Monday 13th of July 2026
We don’t need no education. We don’t need no thought control.
We Don't Need No Education – A Reflection on Punishment, Pedagogy, and EdTech
The title of this post is a gentle irony. While I develop a system for the operational management of secondary schools, I increasingly feel that the "tools" offered to any institution already remind me of Michel Foucault's work "Discipline and Punish."
What's happening is that beyond Foucault's stance back then—alluding to the psychiatrist or psychologist as a jailer of the system—I observe that what's being sold to educational institutions is 100% punitive software, such as Copyleaks and the infamous Turnitin with its extremely high rate of false positives for analyzing that the text is "too good" to be human, analyzing perplexity (how likely or unlikely the word choice is) and the change between paragraphs (burstiness) in smooth transitions or a monstrosity where paragraphs jump almost from a cliff to the abyss. What I'm creating is significantly different.
I focused on the service for students, teachers, and administrators alike, but focusing on academic excellence rather than failing if a meter turns red and says "AI was used." Behind that madness, false positives, and the swarm of privacy problems it carries—those sites sell their API and use every text they receive to train their own AIs.
That's when I thought, "Damn... You can't punish these kids just because there's a probability they assisted themselves with text." So instead of the red meter, I decided to focus on the most appropriate pedagogical guidance for cases where the literary quality or any work doesn't correspond to the characteristics the teacher observes over time. Which requires precisely that the teacher works instead of punishing with a button.
It's a decision born from knowing that evils can't be done by an AI but by a human stubborn to infinity with the loss of control over what happens among their students.
It's inevitable to think of The Wall and the iconic song "Another Brick in the Wall Part II" in a situation like this.
Who knows, my good reader, that in the end I could be completely wrong, and what's really being sought is the punitive tool because it "facilitates the recovery of strict control" rather than a tool that simply delegates certain aspects of evaluation, leaving more time to dedicate to the student rather than simply pressing "detect AI" and then failing. After all, the second option is undoubtedly the fastest path.
It's for this reason that I had to think like Firestone when he noticed that Goodyear was winning by offering versatility over a creation he had generated with his engineers, but that ultimately, Goodyear had subtly improved and advertised better.
That's why I can't afford to fall into "pay to play" on social media, since what's spent on a B2B sale is a disaster, and each acquired client eats such a sharp part of the return on investment that it could make you drown in your own attempt to gain visibility.
More than that, I see it as the sale of old-school textbooks, where the seller took a percentage from the sale of a bulk of manuals from a publisher. It's a cleaner type of business where, ultimately and in the end, everyone wins.
Creating this type of software entails analyzing what the competition offers. Well, the competition offers swords and morning stars for better punishment. That doesn't seem appropriate, ethical, or negotiable to me.
Even less so when I was a chaos regarding my behavior as a kid, but my grades were outstanding even though I didn't bother to carry a folder with me and my classmates were my headphones and my music.
Of course, this generates misfits like me. That's not a good sales argument, so I have to move to the more comfortable zone of good grades without getting into the heart of the chaos or punishment.
It's a reflection on an educational system in which many students pay for the same detectors, but for different reasons. In some cases, out of fear of being labeled as AI when their work is written by hand. In others—and they're not few—to make their AI-assisted work pass as human... Obviously, the unfortunate ambivalence regarding the use of those tools is ultimately an Achilles' heel that sooner or later will make them become either a joke or something obsolete and absurd.
It's not a decision in favor of progress, per se, it's more of a decision in favor of the human, understanding that the machine is there to improve, collaborate with education, and make those dangerous teacher insecurities not exist anymore rather than deliberately feeding them to boost sales.
Anyway... Another day in the paradise of the sternest ignorance, but creating with the idea that ignorance can't contribute anything favorable even in the medium term, and that no one should pay for an AI-assisted paragraph.
Thanks for reading this post to the end (if you made it...) And have an excellent start to the week.
© 2026 Armando -Armand- Grandinetti. All rights reserved. No part of this post may be reproduced without prior written permission. Built with pedagogy in mind, not punishment.
What's happening is that beyond Foucault's stance back then—alluding to the psychiatrist or psychologist as a jailer of the system—I observe that what's being sold to educational institutions is 100% punitive software, such as Copyleaks and the infamous Turnitin with its extremely high rate of false positives for analyzing that the text is "too good" to be human, analyzing perplexity (how likely or unlikely the word choice is) and the change between paragraphs (burstiness) in smooth transitions or a monstrosity where paragraphs jump almost from a cliff to the abyss. What I'm creating is significantly different.
I focused on the service for students, teachers, and administrators alike, but focusing on academic excellence rather than failing if a meter turns red and says "AI was used." Behind that madness, false positives, and the swarm of privacy problems it carries—those sites sell their API and use every text they receive to train their own AIs.
That's when I thought, "Damn... You can't punish these kids just because there's a probability they assisted themselves with text." So instead of the red meter, I decided to focus on the most appropriate pedagogical guidance for cases where the literary quality or any work doesn't correspond to the characteristics the teacher observes over time. Which requires precisely that the teacher works instead of punishing with a button.
It's a decision born from knowing that evils can't be done by an AI but by a human stubborn to infinity with the loss of control over what happens among their students.
It's inevitable to think of The Wall and the iconic song "Another Brick in the Wall Part II" in a situation like this.
Who knows, my good reader, that in the end I could be completely wrong, and what's really being sought is the punitive tool because it "facilitates the recovery of strict control" rather than a tool that simply delegates certain aspects of evaluation, leaving more time to dedicate to the student rather than simply pressing "detect AI" and then failing. After all, the second option is undoubtedly the fastest path.
It's for this reason that I had to think like Firestone when he noticed that Goodyear was winning by offering versatility over a creation he had generated with his engineers, but that ultimately, Goodyear had subtly improved and advertised better.
That's why I can't afford to fall into "pay to play" on social media, since what's spent on a B2B sale is a disaster, and each acquired client eats such a sharp part of the return on investment that it could make you drown in your own attempt to gain visibility.
More than that, I see it as the sale of old-school textbooks, where the seller took a percentage from the sale of a bulk of manuals from a publisher. It's a cleaner type of business where, ultimately and in the end, everyone wins.
Creating this type of software entails analyzing what the competition offers. Well, the competition offers swords and morning stars for better punishment. That doesn't seem appropriate, ethical, or negotiable to me.
Even less so when I was a chaos regarding my behavior as a kid, but my grades were outstanding even though I didn't bother to carry a folder with me and my classmates were my headphones and my music.
Of course, this generates misfits like me. That's not a good sales argument, so I have to move to the more comfortable zone of good grades without getting into the heart of the chaos or punishment.
It's a reflection on an educational system in which many students pay for the same detectors, but for different reasons. In some cases, out of fear of being labeled as AI when their work is written by hand. In others—and they're not few—to make their AI-assisted work pass as human... Obviously, the unfortunate ambivalence regarding the use of those tools is ultimately an Achilles' heel that sooner or later will make them become either a joke or something obsolete and absurd.
It's not a decision in favor of progress, per se, it's more of a decision in favor of the human, understanding that the machine is there to improve, collaborate with education, and make those dangerous teacher insecurities not exist anymore rather than deliberately feeding them to boost sales.
Anyway... Another day in the paradise of the sternest ignorance, but creating with the idea that ignorance can't contribute anything favorable even in the medium term, and that no one should pay for an AI-assisted paragraph.
Thanks for reading this post to the end (if you made it...) And have an excellent start to the week.
© 2026 Armando -Armand- Grandinetti. All rights reserved. No part of this post may be reproduced without prior written permission. Built with pedagogy in mind, not punishment.