AI and the Bleak Desert of Uninteresting and Boring Content
Humans provide the spice of existence, not the bytes of indifference.
The digital world has seemingly morphed into a hellscape game of whack-a-mole, with classes and tutorials, prompts for sale, and super-secret-exposing ebooks popping up everywhere.
You simply can’t escape it.
Generative AI has become the darling of business writing and the holy grail of marketing and copywriting. It has caught the eye of internet grifters as well, so the swelling of ads selling AI classes will continue to grow.
It’s touted as the ultimate Swiss Army knife of creative technology: it spots trending topics, crafts attention-grabbing narratives, and pumps out content at lightning speed.
Content.
When I hear that word, I am reminded that a 2-year-old’s diaper is filled with content, so ‘content’ in and of itself is not necessarily a good thing.
Now we can load our pages with these AI-generated word salads, basking in the glow of endless high-ranking keywords and ROI while skyrocketing to the moon.
It sounds too good to be true, right?
Yep. It is.
But the problem is not as simple as that.
Generative tools are not great marketing saviors.
The irreplaceable magic of human intellect, emotion, and creativity is still crucial for every instance of boosting corporate and business goals.
Content, it seems, must first be interesting in order to be effective.
A 1,000-word article that has 2% interest is going to end up driving people away rather than pulling them in.
The number one culprit is, of course, ChatGPT.
Originally launched as a chat interface, it soon found a nice home in the content marketing arena. Where they make, you know, content.
With the ease of setup and the semi-interesting generated products, it’s damn impressive stuff. Far better than expected, it seems.
It can defend an argument on a specified topic or shape content in a set format (a recipe, a series of steps, etc.).
And it handles a massive amount of data at an incredibly rapid pace.
A hell of a range and impressive agility, without the brutal confines of the human mind. Or the constraints of limited time.
And it has led to many late-night existential discussions at the latte bars and shiny chrome bistros of Silicon Valley.
And New York.
And Dallas.
And Fresno, Tulsa, and Little Rock.
“Are the days of the creative professional waning?”
“Is this the end for artists?”
“Do we know what we are doing?”
“What the hell is a prompt engineer supposed to study?”
“Will we all be living on food stamps while Arnold tries to rewrite history?”
Maybe.
But that begs another question:
What’s really behind this content-spewing beast?
The Large Language Model underpinning ChatGPT is a technical marvel, no doubt. But the secret sauce is the massive corpus of knowledge it was trained on.
Millions of articles and texts on tens of thousands of topics, each penned in a unique style, all ingested by the insatiable AI.
Gobbled. Up.
However, here’s the bitchslap of reality; while it can mimic the content and style of nearly any writer, it doesn’t really comprehend any of what it is doing.
It has no unique ideas, no authentic creativity, no aesthetic viewpoint, no point of view at all, and absolutely no personal beliefs on which to draw.
It really isn’t creative.
Not in the least.
AI is a glorified photocopier, a convenient idea-scanner, the heartfelt journals of a billion writers digitized into a single database of bits.
Thinking of AI tools as anything more than an incredibly sophisticated mimic, a digital pantomime who doesn’t speak our language in any meaningful way is to give them anthropomorphic traits that rival any kid’s animated movie.
And because we are human we see patterns and assign to those patterns the only things we have to contextually explain them — human traits.
“AI writes beautiful poetry.”
No, you see the pattern and accept it as poetry.
“AI makes beautiful photographs.”
It doesn’t make anything at all, it charts probability and you morph it into a ‘photo’ because that is the only explanation you can use.
“AI writes sales copy.”
No, you think it looks like sales copy, but in fact, it is a sales word expectorant. The blend of 10,000+ copywriters, some great, most not, into a gooey paste that sorta kinda looks like good copywriting… as long as you don’t look too closely.
Rhetoric engines simulate believability without regard for the substance of what they create or produce.
They can sound passionate. They cannot BE passionate.
And that is a distinction to be aware of.
They may write poetry, but they do not understand poetry.
They can tell a story without having a clue as to what it means.
AI can generate words of love while being incapable of love at any level.
