The conditions of learning

For a long time, the conditions of learning barely shifted. Teachers held the knowledge, taught it directly and students showed understanding through recall, explanation or application. It worked in a world where information was harder to access, slower to obtain and largely reliable.

Then the internet changed the rules. Once information became freely available and instantly searchable, the challenge shifted from finding knowledge to making sense of it. Teaching had to move towards higher-order thinking, because simple recall just wasn’t enough anymore.

Generative AI has pushed that shift faster than anyone expected. It doesn’t just hand over information anymore; it produces fluent, confident answers — even when the thinking behind them is shaky, biased or just plain wrong. Students can now start with something already put together: a tidy summary, a ready-made structure, sometimes even a full explanation.

The cognitive work has shifted. More than ever, learning needs to be about assessing, verifying and improving what already exists.

Information is abundant, instant and of inconsistent quality. Anyone can generate something that sounds right in a few seconds. A polished answer no longer guarantees understanding; it shows that someone knows where to look. Meaning we all, adults and young people, need to question, compare, verify and explain how we know what we know. Teachers need these skills as much as learners.

Teaching thinking explicitly

Whether students use AI, Google, or neither, the need is the same: they must be taught how to think.

The modern challenge isn’t producing an answer; it’s judging one. A fluent response tells us very little about the thinking underneath it. What matters now is whether students can question what they see, test it against other sources, decide whether it holds up, and explain why they believe it does.

These habits don’t develop automatically. Thinking has to be taught on purpose. Teachers need to make these thinking processes explicit, talk them through, show what they look like, and give students regular chances to practise them. These skills cut across every subject, but they grow in different ways depending on the discipline. In practice, this means teachers need a clear sense of how students think, not just what they know.

Making learnings visible

A final polished product has never been a reliable sign of understanding and AI intensifies that reality. A convincing paragraph can be created in seconds.

Teachers need visibility of the reasoning behind the work: the choices, comparisons, explanations and justifications. That’s where understanding shows itself.

Well-designed tasks make this visible. When students explain how they know, compare alternatives or fix flawed reasoning, teachers get a clearer picture of learning than any final draft could offer.

Developing judgement skills

AI can write fluently, but fluency does not guarantee accuracy. Students and teachers must develop judgement skills: the ability to trust information cautiously, challenge it when necessary and verify it before accepting it.  

What this looks like in everyday teaching

If the conditions of learning change, the daily work of teaching naturally shifts with them. Teachers don’t lose their subject expertise; they use it differently. The emphasis shifts from giving information to helping students work with it, to reason with it, test it, and judge its reliability.

This shows up in four practical ways:

1. Tasks change
Tasks need to reveal understanding, not just produce an output. And that’s becoming harder to see, because AI can now generate the answer and a believable explanation, it can even fake bits of “reflection” or “confusion”.

This means one written task on its own no longer tells us enough. Learners need different ways to show their thinking throughout the process, not just at the end.

Evidence might show up in many small ways: a quick comparison, a short explanation, a check against another source, a presentation, a reference, or even notes from a discussion or group task. Working across different formats not only reveals what students really understand, but also helps them build wider skills in communication, research, collaboration and digital confidence.

2. Learning behaviours change

Learning behaviours have to change. Verification needs to become part of the everyday process, not something students do occasionally. This means getting used to checking claims, asking where an idea came from and comparing different versions of information.

Simple routines help. Sometimes it only takes something small, a quick evidence check, asking for another viewpoint, or a short “talk me through what you did”. Even a brief note about how they checked something can give teachers a much clearer sense of what’s really going on. Early on, it can help to offer students a simple prompt or a rough guide to get the habit forming.

3. Assessment changes

Teachers need more than a final product. A final piece of work on its own won’t tell us enough. The bits in between drafts, annotations, quick reasoning notes, and even short decision logs can all feed into assessment decisions. Assessment doesn’t need to happen at one fixed point; it can be gathered across a series of small checkpoints built into the task.

These checkpoints support metacognitive processes such as planning, mid-point monitoring, reviewing strengths and areas for development, and explaining changes or decisions. A quick note, a short comparison or a brief conversation can give teachers clear insight into how learners are thinking.

This spreads assessment across the process, shows students that the steps matter, and places meaningful value on how they arrived at their answer — not just on what they ended up with.

4. Acknowledging the role of AI

AI could have a place in modern learning, but that place needs to be defined. Students should be able to say when and how they used it, what it helped with and where it fell short. Making space for this kind of reflection, the good, the bad, and the indifferent, helps them understand the tool and take responsibility for their choices. Their reflections can sit alongside the other evidence, giving teachers a fuller picture of how they approached the task.

Used this way, AI becomes part of the process rather than a shortcut around it. Students learn to work with it transparently and thoughtfully. As always, this should be done in line with any national, school or institutional guidance on AI use, as expectations differ from place to place.

A system-wide responsibility

The role of the teacher is changing, not disappearing. In an AI-rich world, teachers remain the people who can challenge weak evidence, spot flawed reasoning and help learners make sense of complexity. Their expertise matters more, not less.

But teachers cannot do this alone. If students are to learn how to think, question and verify, teachers need the time, training and clarity to develop these skills themselves. This requires alignment across the system, DfE, curriculum bodies, awarding organisations, schools and government, so that what we teach, what we assess and what we value remain connected.

AI has changed the conditions of learning. It has not reduced the role of the teacher; it has revealed the human parts of teaching that matter most — judgement, understanding and the ability to help learners turn information into meaning.

[For transparency: I use AI as one tool in my work, but it does not think for me. The analysis and conclusions here are my own, grounded in experience, research and everyday work with schools.]

Closing note

This article raises a set of essential questions — not just for teachers, but for the whole system:

• How do we design tasks that genuinely reveal thinking?
• How should those tasks be assessed in a way that values the process as much as the product?
• How do we build a culture of verification and intellectual honesty?
• And whose role is it to prepare us for this shift?

I’ll explore these themes at Bett UK 2026 in my session The changing role of a teacher in the age of AI (Wednesday 21 January at 14:15–14:45 in the Teaching & Learning Theatre). I’ll share practical frameworks and real examples that schools can use straight away, helping them respond with more clarity and confidence.

Connect, learn, and lead with us

Follow us on LinkedIn to stay connected with our latest thinking on AI-age pedagogy, practical tools for schools, and upcoming live sessions and webinars.

For deeper dives, frameworks and insights, subscribe to our LinkedIn newsletter

You can register for your Bett ticket here.