Hello everyone!
I hope you’re all doing well. It’s been a bit of a whirlwind week here at Mr M HQ—the toddler has discovered the word "why" and is currently using it with the relentless precision of a prosecuting barrister. It’s exhausting, but it does make you think about how we learn, doesn't it? We ask questions. We struggle with answers. We grow.
Speaking of learning, I’ve been diving deep into a rather heavy academic report recently titled The Algorithmic Academy, and honestly? It’s kept me up at night.
We all know Artificial Intelligence (AI) is here. Your kids probably know more about ChatGPT than we do. But while schools are scrambling to decide whether to ban it or embrace it, there are some serious hidden issues that we, as parents, need to know about.
This isn't just about "cheating" on homework. It’s about fairness, creativity, and whether our children are actually learning to think or just learning to manage a robot.
So, grab a cuppa, and let’s unpack the "Algorithmic Academy" together.
The "Use It or Lose It" of Thinking
The report talks about something called "cognitive atrophy". It sounds terrifying, doesn't it? Essentially, it means if our kids outsource the hard work of learning—the struggle to write a sentence or solve a problem—to an AI, their brains stop building those muscles.
We want our children to "flourish"—to develop empathy, creativity, and complex reasoning. But if they become passive consumers of AI answers, they risk becoming "hollowed out". The goal of school shouldn't just be to get the right answer; it should be to prove that a human brain did the work to get there.
Why Does Everyone Sound the Same? (The Coded Gaze)
I watched a TedTalk on YouTube recently that asked the following question: Have you noticed people using the word "delve" a lot recently? Apparently, ChatGPT loves that word.
The report highlights a "feedback loop" where AI models (trained on the internet) spit out a generic, "average" style of writing, and then humans start copying it. This is creating a "Coded Gaze"—a distorted, flattened version of reality.
If our kids are constantly using AI to write or research, they risk losing their unique voice. They might start sounding like a corporate press release rather than a creative young person. We don't want a generation of children who all sound exactly the same because an algorithm decided that's what "good writing" looks like.
The "Linguistic Panopticon" (Or: Why Detectors are Unfair)
This is the part that really got my blood boiling. Schools are rushing to buy "AI detection" software to catch cheaters. But guess what? These tools are incredibly biased.
They measure "perplexity"—basically, how unpredictable a sentence is. If you write in complex, quirky sentences, you pass. If you write in simple, clear, standard English, the computer thinks you’re a robot.
Studies show these detectors are biased against non-native English speakers, flagging their work as AI-generated over 60% of the time, even when they wrote it themselves. Imagine your child working hard on an essay, only to be accused of cheating because their writing style is "too neat" or they are writing in a second language. It creates a hostile environment where kids are guilty until proven innocent.
The "Pay-to-Think" Gap
We talk a lot about the digital divide, but AI is creating a new one: the "Pay-to-Think" gap.
There is a huge difference between the free version of AI (like basic ChatGPT) and the paid versions (like GPT-4). The paid versions act like a private tutor—they can read whole books, analyse rubrics, and give detailed feedback. The free versions often just make things up (hallucinate).
If schools rely on these tools without providing them to everyone, we end up in a situation where wealthy students get a super-powered study buddy, and everyone else gets a confused intern. Grades shouldn't depend on whether you can afford a £20 monthly subscription.
Robo-Grading: The Industrialisation of Marking
If AI writing essays is bad, AI marking them might be worse. "Robo-grading" often looks for keywords and sentence length rather than actual logic or creativity.
Kids are smart. If they know a computer is marking their work, they won't write to communicate or argue a point; they'll write to trick the machine. They'll stuff their essays with long words and transition phrases just to get a high score.
Plus, assessment is supposed to be a relationship. When a teacher marks a piece of work, they are connecting with your child's mind. If we outsource that to a bot, we lose the "caring" part of education.
So, What Should Schools Be Doing?
The report argues we need to move from "policing to pedagogy". Instead of banning AI (which doesn't work) or trusting detectors (which are biased), schools need to change how they assess.
Here are three things we should be asking our schools to consider (in my personal preference - least desirable to most):
Bring back the Oral Exam: It’s old school, but having a chat with a student (a viva) is the best way to check if they actually know their stuff.
The "Two-Box" Method: Ask students to generate an AI answer (Box A) and then critique and improve it (Box B). This teaches them to be smarter than the robot.
Show Your Work: Just like in maths, students should show their chat logs and version history. It’s about assessing the process, not just the final essay.
The Future is a "Centaur"
The future isn't Human vs. AI. It's Human plus AI—a "Centaur". But for that to work, our kids need to be the ones holding the reins. They need to provide the ethics, the strategy, and the creativity.
It’s a brave new world, and it’s a bit scary. But by understanding these biases—the unfair detectors, the economic gaps, and the generic writing—we can better support our kids in navigating it.
Let me know what you think. Has your child’s school mentioned AI policies yet? Are you worried about "cognitive atrophy"? Drop a comment below!
Until next time, take care of yourself; check in on your friends; and remember: you can do this. You're awesome!
Comments
Post a Comment