Is Generative AI in the Classroom, Inevitable?

I've been looking through a document that is available on the Hwb platform, called 'Generative artificial intelligence in education'. It's a document aimed at schools in Wales, that in their words focuses on "opportunities and considerations for schools and settings relating to the use of Gen AI". I personally found it quite an eye-opener and I think that at some point soon I will write in more detail about the content, to help me to get some things off my chest. However, what I thought came across clearly was that there was a sense of urgency, in fact the opening sentence states that, "Schools and settings need to be prepared...", but also what bothered me most was that it was written from the premise that generative AI is here. To me, the document reads that the assumption has already been made that we are going to be using it for teaching and learning, with what I believe to be an underlying context that it is inevitable and here to stay. I've recently seen the inevitability argument being used whenever there is a questioning or dissenting voice against generative AI, statements such as, "It's inevitable, it's going nowhere!", "Well it's here to stay!" Therefore, what I want to explore in this post is the idea that generative AI in schools is inevitable, that it is here to stay, and also explore the power that companies and governments have in shaping the inevitability narrative. 

What is inevitable? Well, I know that night will always follow day. That you and I will die one day. Both those statements are undeniable, unquestionable and as much as I would rather not think about my mortality, inevitable. These events will happen, but can we say the same things about generative AI? Well in my opinion, no. Firstly, I'm going to call generative AI a product and not a tool as I think it helps us to think slightly differently about what this technology actually is. In fact the general term AI or artificial intelligence is, as Dr Emily Bender describes it, simply a marketing term for a group of different technologies, one group of which is generative AI chatbots which are very much the focus in this particular document I've been looking at. They are products that have been created by a relatively small number of extremely wealthy technology companies. But, developing the Large Language Models (LLMs), which your generative AI chatbot interacts with, is an hugely expensive undertaking. For example, it cost an estimated $78 - $100 million to train GPT-4 and an estimated $192 million to train Google's Gemini Ultra 1.0. OpenAI, the poster boy for the generative AI 'revolution' with ChatGPT, is a hugely unprofitable company, with its partners carrying $96 billion in debt. Five AI related tech companies (Alphabet, Amazon, Meta, Microsoft, Nvidia and Oracle) owe a staggering $1.65 trillion in obligations that never appear on their balance sheet. The hyperscalers, including Amazon, Microsoft and Google will have sunk over $1.3 trillion into generative AI by the end of 2026 and have plans to spend a trillion dollars more next year. Why am I telling you this? Well hopefully you now get a sense that these companies urgently need their AI products to somehow make them a profit, because the evidence is pretty clear at the moment, that that is not the case. Consequently, I believe the inevitability narrative needs to be constructed by big tech, so that people "get on board" for fear of missing out, or even out fear of losing their jobs, derailing future careers or losing their lives! Whether it is through the AI boosters or the AI doomers, who are arguably doing the same thing in raising the profile of what AI might be capable of, but from different ends of the spectrum, these companies need you to use their products. After the obscene amount of investment into AI, a failure to become profitable for some could be catastrophic and the current AI bubble could well burst, as some are now predicting. The inevitably narrative is therefore driven by the people who have created it, arguably helped by a friendly, non-critical press. This is not new though, think back recently to the inevitability of the 'Metaverse' and how we would soon be wearing headsets and via our legless avatars we would be talking to friends, playing games, working and shopping in an online world. According to a Barron's article, McKinsey analysts in 2022 wrote, "The metaverse is simply too big to be ignored,” and that “It will have a major impact on our commercial and personal lives." That sounds like a familiar trope around the generative AI hype. Around $100 billion was lost by Meta on the Metaverse and I think it is fair to say that the Metaverse was a failure. However, the Metaverse still exists but no one is talking about the inevitability that we will all be living in Mark Zuckerberg's online world. From an education perspective, think back to the tech industry drive in 2011 to introduce computer science into the curriculum, arguably initiated by Google's Eric Schmidt basically saying that education in Britain is holding back the country's chances of success in the digital media economy. I definitely remember around that time the implication that this generation of learners would become the programmers of the future. Even the Prime Minister at the time, David Cameron, admitted that "we're not doing enough to teach the next generation of programmers", adding that there would be action on addressing this. This was the inevitable future and reform to the curriculum was urgently needed. I do find it a little ironic that on the back of governmental change to the ICT curriculum, in both England (2014) and Wales (2016 & 2022), which was initiated by tech companies, that the generation of programmers may well have entered into a now precarious career, one that AI has quite possibly had the greatest impact upon. Based on the history you would therefore think that the Welsh Government would be cautious when it comes to allowing big tech to shape the inevitability narrative for our schools and learners. We could end up changing our curriculum to suit current big tech priorities, only to see the educational goalposts change in 10 years as yet again a new digital product is inflicted upon the education community. In my opinion, big tech don't care about our learners, what they care about is making sure their AI investment is going to pay off for their shareholders and in having a pipeline of future users for their products. Another question, probably for another post, is how does big tech have such an influence on what we teach our learners in the first place? I can't thing of many (if any) instances where the curriculum (in particular, the DCF) is, in my opinion, 'Microsoft and Google shaped' in particular. Can you think of any other part of our Curriculum for Wales that has arguably been shaped by a company? Something that I have briefly mentioned in a previous post.

