backgrounds in science while 42% are lawyers. They are trained to win arguments, according to this author, not to tell the truth. The religious right is also somewhat responsible, he admits, but the real problem is the second world war. This is when science was partly transformed from an exploration of nature into a weapon of destruction. The rise of postmodernism in universities also suggests, unlike science, that there is no objective truth.

The percentage of Americans who follow science and technology news very closely is declining. Public acceptance of evolution is also low, compared to other countries and only Turkey believes in it less. Senior high school students have rejected science as a subject in “staggering” numbers in Australia as 90% of them studied it 20 years ago, compared to only 50% today.

One way to change people’s minds is to find someone they can relate to, to argue the case. They start off relating more to the person than to the subject per se. It is important to frame a message according to cultural bias, for example, biblical creationism was reframed as ‘intelligent design”. It appears strong Republicans (climate change sceptics) are more likely to believe unambiguous graphs than text.

Sport goes sci-fi

It makes sense that McLaren, a company skilled in creating racing cars, could use the same technologies in bicycles, yachts or canoes. But what if it could take the same designs used for monitoring athletes and improving their performance into health care?
Sensors can be used inside an oar while rowing, for example, to measure speed, movement or angle. When these are linked to the rower’s heart, it is possible to assess how efficient or tired they are, and where time and energy is being lost.

McLaren understands that it is easy to take measurements but nobody can interpret the data fast enough to make it useful. Athletes can talk about it after training, but it would be more effective to be able to respond as it happens.However, many athletes dislike wearing heart-belts or wristwatches, so McLaren has developed wearable sensors embedded in clothing. McLaren is looking for a health care partner to market clothing sensors that could be worn by patients, or even by people who might appreciate an early warning system. Yes, another form of surveillance that seems like a good thing?