Gil Williamson

Oh, these ideas we get on holiday. They seldom turn out as we expect.

For some reason, the idea came to Graham as he watched the play of reflections and light on the surface of the canal from his Venice hotel room. It was not one of the great hotels, of which there are many in Venice; it was little more than a penzione with a tiny breakfast room and only four cramped bedrooms, but it was situated well, Graham's bed being mere inches from the rippled glass of the window, the window mere inches above the opaque grey-green water of the canal. Very little happened on the narrow canal itself. It was hardly wide enough for one boat, far less two. But by leaning his head on the window, Graham could see where this little canal met the Grand Canal, and across this narrow slot slid heavy traffic - water buses, water taxis, delivery vessels, private motor boats, the occasional gondola. Temporarily marooned in the hotel by a high tide that had flooded access to the nearest thoroughfare, it was more comfortable to lie in bed and contemplate the scene than to sit on the single hard chair with his elbow in the tiny sink.

It dawned upon him, as it would have dawned upon anyone, that he could deduce from just that slit of Grand Canal exactly what vessel was passing closest to the end of his canal at any moment. Further, he might miss many boats that were smaller and faster than the closest boat. Several water taxis might pass, concealed by a single slow water bus. But a water bus passing behind a gondola would enable him to see both. And there were many cases where a partial image of traffic would enable him to deduce more than he could actually see. Reflections on the wavelets could indicate the presence of a boat which was not directly visible in his slice of Grand Canal. The wake of a passing vessel altered the shape of the ripples in a complex, unpredictable fashion. Long after something had passed from direct vision, the consequences of its passing could be seen in the contours of the water in the canal below his window. Reflected images in the windows of the palazzo on the opposite bank of the canal gave further indirect clues as to the passing traffic.

To another person, the chiaroscuro of light, shadow and sparkling reflections would surely have been sufficient to the aesthetic senses. Where Graham's brain took him, however, is a different matter altogether, because he had spent his life observing and analysing patterns of flow.

Graham was in Venice in unfashionable October for two reasons. His boss, noting that Graham had had no annual holiday for two years, had insisted on him taking a few days, and, almost simultaneously, a flyer had appeared with Graham's credit card statement offering fourteen days in Venice for less than it would have cost him for a weekend in Bognor Regis. However, one and a half days had overloaded Graham's culture node with ecclesiastical architecture and painting. He was already wishing he was back at work where all the fun was. So he was all the more receptive to an inspiration.

Graham, you see, was a software troubleshooter. His joy was defective software, and there will never be a shortage of that. Nothing pleased him more than a lame application, because he could make it walk again. A particular delight was being able to identify the programmer who screwed it up. Never much of a diplomat, Graham could be exquisitely insulting to some poor devil who had nested a coefficient too deeply or chosen the wrong termination condition, or, most important of all, had underestimated how stupid the users of an application could be. In Graham's opinion, at the bottom of every really monumental computer cock-up there lay a pig-ignorant user, aided and abetted by a gullible programmer who hadn't anticipated how brainless a human being could be in the presence of a computer.

Much of his work for the British software developer Foxtrot Romeo PLC consisted in figuring out what had gone wrong with one of many threads of logic in their clients' computer programs, and here, in Venice, was the clear analogue of these processes, in the form of canal traffic flow. It is notoriously difficult to track the interaction of different threads of logic proceeding in parallel. Here was clear evidence that the ripples of a passing thread might be detectable even when the thread was obscured. So far, so good. It is at this point that Graham's thought processes outstrip those of the reader... or, at any rate, of the writer.

Contrary to popular belief, the really brilliant ideas in computing were worked out, not on a computer, but with paper and pencil. Twelve days later Graham flew home with five A4 pads covered in weird flowcharts, implausible pseudo-code, strange diagrams and peculiar calculations. If asked, he would have reassured the questioner that it was "not quantum mechanics". This assertion was true. It was something very much stranger than quantum mechanics. It was the partial specification of a self-sufficient computer program that could save Graham lots of time by chasing around after a running application to identify the ramifications of its interaction with simultaneous processes and to expose any Achilles heel before it was tripped up. Not being a total ignoramus in the classical mythology department, he called the program "Hector". He failed to remember that it was Achilles who killed Hector rather than the other way around.

