Why The Human Story Still Matters in Education

There is a gap opening between what a student can produce and what that student actually takes away from producing it. Daniela Kon Lieberberg, CEO of SIMA Academy, named this precisely in her essay "Chasing Algorithms: Why the Future of Education Depends on the Stories Machines Can't Tell," and it has stayed with me since I read it.

The gap is not new. Most schooling has run on produce-and-move-on for a century. What is new is that the production side has become effortless, and effortlessness turns out to be the problem.

The half of education we quietly dropped

The Latin root of the word carries a duality that modern schooling has largely forgotten.

Educare means to train and to mould. Call it the fundamentals: the times tables, the spelling, the sequence of events, the technique. It is real and it is necessary, and anyone who sneers at it has usually never had to teach a child to read.

Educere means to lead out, to draw forth, to illuminate. Call it the meaning behind the thing learned. It is the half that asks what any of this is for.

Factory-style mass education kept educare because it scales, and it can be measured, and it produces a number that can be put in a column. Educere does not scale, cannot be reliably scored, and takes an adult who knows the student. So it was allowed to fall away, and we have spent decades pretending the remaining half was the whole thing.

There are people in this world who do not particularly want a population that thinks. They want one that complies. A system that trains without ever leading out serves them very well, and the fact that this outcome was mostly an accident of scale does not make it less useful to them.

AI did not create this problem, it exposed it

We know what these tools can do. They are not coming, they are here, and pretending otherwise in a classroom is a form of malpractice at this point.

The honest framing is not whether to allow AI but where in the process it is permitted to sit.

Here is the rule I would give any teacher or parent, and it is simple enough to explain to a nine year old:

A tool may shape a student's thoughts. It may never create them.

Think of it as a potter's wheel. The wheel keeps the process moving. It does not decide what the pot is, what it is for, or what it should look like. The student's own hands shape the clay, fire it, glaze it and present it. The wheel is genuinely useful and genuinely neutral, right up until someone asks it to design the pot.

In practice that means the thinking happens first and it happens visibly. A student writes the claim, the outline, the shot list or the argument by hand before any tool is opened. What the tool may then do is what a spelling checker does, only more so: tidy, format, suggest a clearer sentence, catch a repetition. What it may not do is supply the position, choose the evidence, or decide what the piece is about.

The slippery slope begins the moment we hand over the keys and ask a machine to create something we should have given full shape to ourselves. That is true for a fourteen year old writing an essay and it is equally true for the adult teaching them.

Which brings me to a disclosure that belongs in a piece like this. I use Claude to give my own concepts blog-ready shape, working from detailed prompts of my own, so that ideas I have already worked out reach the page faster. I do not use it to have the ideas. If I am going to ask students to hold that line, I should be visibly holding it myself, and anyone who writes about AI in education without saying how they use it is asking for a trust they have not offered to earn.

What a film demands that a summary does not

I came to this from film before I came to it from teaching. My first degree was in Film Studies, I spent a stretch in Los Angeles working as a screenwriter, and I have served as a short-entry judge for SIMA. So I will admit a bias, and then argue that the bias is correct.

A well-made documentary cannot be consumed passively. It asks the viewer to weigh perspective, notice what the frame excludes, register where the information came from, catch the bias in the edit, and sit with the most uncomfortable question of all: whose voice is heard here, and whose is missing.

An automated summary asks none of that. It hands over a conclusion with the reasoning stripped out, which is precisely the part that was worth having.

And there is a second gear that most classrooms never reach. To deliver those goods to a viewer, the maker had to work through every one of those problems first. Which means putting a camera in a student's hands is not an art project bolted onto the timetable. It is the fastest route I know to critical judgment, because a student who has had to decide whose voice to include in three minutes of footage will never again watch anything without wondering who was left out.

How to actually run this

Go short. Ten to twenty minutes. Feature-length documentaries fail in classrooms because the discussion, which is the actual lesson, gets squeezed into the last four minutes.

Watch it twice. First time for the story, with no task at all. Second time with one specific question in hand. The second viewing is where everything happens, and skipping it is the most common mistake.

Never open with "what did you think?" It produces nothing but the answer the student believes you want. Ask instead: whose voice is heard and whose is silent? What does the filmmaker want you to feel here, and what did they do to make you feel it? What would this look like made by the other side?

Then make them make something. Three minutes. A phone is enough equipment. One interview, one location, one argument they are willing to defend. The judgment comes from the making, not from the watching.

Apply the potter's wheel rule to their production. The angle, the questions, the choice of subject, the argument: theirs, on paper, first. After that, use whatever tools help them finish it.

For a single student at home, this works unchanged. One documentary, one conversation, one three-minute film about something in your own town. It does not require a media studio or a department.

This is about to be measured

For anyone inclined to treat this as a soft enrichment activity, the OECD has already moved.

PISA 2029 will carry Media and Artificial Intelligence Literacy as its innovative domain, alongside reading as the focal domain. The Australian Council for Educational Research is leading implementation across more than 90 countries, the first draft framework has been published under the title "Navigating an Evolving Digital World," and results are expected in December 2031. It is described as the first international assessment of its kind, aimed at how well the next generation can think critically about what it sees online.

Fifteen year olds in 2029 are in primary school right now. Whatever a system intends to do about this, it is already late.

The point

The skills hardest to automate are the ones we spent a century declining to teach: reading a source with suspicion, judging a claim on its merits, noticing an absence, deciding what you actually believe and being able to say why.

They cannot be bolted onto a programme built for coverage, which is why most schools will struggle to add them. They have to be designed in from the first week, which is the approach The Promontory Academy takes and, more importantly, an approach available to anyone willing to rebuild rather than retrofit.

A student who can interrogate the world with calm confidence rather than hostility is not a product of better content delivery. That comes from being asked, repeatedly and from a young age, what they think and how they know.

Machines can produce almost anything now. They still cannot tell you what it means, or why you should care, or what it cost the person in the frame. That is the work. It always was.

The ideas in this post draw on the essay "Chasing Algorithms: Why the Future of Education Depends on the Stories Machines Can't Tell" by Daniela Kon Lieberberg, CEO of SIMA Academy.

PISA 2029 Media and Artificial Intelligence Literacy: https://www.oecd.org/en/about/projects/pisa-2029-media-and-artificial-intelligence-literacy.html