A hallmark of civilisation has been the drive to create unique stories that explore the human condition. Now robots are learning to write fiction. Is nothing sacred?

No computer has yet written the Great Australian Novel, because computers have some of the handicaps that afflict human writers. Writing is hard.

Although computers can work unhindered by free will, alcohol or divorce, such advantages are outweighed by a lack of life experience or emotions.

Even so, computers are quietly elbowing their way into the workplace. They’re flying planes, driving cars, selecting job candidates and writing news stories. The Associated Press employs a company called Automated Insights to create short news reports from raw data. More advanced software is working on longer pieces.

To be able to write fiction, a machine does not have to think like us but it must ‘understand’ patterns of human experience. While not discounting the life of the mind, computers are being fed millions of novels and short stories to ‘teach’ them character, pace and plot.

This is the goal of the What-If Machine (WHIM) project, a venture involving teams at five universities across Europe. WHIM analyses databases of human prose and then inverts or twists what it has learned to introduce a new idea for a story. Knowing what is typical is the first step in generating atypical stories.

Researchers at the Georgia Institute of Technology have been working for four years on a program called Scheherazade, which analyses crowd-sourced human anecdotes and then produces plausible short stories.

Stories generated by Scheherazade about everyday situations, are nearly indistinguishable from human-written short stories. Using code, they are taught to learn from their mistakes.

I was the former head of the creative writing programs at RMIT, where students were taught the difficult art of writing fiction.

The majority of the students were young, well-educated women, who wrote good middlebrow prose. The majority of the themes involved young women living in the inner suburbs who had gotten themselves involved in drugs, prostitution, a gender change or domestic violence. I never realised things were so grim in Fitzroy and Newtown.

Could a computer write those plots? Absolutely. The reason is they are programmatic. It’s innovative and unique writing that computers have trouble with. Could they write like James Joyce or Ursula Le Guin? Forget it.

The philosopher Leibniz speculated that humans wrote, uttered words and phrases that were derived from individual emotions and perceptions. Behind this was an important supposition. Humans are not only knowable but predictable (sort of). It’s this reductionism that angers ‘creatives’, because it harbours a nasty anti-humanist sentiment.

If our thinking and how we express our thoughts in language can be expressed in an equation or by a ‘middlebrow algorithm’, then we can also model the product of that thinking. That means a computer can write novels and poetry. Here is an example:

“Dave Striver loved the university – its ivy-covered clocktowers, its ancient and sturdy brick, and its sun-splashed verdant greens and eager youth. The university, contrary to popular opinion, is far from free of the stark unforgiving trials of the business world: academia has its own tests, and some are as merciless as any in the marketplace. A prime example is the dissertation defense: to earn the Ph.D., to become a doctor, one must pass an oral examination on one’s dissertation. This was a test Professor Edward Hart enjoyed giving.”

That opening paragraph was written by a computer at the Rensselaer Polytechnic Institute in the US. JM Coetzee can sleep safely. But it shows that computers are learning to obey syntax. They are learning to learn.

I know what you’re thinking. Computers aren’t alive. Their “writing” is flat and lifeless. There’s no surprises, nothing that rises up off the page and goes for your throat or heart. But such criticism may be premature.

Recently the Go grandmaster, Lee Se-Dol was defeated 4-1 by Google’s AlphaGo computer. In the second game, AlphaGo made a move so surprising – “not a human move” in the words of a commentator – that Lee Se-Dol had to leave the room to recover his composure. The computer threw a “crazy Ivan”. If it can do that, it can create radical plot shifts.

It’s intriguing that human beings, whose very consciousness is born from experience, may relate in some way to stories that have no experiential grounding. That truly would be science fiction.