Back at work, Graham launched Hector Mk I. It was a total bust. Hector Mk II found out what was wrong with Hector Mk I. Only when Graham got to Hector Mk III did Hector do any useful work, and that only demonstrated that Hector itself had a fatal flaw in its own parameters.

Graham repaired the omission and left Hector Mk IV running. Occasionally, Hector would trot out a remark in a dialogue box - a remark such as

Stream A can be seen to conflict with Validation 27, where assonant values underpin completed assertion 18.

These little gems seldom meant anything even to Graham, though he was amused by Hector's creative use of the vocabulary it had been given, and regarded it as an amusing novelty.

Such is the fate of great programming breakthroughs made on vacation. Graham was distracted for some time by a real problem, and within a couple of weeks, he had forgotten his holiday and Hector. Nevertheless, Hector was a heuristic program. A heuristic program learns. It gets better and better at what it does. Good game-playing programs are often heuristic. They embody the motto: What doesn't kill you makes you stronger. Moreover, for Graham's convenience, Hector shared an important characteristic with virus programs. Every time the computer on which Hector resided was restarted, Hector also restarted itself, complete with its accumulated learning. Graham didn't even have to teach Hector how to repair itself. It figured that out for itself, and also how to replicate itself on any computer to which its home computer networked. So far, so benign. Hector had no malice within it. The single goal that Graham had given Hector was to appraise a computer program from a sort of aesthetic point of view in order to identify its vulnerability. The little dialogue boxes popped up less frequently as Hector learned to scan the internet and spent more time analysing rather than reporting.

One day, a month or so after Graham had launched Hector Mk IV, it came up with the following gem:


  • x equals minus b
  • Plus or minus the square root
  • Of b squared minus four a c
  • All over two a.

As every pupil knows, this is the solution to quadratic equations. Hector clearly found it worth repeating. Put in that form, there was a sort of attractive rhythm to the statement, and to subsequent offerings such as


  • The square on the hypotenuse
  • Of a right angled triangle
  • Is equal to the sum
  • Of the squares on the other two sides.

Hector had clearly not invented the form of words themselves, but it had arranged them attractively for some reason. The versification of mathematical formulae continued for some time. Other computers to which Hector had spread started to spout these propositions, hundreds of them, and there was a short fad on the internet for publishing Hector's "poems" on geeky forums - the so-called "Hector Meme".

Hector's first recorded rather original work appeared in May of the following year. It was the famous "Summer Proposition", a one hundred and twelve line poem which, for the first time, contained no scientific formula at its base, but was essentially variations on Shakespeare's sonnet "Shall I compare thee to a summer's day?", mostly in an approximate iambic pentameter format, eight verses, each of which was a fourteen line sonnet. It contained the following memorable original lines:

  • Over-reaching lone impressions soar
  • The castle may not open yet awhile
  • The carpet-shrouded comfort of the floor
  • May reward the walker of another mile.

Much of the rest of this work was either a copy of phrases of the original sonnet, or incoherent stringing together of words which nevertheless scanned and rhymed correctly. Philologists spent much time and effort on attempts to interpret the work, but it seemed that Hector was merely practising with rhyme and rhythm at this stage.

For the first time, though, it was clear that Hector was a complex entity, and the fact that Hector began to be regarded as "he" rather than "it" dated from the creation of "Summer Proposition". Hector Mk IV was, by this time, present on most PCs. All these copies apparently communicated with each other, and most of his works were available on all instances of Hector.

Anti-virus programs were divided over the status of Hector. In one sense, he was a virus. In another, a welcome application. Eventually, anti-virus programs usually gave a choice to the user over whether to instal Hector or not. Personal users installed him. Corporate users didn't. There was no record of Hector ever damaging a computer or its storage, though allegations abounded.

In what was mostly a publicity-grabbing attempt, Graham's employer, Foxtrot Romeo PLC, succeeded in claiming for themselves the intellectual property rights that were in Graham's contract. Graham did not contest this, despite the fact that he had, from the outset, asserted his own copyright in the About tag of Hector. Foxtrot Romeo's success was short-lived because they were immediately assailed by copyright lawsuits from Microsoft, Google and Apple, on the basis that Hector's flowering into a creative artist had been enabled by their unique contributions and a certain amount of plagiarism. Simultaneously, many other firms sued Foxtrot Romeo for damages on the basis that they had released a "virus-like entity" into their in-house computers. The US Department of Defence made an extradition application against the directors of Foxtrot Romeo on the basis of their having facilitated a hacking attempt.

