The Map Has Changed. Are We Still Teaching Students to Read the Old One?

The interface has always shaped the thinking

I realised something odd last week. I could not remember the last time I had typed a URL into a browser with any real intention.

Not for lesson planning. Not for finding an extract. Not for keeping up with what is happening in economics. At some point, gradually and then quite suddenly, I stopped navigating to information and started having conversations with it.

I doubt I am alone in that. I have heard teachers, developers and researchers all describing the same quiet shift. The browser tab is starting to feel secondary. The AI workspace is increasingly where the thinking happens.

But I do not think this is really a story about tools.

When the printing press arrived, it did not just make books cheaper. It changed what literacy meant. Reading became a more private, interior act. Arguments could be revisited, cross-referenced and disputed. The interface itself, the codex, the page, the index, shaped how knowledge was organised, shared and ultimately how people thought.

When the web arrived, we built a new map. Hyperlinks created a geography of information. You navigated to things. You developed instincts: this site is reliable, that one is not, here is where the primary source lives. We taught students how to read that map. How to evaluate a webpage, check authorship and compare sources. Those were valuable skills for a genuinely new information landscape.

Most of us were trained to teach pupils how to do exactly that.

What I am not sure we have fully registered is that the map has changed again.

From navigation to conversation

The shift is not just that AI is faster than Google. It is that the nature of the act is different.

Searching was navigational. You moved through a space, making judgements about territory. You found material, brought it back, and then made sense of it yourself.

Conversational retrieval is something else. You articulate what you are trying to understand, often imprecisely and often through a few false starts, and the interface meets you there. The synthesis happens within the exchange. Your thinking and the response develop together.

That has real cognitive implications. The skills that matter now are less about navigation and more about articulation: the ability to frame a question well, to push back on an answer, to notice when something feels off, and to recognise what you actually want to know.

In some ways, it is closer to a tutorial than a library visit.

What this might mean for our students

This is the point where it starts to feel important.

We have spent years teaching students to read the web critically. Check the source. Look for bias. Compare accounts. Triangulate. Those are still worthwhile habits. But they were designed for an interface that assumed students were receiving separate pieces of information and then evaluating and assembling them for themselves.

In a conversational interface, the challenge is different and, in some ways, harder.

The response arrives already synthesised. It feels coherent, fluent and often authoritative. There is no obvious author to scrutinise in the same way. The critical question is no longer just, "Is this website reliable?" It may be something more difficult: "Is this framing of the issue narrowing what I can see?"

That feels like a more sophisticated intellectual move. And I am not sure our current approach to information literacy fully prepares students for it.

I also think there is a generational angle here that matters.

I am old enough to feel the shift. I can see both the gains and the losses because I spent most of my life learning, reading and thinking in a different environment. I know what it feels like to search widely, to get lost down a trail of links, to stumble across something unexpected, and to read a whole article or chapter before arriving at a view.

For many young people, though, conversational AI may simply become the default way of encountering knowledge. That is a different kind of challenge. If this is the only map they really know, they may be less able to sense its limitations. They may not feel when a line of enquiry has started inside too narrow a frame and ended there too. They may not know what they do not know, because the system is so good at giving a plausible answer within the boundaries of the question asked.

If pupils are increasingly going to use AI to help them research, revise and think, then I suspect we need to teach some slightly different habits alongside the old ones:

• how to ask better questions
• how to test an AI's framing
• how to spot what has been left out
• when to step away from the chat window and go back to a full source

Something else worth sitting with

I do not want to be naively celebratory about this shift. There are things I genuinely miss about navigating information.

There was serendipity in it. You went looking for one thing and found another. The unexpected adjacent link. The source you were not searching for. The writer you had never heard of who changed how you thought about a topic. Conversational interfaces are very good at giving you what you asked for. They are less good at giving you what you did not know to ask.

There is also a question about depth. When you navigated to a source, you were sometimes encountering a sustained argument: a full essay, a proper article, a developed chapter. Conversational synthesis tends towards compression. Often that is exactly what you need. But not always.

And then there is attention. Slow, navigational reading cultivated certain habits of mind: patience, sustained focus, tolerance for ambiguity and a willingness to sit with complexity before reducing it. Whether we are losing something important there, or whether that worry is just nostalgia dressed up as intellectual seriousness, I honestly do not know yet.

The question I keep coming back to

If the interface through which most educated people encounter knowledge has fundamentally shifted, and I think it has, then what does that mean for what we teach?

Not just in terms of tools. In terms of the intellectual habits we want to cultivate. The kinds of questions we want students to ask. The relationship to uncertainty, synthesis and judgement we want them to develop.

The students in front of us now are going to spend their adult lives thinking through conversational AI in the way many of us spent ours thinking through the web. The map they will need to read fluently is the one forming now.

I do not have a neat answer to any of this. But I do think it is one of the more important questions in education at the moment, and one that still is not being discussed anything like enough.

What is your read on it?

Gary Roebuck is an A Level Economics teacher and founder of TeachEdge, an AI-powered marking and feedback platform built for UK secondary schools.