I used to believe AI would save education.
Grades were dropping, but now students could make 1000+ flashcards w/ AI.
But after seeing numerous AI learning software, things are about to get much worse for students, education, and society.
Here's why:
My argument is simple.
First, here are 2 definitions for you:
1. What is learning software/ edtech?
Any AI tools that facilitate learning/ studying.
Example:
- AI writing app.
- AI tutoring app.
- Even AI that can create 100 practice questions in 60 seconds.
You get the point.
2. How does learning works?
Learning has 2 main parts: representing and processing (encoding + retrieving) information.
AI can help with the representation, but as a result, it may damage students' already poor processing skills.
This piece will explain why and what this implies in the bigger picture of education and society.
Let's step back... what does it mean to process information?
This is the most in-depth part, with 2 crucial points I want to state, so bear with me.
1. Processing occurs when you internalize the concepts and consider their connections and relevancy to build a web of knowledge.
With the Bloom Revised Taxonomy and the SOLO Taxonomy, we understand that higher-order thinking (analyzing, evaluating, and comparing concepts) creates a more connected network of ideas.
Which helps us skip the tedious memorization.
2. Through the Cognitive Load Theory, we understand a clear correlation:
Mental discomfort (struggle, confusion, frustration...) = more engagement = more learning.
3 types of cognitive loads combine to create the estimated load.
- Extraneous load: unnecessary distraction
- Intrinsic load: Complexity of information
- Germane Load: The process of learning
Just like progressive overload in the gym, the optimal zone (desirable difficulty) stretches you just enough.
(Please correct me if I'm wrong about any of this.)
From these theories and the general trend of students avoiding strategies requiring more effort even though they may increase learning efficiency (misinterpreted effort hypothesis)...
AI edtechs are making things worse.
3 reasons why:
1. Most EdTechs don’t understand how learning works.
AI reducing the extraneous load is good, but things tend to go wrong when it starts reducing the germane load.
When you skip higher-order thinking and rely on flashcards, rereading, linear notes, etc., you are now studying, not learning.
This results in companies falling into the trap of facilitating the misinterpreted effort hypothesis—taking manual tasks (flashcards, practice questions) and letting AI spam them out.
For example, an AI can create 100 practice questions in 60 seconds.
On the surface, that is attractive.
Still, when we remember that creating the questions is part of the learning, we realize that students skip both the thinking and struggle (you see the problem now?)
2. Conflicted interest on a business level:
I've looked at Instagram + TikTok lately, and the amount of misinformation pushed by AI EdTechs and influencers is insane.
But here's a quick thought experiment:
You see an AI edtech making crazy promises…
“why are they saying so?"
“Is it because they experimented and found a scientific breakthrough, or are they desperate to get user sign-ups?”
I think if the business has a clear intention to help students and their methods work, scaling is a win-win for everyone.
But the problem is the operational conflict.
When an edtech operates from the standpoint of adding value, like a business:
• Achieve the dream outcome
• Showcase the perceived likelihood of achievement
• Reduce time delay
• Minimize effort and sacrifice
We get something like this:
"We help students get straight As in under 15 minutes per day while spending less time and minimal effort. Here are 10 case studies of other students just like you."
That's a great offer.
But if we give students the reality that:
• A single trick won't save you.
• You can do it right and still fail.
• Struggling is important for learning.
• Studying is difficult and time-consuming.
It's very hard to sell anything.
I don't blame the edtechs. At the end, they have stakeholders to satisfy.
A few committed to helping students will seek consultation from learning experts. However, most will apply them partially or bend the advice because it could be more profitable.
3. AI will widen the gap between smart and curious students and average students.
Let's zoom out into the big picture — school.
Smart students become smarter, and less engaged students will become even worse.
Like human nature, we always choose the path of least resistance.
With education becoming meaningless, students might fall into the AI rabbit hole.
It might just start with:
"AI summarizing my notes couldn't do that much harm, could it? 🤔 "
" AI to research ideas for assignments 🙋 "
and soon enough... the AI has written everything. 😥
And this brings us to an important point:
Even if school admins try to compete with students on any front related to technology, they will F A I L.
For every new Turnitin update, hundreds of new AI writing tools will walk past it.
"But schools can ban AI and phones as a whole... right?"
• Hardware level: Yes, they can ban the tech.
• Software level: We are wired to solve problems. In this case, schoolwork or studying is the problem.
Once the students see an Instagram reel about a new AI that helps them avoid a problem, they will take the bait.
This is like their escape route from the pain (study). And once the intention is there, it's over.
For a much much better explanation, this video by @voxdotcom will do justice:
https://t.co/vr0a8IT2lB
Now let’s imagine this in the bigger picture of fast-forwarding 10 or 15 years.
Some of those high schoolers are now surgeons, lawyers, and accountants—people with power and responsibilities.
They finished 4+ years of schooling without basic knowledge of anything while being $100k+ in debt.
Another 10 or 15 years later, and some will be decision-makers who impact the lives of millions. At that point, some say we might already be like cyborgs and don't need basic facts, but that dependence will worsen us.
But now, back to the present...
I am NOT saying to ban AI from schools -- that's a losing game.
It's about changing how we teach and examine so that AI becomes the foundation and lets us focus on thinking & learning.
I won't dive into it here, as I feel the piece is already too long 😅
But here's one hypothetical solution.
Like Buffet would solve the debt ceiling problem in 5 minutes, people in power need more skin in the game.
AI EdTechs need accountability for their claims.
There should be an independent organization operated by learning science experts who would verify the claims of the EdTechs.
Under scrutiny, if the business can't submit a report on learning or research to back up its claim, it would go into the "wall of shame."
The site will be available online, and they will have a media team to call people out, making it a trusted source for the public.
Like the Community Notes on X, we need moderators who call out the BS.
At first, Edtech would get a warning, but if repeated, we would publish public articles and social media posts about them.
FYI, this will never happen. Maybe I should start this one day 😅
For now, I hope students, educators, and EdTech people find this post useful and it slightly changes the way they operate in their day-to-day lives.
AI can do great things for humanity, and the future is bright. If it's not, I will make it bright.
Share this piece with your friends if you want to spread the word.
If you want more content like this follow @LearningToan for more.
Toan.
I’m in love with this sentence:
“The degree to which a person can grow is directly proportional to the amount of truth he can accept about himself without running away.”
Nobody tells you this: Dopamine from information gathering is a dangerous drug. It’s the dopamine from reading, planning, or learning, but never doing. Stop looking for more information and start acting on the information you already have. Get your dopamine from action.
I don't understand why people don't just lock in on YouTube
My faceless YouTube channels make me $20,000+/month (long form only)
YouTube is not luck. It's a formula.
Let me send you a free course on exactly how to launch a Faceless YouTube Channel now & you could be making $11,000/month in January 2026.
To get: -
1. Follow (So I can DM you )
2. Like & Retweet (MUST)
3. Reply " YT " Must follow me to get DM.
I could show you how to;
- whip up hot-selling digital products with AI
-make people pay for them immediately they set eyes on them
- sell them to customers all over the world
- do all this without ever showing your face on the internet
RT and Comment "PDF"
@Codie_Sanchez Perfect tension. Faith without works is just wishing; work without faith is just grinding. Both hands on the wheel, both knees on the ground. That's how things move.