Will AI Replace Lecturers?

OpenAI has just switched on its new Study Mode for ChatGPT. Learners can now drop a syllabus, ask for bite-sized explanations, and get instant practice questions while they sit on the bus. If that is already happening on their phones, how long can colleges and ITPs afford to carry on as if nothing has changed?

FE is Locked in a Perfect Storm

Teaching vacancies sit at 3.9%, retention rates have fallen off a cliff, and real terms pay is stuck in reverse. The Institute for Fiscal Studies confirms that college teacher pay has crashed by 18% in real terms since 2010, worse than the 13% fall for school teachers. A quarter of college teachers leave within just one year, and almost half are gone within three years. Less than a quarter of those who joined a decade ago remain in post today, compared to over 60% of school teachers.

The pay gap between colleges and schools has widened from 14% in 2010 to 21% today, with college teachers now earning £6,000-11,000 less than their school counterparts. It’s gotten so desperate that 172 college leaders have written jointly to the Chancellor demanding action, this isn’t isolated whinging, it’s system-wide collapse.

Meanwhile, colleges need an extra 800 maths teachers and 400 extra English teachers just to meet new GCSE requirements. At current pay levels, that’s pure fantasy. Schools paying £10,000+ more can’t find maths teachers, what hope do we have?

Add the national skills crunch, 251,500 extra construction workers needed by 2028 (with the industry losing 210,000 workers annually while only recruiting 200,000), 111,000 unfilled NHS posts, and a digital skills shortage that could cost £27.6 billion by 2030, and we’re asking understaffed colleges and Training Providers to solve a national crisis with one hand tied behind their backs. Even the government admits that without better pay and workforce strategy, their own Skills England ambitions “are in danger of failing.“

People First, Robots Second

Think of your best bricklaying tutor. Years on site, stories that grab a room, the instinct to spot a sloppy mortar joint before it collapses, along with actual contacts at local firms who will actually take your apprentices. No chatbot can give you that. Practical skills need live eyes on angle grinders and welding torches, plus the pastoral nod that says, “I know you are struggling, let us talk after class.” Anyone who claims that something like a Study Mode can replace that must have an agenda.

Where the Machines Should Step in

What the machines can do is stop us drowning in admin. UCU workload research tells us that 41% of staff describe their workload as unmanageable, with lecturers clocking an extra 13 unpaid hours every single week. Bolton College’s Ada chatbot has already helped more than 70,000 learner queries, freeing staff for deeper conversations. A Jisc TeacherMatic pilot found that nearly 400 lecturers across eight colleges saved just over 2 hours a week on planning and resources, with almost 70% reporting a better learner experience. In KS3 science, an Education Endowment Foundation trial showed that ChatGPT assisted preparation cut lesson planning time by nearly a third with no dip in quality. Even in high-risk trades, virtual reality rigs paired with AI feedback let trainees practise scaffolding or live circuit work repeatedly without danger.

How to Jump in Without Blowing the Budget

How do you actually get started without waiting? Start with the licences you already pay for, usually Microsoft or Google, and nominate a couple of curious staff as AI champions. Ofsted’s latest study found that colleges are already successfully using AI to reduce workload and support learners, proving that FE’s mature learners and work-readiness focus make it perfectly positioned for AI adoption. Give your champions time to explore, experiment, and then show the rest of us what works, this might be an afternoon to start, but realistically could take weeks or months since it’s a complex process that’s best not to rush. Use the free Jisc AI maturity toolkit to map your next steps and understand where your college sits on the AI journey. Then run a lunch-and-learn session showing how other colleges are making this work.

Do not ignore equity. If learners are scraping data on pay-as-you-go mobiles, an AI-enriched course widens the gap. Wi-Fi, device loans, evening access rooms, whatever it takes, because the widening digital divide is only getting bigger.

The Cost of Sitting Still

Replacing a burned out lecturer can cost £30,000+ once you add recruitment, cover and induction. Every learner who drops out because support was not there drags away funding.

Employers vote with their feet! 71% already see skills shortages as a threat to their competitiveness. They will choose colleges that choose to utilise things like Study Mode to help them get better attendance, faster completion and staff who still smile in week twelve.

No, AI Will NOT Take your Job

It will polish it, extend it, and make the human parts matter even more. Colleges that move now will keep staff, employers, and funding. Those who wait for the “perfect” policy paper will watch their learners swipe open ChatGPT Study Mode anyway.

The time for another committee paper has gone. Pick a pilot group this term, carve out the rules and let the machines grind through the drudge so people can get back to teaching and providing more in person time to learners and employers.

By Danny O’Meara, Operations Manager, FE News