LO
Education and teacher training
Trains you to take something you understand and get it into someone else's head — by working out what they already believe, where they will go wrong, and what evidence would show they have it.
Take a homework task from your own school years, give it to a model, and look at the output. Then rewrite the task so that the same model's answer would no longer earn full marks. That rewrite is the single most transferable thing you can practise this term, and schools are currently short of people who can do it.
A single number for a whole major would hide the thing that matters: this degree trains several separate competencies, and they are not all moving in the same direction. Automation acts on tasks, so any assessment lives on the occupation pages below — not here.
Written for undergraduate education and teacher-training programmes in mainland China and Singapore, including subject-specific teaching tracks. Early-childhood, special education and educational-technology programmes diverge and are not covered by this assessment.
What this degree actually trains#
Not the course list — the competencies underneath it, and whether each one is worth more or less than it was.
Working out what someone has actually misunderstood
Worth more than beforeLooking at a wrong answer and telling which wrong model produced it — not marking it wrong, but knowing which of four plausible confusions this particular student is carrying.
A model can generate an explanation instantly and can grade against a rubric, but it does not sit with one child and notice that the arithmetic is fine and the reading of the question is not. That diagnosis comes from watching a specific person work, and it is the part of teaching that generative tools have made more valuable rather than less, because explanations are now free and the scarce thing is knowing which explanation this person needs.
Holding a room of people who did not choose to be there
Worth more than beforeKeeping thirty people who would rather be elsewhere engaged enough to learn, and handling the moment when one of them decides to test whether you mean it.
This is the part of the job that has no software substitute and that outsiders consistently underestimate. It is also the main reason people leave the profession, which means it is the capability that decides whether the rest of the degree ever gets used.
Designing an assessment that measures the thing
Worth more than beforeBuilding a task where doing well requires the understanding you care about, and where it is visible whether the work is the student's.
Every take-home assessment written before 2023 now has a free way around it, and every school has discovered this at once. Rewriting assessment so that it still measures something is live, unglamorous work that has to be done by someone who knows the subject and the students, and it is currently one of the clearest places a new teacher can be immediately useful.
Preparing the lesson material
Worth less on its ownProducing the slides, worksheets, exercise sets and marking schemes that a lesson runs on.
Drafting these is the task generative tools do most convincingly, and it is a large share of the unpaid evening hours the job has always carried. The judgement about whether a worksheet suits this class is not automated; the typing of it substantially is, and treating the typing as the skill is the mistake to avoid.
Explaining a child's situation to an adult who is worried
HoldingTelling a parent something they do not want to hear, in a way that leaves them able to act on it and still willing to talk to you next term.
Neither newly scarce nor displaced: it was always the part of the job least visible from outside, and nothing about automation has changed who has to sit in that conversation.
Where it can lead#
4 directions, never one. Each is written out in full on its own page — what your training reuses, what graduates typically lack, the entry reality, and one thing to test this term.
What to add outside the classroom#
This is about what graduates commonly lack in practice — not a claim that your school failed to teach it.
Hours in front of real students, as early and as often as possible. Everything else on this list is a distant second, because this is the capability that decides whether you stay — and it is the one no coursework can simulate.
A worked view of what generative tools do to your subject specifically. Not a general opinion about AI in education — the concrete answer to what happens when a student in your subject has one, which differs enormously between mathematics and composition.
Some familiarity with what your education system actually requires of teachers on this subject, since several markets now mandate AI-related teaching content and that lands on you as work rather than as a policy document.
This term#
One or two actions, each producing something you can show someone. Not a reading list.
Take a homework task from your own school years, give it to a model, and look at the output. Then rewrite the task so that the same model's answer would no longer earn full marks. That rewrite is the single most transferable thing you can practise this term, and schools are currently short of people who can do it.
Ask a teacher who has been in a classroom for more than five years what share of their week is administrative rather than teaching. Their number is the one to plan around, and it is rarely the one prospective teachers assume.
Common questions#
Not on the strength of a score, because this site will not give one to a degree. The relevant evidence is narrower and more useful: on the teaching page here, drafting lesson material and noticing who did not get it are assessed as augmenting, where tools help but the teacher still does the work, while holding a class is assessed as human-led. If those are the parts you want to spend a career on, the degree still leads there. If what appealed to you was preparing beautiful materials, that is the part under most pressure, and it is worth knowing now.
The records on this site point the other way, and in a specific direction rather than a reassuring one. China's education ministry issued guidance in 2025 requiring AI general education across primary and secondary schools and assigning teachers a strengthened guidance role — that is policy adding a duty to the job, not removing the job. A New York district's attempt to bring in a humanoid teaching robot was paused after objections from the state, the union and parents, and what stopped it was procurement and student-data agreements rather than the technology. Both records are on this site with their sources.
Rewriting assessment so it still measures something is currently among the most useful things a new teacher can be good at, and most schools are doing it in a hurry. The practical move is to shift what is graded toward work you can see happening — in class, out loud, or in a form where the reasoning has to be shown rather than the answer produced. That is not a policing strategy; it is a design one, and it is a skill you can start practising before you have a class of your own.
That framing does not survive contact with the evidence, and the honest answer is that the risk in this profession has never mainly been automation. People leave teaching over workload, behaviour and pay far more than over technology, and nothing on this site changes that. What the evidence does say is narrower: one part of the week — drafting material — is being assisted by tools rather than taken over, and one part — the policy duty to teach AI content — is being added.
Real hours with real students, and a clear sense of which age group and which subject you can sustain — this profession's binding constraint is attrition, and the people who last usually found that out early. Beyond that: one assessment you redesigned so a model could not simply produce the answer, kept with both versions. It demonstrates subject knowledge, an understanding of how students actually work, and that you have engaged with the thing every school is currently worried about.
Method#
Assessments live on tasks, not on majors. Follow any direction above to its occupation page to see which tasks are changing, how strong the evidence is, and what it does not yet show.