Daily Archives: July 20, 2026

As a high school teacher, I can attest that people are in fact getting dumber at an alarming rate. I remember the first time this reality hit me. It was a little more than ten years ago. I wrote something on the board. I was already annoyed that it was a white board and not a good, old fashioned chalk board which I really preferred. I effortlessly scribbled a few words in dry erase marker in my best penmanship. Catholic school, Palmer method. My handwriting is impeccable. Sister would be proud. As I turned around, I noticed a hand was up. I hadn’t written anything more than: “Homework: Read pp. n.-n.” I allowed the young lady to pose her question, curious as to what part of this was confusing. “Um,” she said, “I don’t know what that says.” After a brief back and forth with the student I discovered the reason behind her lack of cognition. It turns out the girl could not read cursive writing. I had heard this was becoming a thing but hadn’t believed it until that moment. She was so sincere in her humility, too. Until she was no longer humbler when I told her that I would not, under any circumstances, re-write my words. Then she was pissed.

I can also attest that over the course of almost a decade at this one particular high school, basic things just seemed to be eluding the majority of the students on a yearly basis like a snowball gathering steam into an avalanche. Case in point: part of my curriculum involved reading CS Lewis’ The Screwtape Letters. The book, in case you’ve never read it, is broken into twenty-odd “chapters” that are really just short, epistolary notes. In other words, each chapter is simply a 2-3 page letter written from one demon to another. The work, not more than 70 years old at the time I was teaching it, contained a few by-then-archaic English phrases; but not enough to be incomprehensible, especially given the context. When I began teaching that course, I could assign a chapter or two of reading per night and then jump into a discussion of the text the next day. By the time I left that school I had to resort to reading the text aloud to the students in class, stopping every minute or two to explain many of the words.

Enter AI.

Man’s propensity toward laziness is on full display with this nonsense. It is easier to accept what someone else said than to do one’s own research. And in many cases, one could rightly expect that certain sources are trustworthy. I remember asking students to research a topic and telling them not to use Wikipedia for the very reason that it was crowd-sourced. They used it anyway and I could tell. How could I tell? I had done the research already in my own studies and could discuss the material without a book in front of me. More to the point, I could take basic information and extract other points because I know how to think. With AI? Oh no… It’s far worse.

Take a look at this video and sit in terror with the implications.

Again, it’s one thing to use an app like ChatGPT to quickly search for something. It’s very fast and most of the time it’s not too wrong. But do you see what’s happening? AI models are feeding off of other AI models so all the errors get incorporated. This happens on a scale so rapid and so vast that within no time, the “slop” becomes accepted reality. Just last Sunday night, I used ChatGPT to find out which of the Republican Congressmen who had first been elected in 1994 were still serving in either the House or Senate. Odd topic, I know. Here’s the thing. I remember 1994. I was 16 years-old when that election took place. I remember Newt Gingrich and the Republican Revolution and the 54 House seat pick-up. I remember that two sitting Democrat Senators switched parties within the following days. I even remember their names! Colorado’s Ben “Nighthorse” Campbell and Alabama’s Richard Shelby. ChatGPT spat out the following: “Three members of the US Senate are the only remaining members of the class of ’94.” Really? Who are they? “Roger Wicker of Mississippi, Susan Collins of Maine, and Lindsay Graham of South Carolina” I then argued with the chat for the next five minutes that Graham had, in fact, died 24 hours earlier. Now, 24 hours was enough time that his death was universally reported, his Wikipedia page had been updated, and every computer in the world had recorded evidence of it. The chat? “If Lindsay Graham died, this is news to me. It must have been very recent.” The damn thing insisted that Graham was still alive and that I was wrong because it couldn’t find evidence of something for which the internet was dripping in evidence.

Do you know who loves this kind of confusion, this kind of stupidity, and this kind of loss of trust in verifiable truth?

I’ll wait.

Yeah, him. Friends, this is scary and I fear there simply are not enough of us left who care. I also fear that only divine intervention can end this and set man aright. We are at a Babel moment here. AI chatbots are now merging with humanoid robots. Sounds like something out of an Asimov novel. The very fact that we can use the word “hallucination” in reference to a machine ought to frighten each of us.

So what should we do?

Stay confessed.

Pray.

Frequent Mass and the sacraments.

That’s all I’ve got.