AI has never had a broken heart, survived the death of a loved one, spent a perfect day in the sun, slipped into cool sheets on a warm evening, heard the sound of children playing, tasted a deep dish Chicago pizza with sausage, experienced an ocean wave for the first time, or revelled in the magnificence of a symphony orchestra belting out Beethoven’s Ninth Symphony.
It simply produces… content.
And considering that misquotes, fabricated sources, and sometimes citations of authors or scientists that don’t even exist may be what you risk with machine-generated content, it will require more babysitting than a house full of malicious 3-year-olds.
(3-year-olds cannot be malicious you say? Tell me you haven’t had kids without telling me you haven’t had kids.)
One lawyer was disbarred for using AI-generated legal precedents that never existed in an important case. (EDIT: A correction. The lawyer was sanctioned, not disbarred.)
Oops.
Did the AI do it on purpose?
No, purpose-driven action is human.
Was the AI bribed by opposing counsel?
No, being corrupted by money is human.
Did the AI do it out of ignorance?
No, ignorance is a human state.
Was the AI contrite, apologetic, or even capable of giving a shit?
No, those traits are human, not machine.
AI can’t discern truth.
It is unable to grasp meaning.
AI doesn’t know or care when it fucks up.
Its primary function is to assemble words into coherent and plausible sentences, which can possibly mislead the reader with unfounded sources. Based on the coherent and plausible sentences it was trained on, but actually written by humans with massive amounts of humanity.
That humanity is lost in the flood of, and I am sorry to say it over and over again… content.
Can we even detect the potential errors embedded in the petabytes of data it was trained on?
Should we?
See, the biggest problem is that it cannot question the veracity of its own output, only its plausibility.
So, it often concocts content that seems perfectly logical and meaningful in the algorithm but has no relationship to reality.
Nor does it take any responsibility when the shit hits the fan, and the depths of its indifference are revealed.
ChatGPT’s own disclaimer: “ChatGPT may produce inaccurate information about people, places, or facts”.
Is it worth it?
Do you dare tap this magnificent information beast?
Do you dare not to?
Conundrum.
No degree of algorithmic sophistication has yet been able to emulate the quintessentially human ability to breathe life into an idea through carefully crafted words.
Human words.
For humans to read, listen to, be engaged by, and fall in love with.
We want to believe it can, but we know — I mean, we really know down deep — that the crap it spews out sounds sort of, what… weird? Mechanical? Slightly fake, like the accents of B-list TV actors trying to emulate a Scot?
AI cannot originate even banal content without explicit, human instructions.
That means that at least half of its moniker as Artificial Intelligence is a lie.
It is anything but intelligent.
Fast? You betcha.
Predictable? Yep.
Able to fill a Wix Blog with a bazillion words about lawn equipment for your millionaire-creating affiliate site? Sure.
As platforms like ChatGPT ‘democratize’ content creation (whatever the hell that means), enabling companies to churn out word vomit en masse, we will inundate an already bloated content marketplace.
A marketplace with no increased demand.
Blog traffic is down.
Social media traffic is flat.
Online news and information sites struggle for eyeballs.
And this is the time to flood the pipes with lame, minimally interesting content written on the level of an Ikea instructional manual without the inimitable Swedish humor?
Today alone on my FB feed, there were two dozen sales pages offering to make me a millionaire if I paid $1 for 11,000 prompts “done for you”, ready to cut and paste while ordering that spectacular Ferrari.
Will AI be the next Metaverse?
Maybe.
I think it will gain traction more as a tool than a creative entity. There is no pent-up demand for AI stuff. We will still need humans.
Authenticity still wins.
Authenticity: flawed, dirty, glorious, hyperbolic, understated, challenging, mixed up, confused, flailing, stoic, emotional, and occasionally crazed.
That still requires humanity.
Every time.
Photo of the author by Carol Rioux, light-painted in Calgary.
Hi, I’m Don Giannatti, a photographer and mentor for up-and-coming photographers. You can find me on my own site, Don Giannatti, and at my Substack site, where I also publish for creative people. All free subscribers to my Substack have access to a free long-form workshop on the business of commercial and professional photography as well as my latest book.
