At New English Review, Ken Francis offers some thoughts on current efforts to get an attempt to get Chat-GPT or other LLMs to do one’s creative writing. He tried and the outcome was not happy.

It’s not that there is no voice in there but rather, it is a voice a writer may not want:

This should come as no surprise, as the multiple teams of hi-tech clones who input the narrative/literary instructions into AI are predominantly composed (not all) of the Woke sages of Silicon Valley. Many minds might make light work but too many ‘geeks’ can spoil the aesthetic ‘broth’ by saturating it with the groupthink trite of groundlings.

To draw an analogy in technological other art forms, it is like using Terragen (a scenery generator program) to create renderings of a mountain-range landscape and forest. The result would more than likely resemble the ‘happy’ trees, mountains, and lakes from a Bob Ross painting (almost photographic), as opposed to a Van Gogh or a Turner…

But AI lacks, or has its limits, in organic literature permeated with the creative imagination from the creative individual, human touch. It will never understand the immaterial mental states of the human Mind (more accurately described as the Soul). Machines do not have spiritual minds; they have technological artificial brains.

The “Virtues and Limitations of AI,” April 2025

But to whom does that matter?

Of course, the critical question is, how much does the audience actually want originality and how much does it want the old familiar storyline — the one that sometimes fails to entertain but never challenges or disturbs? George Orwell (1903–1950) famously thought that machines could be programmed to write formula fiction. In fact, in his dystopic novel Nineteen Eighty-Four, his central female character Julia has a job minding just such a machine.

Francis senses that, for he ends his essay this way:

I hope this essay has brought some comfort to the few independent talented screenwriters out there. As for the teams of Hollywood mainstream laptop-tappers in the cafes downtown LA, I’m afraid things ain’t looking too good: They might have a new rival AI kid on the block, judging by the tonnes of predictable popcorn scripts churned out in recent years. ChatGPT might replace that—if it has not done it already!

If Francis is right, good, challenging literature will rise above the tide of automated junk just as it does today above the tide of potboilers. Perhaps the bigger worry in many places is, as Orwell foresaw in Nineteen Eighty-Four, that government will forbid writers to issue such challenges.