Following last week’s outline of some of the recent trends regarding the use of AI-generated content, in this article I will discuss how we as readers might approach the issue.
This question derives from a recent episode in my reading circles. A fellow reviewer refused to read a novel because they suspected AI use in the writing. According to them, at the start of that book, there was an overabundance of certain words, phrases and punctuations, as well as certain ways of describing inanimate objects.
While none of the provided examples alone could indicate AI use, this fellow reviewer insisted that because so many of them were found together in a relatively short text (e.g. the first five pages), it was a sure sign the novel was at least partially AI-generated. Therefore, they determined the book was unworthy of reading.
Which I disagree. Particularly when I later discovered this statement on the novel’s author website: “All written content is entirely original and crafted as my own work without the use of generative AI.”
Should we trust the author, or accuse them of lying? Do we read for ourselves and for pleasure, or do we abandon a book because others cast doubt on its merits? Do our views and opinions matter as readers, not just to each other, but especially to all the writers/authors out there? Does anyone even bother to care?
To borrow from Monty Python, my position is to “always look on the bright side of life”. Reading is and will always be a subjective experience. Indeed, based on my own perusal of the aforementioned novel, various common and minor imperfections have been identified that can only be found in human writing.
Did I have fun reading that novel? Yes. And that is all that matters – to me.
With that said, for those who are interested in detecting AI-generated content, many experts in the field have offered their advice.
The most frequently cited red flag is when a piece of writing feels “off”. It can be bland, overly verbose or repetitive, too perfectly structured, too consistent, too polite, or simply too neat. In other words, there is no “human touch” – authentic, personal, real-world and often messy emotions and thoughts.
But the most crucial criterion – which I think applies to both fiction and non-fiction – is whether or not the writing “moves” you. Does it offers something original and refreshing, not just informative but also exciting and inspiring? Does it contain “lived experiences” – unique knowledge, understanding and insight that real people have gained after going through significant events in their lives?
Ultimately, AI is trained on massive datasets sourced from content created by humans, and has learned to categorise information and identify patterns. While technological advancements and human feedback will continue to help improve the quality of AI-generated content, whether or not this will make reading more pleasurable remains unknown. But one thing is certain – all human writers/authors out there need more support than ever before.
Note: This article was originally titled “Regarding AI-generated content (part 2 of 2)” and published under the title “How readers can approach AI-content” by Ranges Trader Star Mail, February 3, 2026, P.24.
Voices under the Sun 陽光下的聲音 