Science, art, magic and expert systems

Citation

Loughlin, C. (2010), "Science, art, magic and expert systems", Industrial Robot, Vol. 37 No. 6. https://doi.org/10.1108/ir.2010.04937faa.001

Science, art, magic and expert systems

Article Type: Editorial From: Industrial Robot: An International Journal, Volume 37, Issue 6

Arthur C. Clarke once said that “any sufficiently advanced technology is indistinguishable from magic”. Playing around with some of the toys we now have, like iPhones, Google Earth and Global Positioning System (GPS) positioning, it still seems like magic even though we all have at least some vague understanding of how they work.

If we were to step back 2,000 years or further and show our GPS plotter to an ancient druid, then they would be very impressed, but if we informed them that it uses satellites to fix the position they would understand the concept at once – after all, they would have been very used to studying the night sky looking at the stars and the position of the Moon (also a satellite), to both give them an idea of where they are and also to be able to predict impressive events like the summer solstice and eclipses, and even foretell the future with modest certainty.

To the common man listening to our druid, it would all have seemed like magic and comforting proof of their religious doctrine. To the druid himself, it would have been a very welcome meal ticket.

The tricks we see practised by magicians and illusionists cease being “magic” as soon as we understand how they are done. They remain very clever and a lasting tribute to the skill of the exponent – but they are no longer magic.

If magic is therefore something clever that we do not yet understand what about art? At university the subjects are frequently segregated between “art” and “sciences”. But what is art? Why is it that one picture that is made up of patterns of coloured paint is considered art, while another is written off as an amateurish mess? The difference can be hard to define but normally “meaning” or “pleasing to the eye” enter into the critical debate. What also is the difference between music and a random sequence fed into a tone generator?

Music may give us some clues – many mathematicians have a natural inclination towards the writing and appreciation of music, and can wax lyrical about sequences and interlocking cycles of modulated tones. Such an analysis requires study after the event, but the original real-time listening gives instant and instinctive pleasure.

There must be some underlying rules that make music, and other art forms enjoyable, if there were not then we would never have musical hits. Some of these rules may be easy to identify such as a beat that synchronises with our heartbeat and other natural rhythms, while others may be more obscure and, as yet, intangible.

It is not unreasonable to consider that in time we will understand why a particular piece of music or a particular painting gives us pleasure. When that time arrives we will still gain pleasure from the art and our emotions will still be roused, but we will perhaps have lost at least some of the “magic” that the art currently evokes?

The last 30 years or so have seen the development of expert systems and neural networks. If an industrial process is difficult to control or an image is difficult to analyse then we turn to expert systems or neural networks to come up with a solution. These work by feeding the incoming data into a long list of weighted rules and – hey presto – the answer pops out as if by magic. The rules are created either by asking an experienced person what should be done under an array of circumstances or by teaching (programming) a neural network by showing it a broad range of experiences and telling it the desired outcome. These systems use experience and instinct to come up with a solution when conventional logic and If-Then-Else constructs have been found wanting. If we had a better understanding of the original task then our new found knowledge would make such systems redundant, or at least unnecessary and optional.

In the end it all comes down to science – we just do not know it yet.

Clive Loughlin