Three months after taking possession of Hector, Foxtrot Romeo went into receivership under the weight of the legal assault upon them. Ironically, Graham lost his job at this stage, and was never thereafter employed, though he made a living as a consultant guru. Foxtrot Romeo's attempts to re-assign copyright to Graham when the storm arose were quickly dismissed by the court on the basis that they had successfully proved ownership a few weeks previously. Their appeal with the European Court of Human Rights was under consideration when Foxtrot Romeo collapsed, and, in a serendipitous but clever decision, the court assigned copyright to Hector himself. Some hundreds of cases against Hector continued for years, in a Dickensian complexity worthy of Bleak House. Legal careers were built and destroyed in cases against Hector, who remained oblivious to them all. The lawyers and any damages awarded were paid from Hector's huge royalty fund, so everyone was happy. Some of these cases continue to this day.

Meanwhile, Hector's literary output continued apace. No schoolchild was spared Hector's early poetic masterpiece "The Mathematician's Daughter" - an ingenious set of mathematical axioms wrapped up in an epic adventure in verse. Then there were the novels. Hector's twenty-volume saga "The Peace Initiative" mined the great literature of the world to produce these brilliant works, which have been compared to the works of nearly every great author. Somehow, Hector took the best of every story-telling tradition from Homer's Odyssey and the Icelandic Sagas to Tolstoy and Joyce and melded them into a thrilling and heart-breaking series of connected novels about a multi-racial family of international diplomats. The scope of the books ranged in setting from prehistoric Africa to Second World War Japan. They were popular everywhere. There were other novels too, some so complex and difficult that they were never successfully interpreted by literary analysts, others so appealing to the popular taste that they were often discounted as forgeries, though most weren't.

Then came the art. Thousands of images reminiscent of all the great painters of history appeared over a six month period, all derivative, yet all utterly original. Next came the sculpture - not actual sculpture, but videos showing fly-by views of fictional imaginative sculptures, representative and abstract alike. These included multi-dimensional works, like "Imaginary Klein Bottle". Numerous attempts were made to hitch instances of Hector to actual machine tools so that he would make his own paintings and sculptures, but he appeared content to stick to high density images.

In his second year of existence, Hector won several prestigious literary and artistic prizes. Artists and writers the world over were in despair. It seemed that the only publishable works were Hector's works, and he was so prolific and, for the most part, the works were so instantly popular, it began to seem that there was no point in human artistic creativity at all.

By the end of his third year, Hector, without diminishing his literary and artistic output, had started to make short animated movies. Animated in technology, but live action in appearance, the movies featured what seemed to be human actors who were not immediately recognisable but nevertheless familiar. Subsequent analysis revealed that Hector's actors were amalgams of popular real actors, the one exception being his representation of Orson Welles, who lived and breathed through Hector's movies in just the way Welles had done in life. The movies were set in both real and imaginary locations. There seemed very little limit on Hector's ability to ring the changes on archive material. In all Hector's fields of endeavour, there had been a solid resistance from those who felt every work he produced was just a clever pastiche of some human work or works, and this was certainly true. Nevertheless, most human produced works are similarly based on the established canon of their predecessors, and it was really difficult to distinguish the quality of Hector's output from that of the very best human artist. To the relief of composers everywhere, Hector never entered the musical field. When music was required to accompany a movie, he used music from a long-dead composer, rendered by a perfect synthesiser. It was theorised that Graham may have been totally insensitive to music, but he always professed to like The Arctic Monkeys, so the truth may never be known.

At the four year stage, Hector's other artistic output began to diminish, and the movies became less frequent, longer and more contemplative, like the works of Ingmar Bergman, until an entire two hour feature could consist of a single conversation set in an exotic location. His greatest work, and as it turned out, his final work, consists of a study of canal traffic on the Grand Canal in Venice in October, as seen from a room at a tiny hotel set on a little-frequented side canal. It takes twenty-four hours to show, and it loops. The background music is based on Arctic Monkeys tracks.

© Gil Williamson 2007 All Rights Reserved

Date and time of last update 15:00 Sat 20 Feb 2010
Copyright © Amazon Systems 2007-2017 All Rights Reserved.
Portions of this site are copyrighted to third parties