AUTHOR’S NOTE: In online discussion groups I frequently encounter ill-tempered sniping over whether specific written content is an original creation or whether it was assembled via internet “scraping” by generative AI. Because of my admittedly literate writing style, I’ve been waiting for the day when I’m accused of such derivative chicanery.

To see where I land on the AI spectrum, I ran the Grammarly detection program on a sample of my work and came up with “zero” AI incursion. Since I’m fond of using dashes for various purposes in my text (primarily for hyphenation and as a substitute for parentheses), and the presence of “en” and “em” dashes is considered a dead giveaway for AI manipulation, I expected to see a different result from Grammarly even though I started writing this blog long before such technological hacking was even remotely possible.

As far as I’m concerned, “a dash is a dash is a dash,” and I pay scant attention to the subtle differences between the various types, although I’ve been wondering why Medium insists on placing what look like excessively long dashes in my sentences despite having typed a hyphen on the keyboard. Generally, I prefer the shorter version so I try to overwrite the “em” dashes but don’t always succeed. Also, I never put spaces around the dashes in a hyphenated word or number but I always do so when inserting parenthetical asides. I just go with whatever looks best to me and let it pass. Regarding my reliance on the “en” dash with or without adjacent spaces, I understand that I’m in good company with J.R.R. Tolkien and other British writers.

When posting my essays on Facebook pages and other internet platforms, I’ve taken to including a disclaimer stating that no AI material was used. I will now take the liberty of quoting one of my followers on that score from a recent exchange:

“It’s frightening to see that the ability to construct longer, compound sentences is now confused with AI use. I put it down to inadequate consumption of actual books, and variety of written material. People are no longer developing “an ear” for language. At this point in time, AI generated text is usually still identifiable by that “sound,” and it has almost nothing to do with em-dashes or en-dashes. So sad that you even need to add the disclaimer.”

I could not have said it better myself, and I responded as follows:

“I don’t know about dashes, but I would be more suspicious of semi-colons. Kurt Vonnegut said “Don’t use them, they only prove that you went to college.”

In another conversation I mentioned that the only time I consciously deal with AI-generated content occurs when asking Google for information because it simply can’t be avoided. However, I take what I receive and translate it into my own thoughts on the subject, letting AI do the “leg-work” but never just copying-and-pasting the results into my writing. I do the same thing when drawing inspiration directly from another author’s books, quoting verbatim and acknowledging the source where appropriate but otherwise turning any insights into personal observations using my own words. I anticipated that Grammarly would call me out on this, but it didn’t.

I believe that AI “scraping” is a disaster for copyright protection and anti-plagiarism efforts, and I don’t see any effective way to stop it. In my own case, everything I post is free to use as long as it isn’t exploited with commercial intent, and I’m not trying to monetize it in any way. But I fully expect to eventually see something I wrote show up intact in someone else’s pirated publication, and the alleged author won’t even know where it came from. I grumble about AI in the same way I complain about genetic engineering: just because we can do something doesn’t mean we should.

If only it were duly credited, I might see this as flattery since I have no fear that AI will be able to tackle the subject of divination in the near term as imaginatively as I have over the last eight years. That’s not hubris talking, it’s just a reasonable assumption based on what I’ve seen so far. After all, a parrot can only repeat what it has been taught. With AI appropriation, the output is only as good as the input and, given the sheer magnitude of its reach, the aphorism “garbage in/garbage out” always applies.

I’ve been told there is such a thing as “ethical AI,” but when it comes to the written word I don’t know how it can distinguish between “fool’s gold” and the real thing without human moderation if the intelligence wasn’t introduced at the point of origin. Soon or later, AI is going to start cannibalizing itself, and we will finally enter the “Age of Idiocracy” when professional writers and artists are forced out of the process and reduced to slinging burgers.