Science, technology & design

Tell me what to do

How can you find something that you didn't know you were looking for? We've written about collaborate filtering software before (which was first suggested by Dave Goldberg at Xerox Parc way back in 1992), but the technology is becoming increasingly common. The idea, of course, is to help people navigate their way through the plethora of everyday choices. Everything from which book or music CD to buy, to what film to see (or avoid). The theory is that you can predict likely 'likes' because similar people share similar tastes. Early incarnations of collaborate filtering required individuals to tick virtual boxes to explicitly express their tastes, but this has been replaced by tracking buying or viewing habits. Tivo video recorders are a good example of where the technology is heading because recommendations are based on your personal viewing habits, not generated by a centralised system. So what's next? Adding a GPS component will make information gathering and recommendations even more specific, because devices will not only know what you like, but when you like to do it. The privacy (and security) concerns of such technology are significant (eg easy government surveillance - or the fact that the biggest retailers will hold the most information thus perpetuating oligopolies). However, such concerns will probably be overridden by the sheer usefulness of such devices. How about using collaborate filtering to help you find holidays, jobs and even spouses you didn't know you'd like?

Innovation shock

The mobile phone is one the most successful innovations in recent history and yet its ubiquity has done little to dispel safety fears surroundings its use. These fears range from brain tumours and crashing planes to exploding petrol stations, none of which has so far been proven to exist. A recent MORI survey in the UK, for example, found that two thirds of people thought that the risk posed by using a mobile phone outweighed the benefits (but bizarrely they still use them). So are such fears a normal consequence of innovation? The answer would appear to be yes. When trains and motor cars were first introduced people were convinced that the speed would cause a variety of physical and mental disorders. Similarly, the invention of the telegraph created a widespread belief that the signals would interfere with the weather. So what's the thought here? Simply that new ideas are generally resisted, and the stronger, or more successful, the idea the more resistance there will be both direct (physical actions) and indirect (the creation of myths). Moreover, whilst it is possible to prove that an innovation is harmful, it is impossible to verify that something is totally safe.

Sleepy head? - stay in bed

Sleep Smart is a smart alarm clock developed by students at Brown University (US) to stop you waking up at the wrong time. As people sleep, they rotate through three distinct phases: light sleep, deep sleep and REM sleep, a pattern that is repeated every 90 minutes on average. The theory is that if you are woken up during the wrong sleep cycle (eg during deep sleep) you will feel 'under the weather' for most of the day. The product works by wirelessly linking a headband, containing electrodes and a microprocessor, to a clock that can be programmed with the earliest and latest desired wake up times. The product is being developed by Axon Sleep Research Laboratories - a spin off from Brown University.

Smart materials (part 3)

Materials are, by definition, made of the same material and thus have the same physical properties throughout - right? Wrong. Scientists at University College London (UK) have invented a way of creating customised structures where the structure and properties of the material can be designed to vary millimetre by millimetre. The trick, apparently, is to treat the whole structure as a series of 'microstructures' whose physical characteristics are modelled individually. Having created a virtual model of each microstructure, the macro structure is then built using rapid prototyping technology. Each layer of the final structure is then literally 'printed' out, one layer after another. The resultant structure is a honeycomb of anything from metals to polymers. A good example of the use of such materials is a crash helmet. Currently a crash helmet is a homogenous structure made to withstand a general force acting across the whole helmet. But what you really need is a structure capable of dealing with localised impacts. Using this technique, in conjunction with metallic foam, it would be possible to vary the size and location of the bubbles in the foam to vary the strength of the helmet.

When machines start to think for themselves

Singularity is the term used by futurists to describe the point where machines have developed to the level where humans can no longer fully understand or forecast what they are capable of. In other words, it is the point at which machines become more intelligent than people. The idea of artificial intelligence (AI) goes back to the mid-fifties, although Isaac Asimov was writing about robots back in 1942 (the word robot comes from the Czech word for drudgery). The true test for artificial machine intelligence also dates back quite a way to 1950, when the British mathematician Alan Turing suggested that we would have AI when is was possible for someone to talk to a machine without realising it was a machine. The sixties and seventies saw a great deal of progress in AI, but breakthroughs failed to come. Instead scientists and developers focussed on specific problems, like speech and text recognition and computer vision. However, we may now be less than a decade away from seeing the AI vision become a reality. A company in Austin,Texas, called Cycorp is only months away from putting a product called Cyc on to the Internet. Cyc is an expert system that will be freely available for experts and ordinary people alike to interact with. Unlike existing 'chatbots' the system is unusual in that it uses common sense. You can ask it a question and if it answers incorrectly you can correct it. The system learns from its mistakes, and the more it knows the more it learns.