I can't stand AI-assisted writing anymore. It's all over Substack. It's in my LinkedIn replies. It's everywhere. Like a disease slowly infecting our literary consciousness. I even see it from people who I know aren't bad writers. For example, a prominent person in the wealth management space recently posted the following:
Your dollar lost 53% of its purchasing power over the past 30 years.
That’s not an anomaly.
That’s the system.
Inflation isn’t just a number – it’s a silent thief.
There's nothing wrong with this message (I agree with it), but it's clearly written by AI. The "It's not X, it's Y" motif is the clearest giveaway of AI-assisted writing, especially when that motif is repeated back to back. The message might as well say "It's not human, it's AI."
But here's the crazy part—even the people complaining about AI are using it in their writing! I was recently reading an article from a history professor about how all of his students use AI to write their essays. As he stated:
I largely receive 400 variations on the same essay. The wording, structure, transitions, tone, even the closing sentences are largely identical.
But then the very next line reads:
This isn’t about a few students cheating. It’s about the collapse of an entire pedagogical model. The mass lecture, the take-home prompt, the standardized rubric—all built for a world that no longer exists.
What AI has done is not simply make cheating easier; it has made my entire form of assessment obsolete.
It's the classic "It's not X, it's Y" motif followed by a list of three items with the em-dash (another AI writing tactic) and finally closing with another "It's not X, it's Y" style line.
Excuse my language, but what the hell is going on?
If the history professor complaining about AI-assisted writing is passing off AI-assisted writing as his own, we're cooked. If the ultra-successful wealth management professional can't help but use AI (or have his social media team use AI), we're done.
Don't get me wrong, using the "It's not X, it's Y" style is not always indicative of AI-assisted writing. But when you see it so many times in the same piece, it's obvious. Why do people do post stuff like this? Do they think no one will notice?
I'm not here to bash anyone in particular (which is why I didn't link to the original posts above). However, I'm disappointed in how much AI has infiltrated the writing space.
This is what it must have felt like for natural bodybuilders when anabolic steroids showed up on the scene. AI is a performance-enhancing drug for mediocre writers. And, like bodybuilders on steroids, it leaves behind obvious clues when used.
But using AI isn't the core issue. How people use AI is. The root of this problem is that no one wants to think anymore. They want to outsource their thinking to machines. And, once you start doing that, it's hard to stop.
Trust me, I know the feeling. In early 2023 I used AI to help me write about various topics for SEO purposes. And it worked! My search engine traffic went up 6x from the beginning of 2023.
But, within a few months, I had to stop. As I publicly admitted, writing for a search engine was soul-killing work. I wasn't writing for me, I was writing for an algorithm. To get my passion for writing back, I had to start thinking for myself again.
I know how tempting it is to have AI write an entire article for you. After all, writing is cognitively demanding. It would be so much easier to just "push the AI button" and watch it do its work. But you have to fight that urge. Because once you cross that line, your writing dies. And your soul dies with it.
More importantly, you have to fight back because this is the last edge you will have. As more and more people outsource their cognition to AI, what do you think is going to happen? Do you think they are going to be smarter? Do you think they are going to be more knowledgeable?
Not a chance. In fact, the opposite is already happening. A recent scandal broke out at UC San Diego (UCSD) after it was revealed that the number of students whose math skills fall below middle-school level "increased nearly thirtyfold, reaching roughly one in eight members of the entering cohort." Shockingly, 25% of students in UCSD's remedial math course got the following question wrong:
Fill in the box: 7 + 2 = ☐ + 6
It's 3. There's no gotcha here. 7 + 2 = 3 + 6.
If this is happening at UCSD, which is a great school, imagine what's happening elsewhere.
We don't have to imagine though because the data paints a clear picture. As the New York Times recently reported, high school seniors had the worst reading scores since 1992 and the worst math scores since 2005. There are many possible reasons for this, and the heavy reliance on AI isn't helping.
College students in particular use AI at exceedingly high rates. This chart from Sherwood News illustrates how ChatGPT use drops off as college students leave school for the summer (in early June) and then starts to pick up again as they return (in late August):
ChatGPT usage by itself isn't a problem. But when so many educators are complaining about the rapid decline in cognition among college students, it is.
I'm not saying this to complain about the next generation. I'm saying it because it foreshadows what could happen to the rest of us if we aren't careful. The more we rely on AI to do our cognitive tasks, the more we will mentally atrophy.
Unfortunately, there is no substitute for thinking.
This will be as true today as it is in a decade. While the AI tools will get "smarter" at various tasks, outsourcing everything to them will only make you more vulnerable. As the American journalist Sydney J. Harris once said:
The real danger is not that computers will begin to think like men, but that men will begin to think like computers.
This is why thinking is the last edge we have. Because while everyone else is having AI draft their emails, write their code, and plan their lives, those who still dare to do some of the mental work themselves will win out.
As the proverb goes, "In the land of the blind, the one-eyed man is king."
Thank you for reading.
If you liked this post, consider signing up for my newsletter.
This is post 479. Any code I have related to this post can be found here with the same numbering: https://github.com/nmaggiulli/of-dollars-and-data