The Intelligence Age
We started with the Information Age.
The Internet levelled the playing field for information worldwide.
Suddenly, no one had to travel to a foreign country for education.
The Internet brought high-quality education everywhere.
MIT OpenCourseWare is a good example, as is Coursera, edX, and Udemy.
But now we live in another, entirely new age.
Some call it the Intelligence Age, and I like that term, so I will use it.
LLMs like ChatGPT, Claude, Perplexity, and Gemini bring high-level intelligence and analytical skills to anyone who can write an English sentence asking for what they want.
Today, the only limit to the information you can learn is your work ethic, discipline, and imagination.
All three factors are important, but the greatest among them is imagination.
So now, educators have trouble.
Students are using LLMs for all homework, assignments, and problems.
AI detection tools and banning LLMs are not the answer.
There is another much better option.
Before I go any further, to give credit where credit is due -
This entire article was inspired by this single video:
https://www.instagram.com/reel/DHlPmCGvU9t/?embedable=true
How to Use LLMs for Education Effectively
You cannot ban a student at home from using ChatGPT/Claude/Gemini/Perplexity any more than you can ban accountants from using calculators.
In the Intelligence Age, LLMs are not to be fought - they must be adopted and embraced.
How?
Revise existing education systems.
Let students use LLMs at home.
From middle school, encourage LLM use at home.
This is how they will work anyway, so let them master prompt engineering and vibe coding at a young age.
Digression:
I am not a fan of vibe coding.
But children who are always supervised by adults on the Internet are rock stars with vibe coding tools.
Let the teens start coding at a young age, in a gamified fashion.
But-
Do not give teens mobiles or tablets, and never let them access the internet without supervision.
Your future self will thank you.
Supervise all Internet usage.
There are far worse things on the Internet than pornography.
I do not even like talking about it.
But we, as I said, digress.
Back to the article!
Adopt LLMs in every country - every education system - every course - almost without exception.
The Methodology That Flips The Process The Right Side Up
This is what hit me so hard in the video.
Allow children to use LLMs at home.
Allow college students to use LLMs for homework.
But do not grade that work for their marks.
It is not theirs to receive credit for it.
When they arrive in class the next day:
Ask them to critique/modify/add to their work - in class - live - without LLMs.
Do not spring this as a surprise - tell every student that this is what will be done.
Use that output for grading.
You might say, one teacher cannot individually grade 120 students in even 6 hours.
Don’t do that.
No need.
Let the LLM grade the in-class submission for you.
Make the evaluation transparent and live.
Bonus marks if the student can spot a hallucination in the grading!
This is the Future
LLMs are not perfect, and in their current form, they may never be perfect.
But the solution to progress is not banning progress.
The solution to progress is to embrace it.
Banning LLMs is like banning calculators.
It helps in learning the basics - but most of the work in the future will be done by LLMs.
While basics are important, let the students learn basic prompt engineering from a very young age.
Let them prepare and learn with LLMs.
You will find some outliers going at supersonic speeds.
Encourage them.
Others will find LLMs enthusiastic and infinitely patient teachers.
People who are smart today -
I predict that they will be eclipsed by adults who have been prompting LLMs since the age of seven.
They will be the AI superstars who will eclipse everything the professionals of today are capable of.
What About UG and PG College Degrees?
Same drill.
Let them use AI in their dorms, away from class.
Tell them in advance that the grading will be done on classroom knowledge without access to AI.
Three, four, or five years of this, and students will be better than they ever were in the past!
Let the LLM evaluate all human-submitted answers instantly, and let students access their grading results from LLMs.
Give them bonus marks for every hallucination they spot!
Caveat: for confirming hallucinations, use a different LLM.
I recommend Perplexity or Claude for this use case.
Let them use every AI tool to prepare.
- NotebookLM
- Learning Mode Across LLMs
- Google Gemini with 1000 free requests a day
- Claude for spotting hallucinations
- Perplexity for citations
- Deep Research across LLMs
- LLM summarization of entire books
But the final exam?
No AI tools allowed.
But tell the students the entire plan in advance.
Well in advance.
Why no AI tools for exams or tests?
Students will be using them for work!
Because we risk dumbing down our students.
A student who never develops critical thinking, the ability to formulate theories and ideas, as well as cogent writing skills will be useless to any employer.
A student who is not resilient, able to handle pressure, and think independently without an LLM will be useless when a project in production goes down or when detailed independent creative thinking is required in critical situations.
I would never hire any student who could not work without an LLM.
And those are the students that the current education system is building.
The administration says - No LLMs allowed.
Students use LLMs for everything.
Especially remote learners.
AI detectors are not very reliable and will hurt those for whom English is a second language.
And What About Remote Learners?
I have bad news and good news for you.
With LLMs, the old norms of remote learning are officially void.
Originality.ai has made some strides in AI detection tools recently.
But until AI detectors can detect human paraphrased AI texts -
No remote learner can be verified reliably.
So, is this the end of remote learning?
Hardly.
Remote Learners need an additional step up.
They must create a project portfolio (partially with LLMs) that is so outstanding and original that it could never have been written by AI alone or coded by AI alone or conceptualized by AI alone.
This is not the end of remote learning.
This is the step for which LLMs level up remote learning to an entirely new ball game.
This is the chance for remote learners to shine.
LLMs can still do the grading, but with human oversight, preferably at least three examiners for the online viva voce.
The standards must be set to the highest of levels.
And to repeat, bonus marks for spotting hallucinations during LLM grading!
Remote learning without human oversight is, with the advent of LLMs, officially not acceptable as a qualification.
However, using those skills to create stunning projects, real-life deployed mobile apps, and incredible writing at the highest level, all available on an online portfolio site, will be all the qualifications an employer needs to see.
The standards need to be at the highest levels.
Dumbing down one student today is dumbing down an entire society tomorrow.
And that is what we risk if we ban LLMs for education.
Because students will use them anyway.
And writers who are not skilled in English as a native language will have their original human work classified as AI.
The future is not banning LLMs, but embracing them.
Kudos for the video, @ Sinead Bovell
This is the future of education, and it must be adopted worldwide.
Conclusion
What about PhD Scholars and Research Scientists, Thomas?
That’s a very carefully nuanced topic that deserves an article of its own.
And I will give you one.
Without AI.
Finally:
Parents: keep an eye on your children with the Internet at all times.
No tablets or smartphones until 16+ years of age.
The risks outweigh the benefits.
Watch this video if you need convincing:
https://www.instagram.com/reel/DPQfBjpiNTX/?embedable=true
But if your children can learn to use LLMs early on in their life - nothing like it.
Welcome to the Intelligence Age.
The Stone Age → The Iron Age → the Industrial Age → The Information Age → The Intelligence Age.
Is the SuperIntelligence Age next?
Your guess is as good as mine.
But - despite all the difficulties and dangers - I strongly believe that there’s never been a better time to be alive.
This is the time of constant change, which is also the time of constant opportunity.
Our education system, worldwide, needs to be revamped.
Let’s make it happen.
No AI was used in the writing of this article.
All Images in this article are AI-generated by the author using NightCafe Studio Creator, available at this link: https://creator.nightcafe.studio/explore