Coming back to the document I referred to at the start of this post. It appears to me that the way the Welsh Government has shaped the document (the narrative) was that the generative AI is going to be used by schools and therefore the only way forward was to make sure schools are generative AI ready and can adapt to these new products. However, before this document was published, I can't remember any democratic discussion with the education profession about how we should move forward with generative AI or even if it should have any place in the classroom or in teaching at all. But, maybe like the tech industry they wanted to move fast and break things. I guess difficult questions or dissenting voices could muddy the waters, so some people made the decision that this was going to be the starting point. Remember this was therefore a political decision, made by Welsh Labour who were in power at the time, and quite possibly influenced by the national UK Labour Government and the then Prime Minister, Keir Starmer, who actually said these very words:

AI and tech makes us more human. 

and also by the big tech companies lobbying them with the inevitability story. Just remember, it was people who made those decisions to embrace AI and generative AI in particular, and in fact a different government could well have made a different decision. This is not like the inevitability of day and night. If you were teaching in Norway, New York or Los Angeles for instance, where they have recently announced a near ban on the use of generative AI for learners aged between 6 and 13/14, generative AI is not an inevitability in the classrooms there, so why should it be an inevitability here here in Wales? We do have a choice.

Schools are told they must prepare students for an AI future.

Workers are told they must become AI literate.

Companies are told they must adopt AI or fall behind....

Everyone is racing.

Very few people seem willing to ask where exactly we are racing to. Or whether every institution needs to arrive there at the same time. (Link)

I can certainly understand why the majority of headteachers would not want to put their 'head above the parapet' and ask why are we doing this and where exactly we are racing to. Once the argument is framed as inevitable by the tech companies, by a friendly press and by governments, any opposition by a school leader could be seen as being irrational and a critic of generative AI for daring to ask questions about its utility. I'm not a school leader, so here is a list of some of the questions that I would be asking those who are encouraging you to use generative AI to 'help with workload' or exploring the potential learner benefits of using generative AI in the classroom:

Who decided this approach? Why? Who benefits from this? What are the risks of following this approach? What are we trying to accomplish? How does this serve education? How does it serve our learners? How does it serve my colleagues and I? How is this approach any different to what we currently do and will it produce better outcomes? Do those 'better outcomes' come at cost to teacher professionalism? What becomes possible because of this change and what might we lose in the process?

I'm going to argue that the last Welsh Government were clearly trying to make generative AI in education inevitable and schools will have to follow. To be honest, I'm really not clear on exactly where our new Plaid Cymru led government stands on generative AI, however, the document I've been referring to is still on the Hwb platform, it hasn't been removed and there is no note added to it to say that this was the policy of the previous government. Therefore, it must still be relevant for this new government too. I am also already aware of a primary schools having generative AI being referred to on their School Improvement Plans (SIPS) for this new school year (26 / 27), of cluster primary schools investigating the use of generative AI products with their learners and of working parties of headteachers exploring how generative AI can reduce teacher workload. If it's on SIPS and there are cluster working parties, then I'm sure generative AI has been highlighted as a priority by challenge advisors from local authorities / consortia who support schools, who will be acting on what the Welsh Government have told them to do with the schools in their areas. In other words, Estyn's recommendations from their AI "A New Era" report recommendations, are still being carried forth by this new government.

It is certainly clear from my reading that both the UK, the last Welsh government and the current Welsh government, therefore see the use of generative AI as some sort of panacea and an inevitability that will 'solve' the teacher workload issue. I have previously questioned this belief and I would suggest reading my posts on "AI - reducing workload and saving time?", "You did not gain an hour, you gained another task", and "We do have to be creative in terms of how we teach in the future..." to understand my concerns in more detail. However, where I believe individual practitioners will have more agency when it comes to generative AI, is in their pedagogic practice in the classroom. This is where I believe the inevitability narrative could certainly falter somewhat. As I mentioned above, one country (Norway) and two US school districts have already made the decision that their learners will not be using generative AI. The Mayor of New York, Zohran Mamdani saying:

The tech industry wants us to believe that A.I.-powered early education is not only inevitable, but necessary. We do not see it that way.

The Norwegian Government and US education districts have mandated that teachers will not be using generative AI with their learners for at least the next year or longer, until there has been a chance to evaluate the impacts of these products. Currently, I'm aware that our primary school users have no access to Copilot or Gemini via the Hwb platform, but do still have access to generative AI image generation through Adobe Express. However, my concern is that the Welsh Government might well mandate the use of generative AI through the Digital Competence Framework (DCF) in an update, that could possible come sooner than later. This was the underlying concern that resulted in my recent letter to the Education Minister. If the learners use of generative AI products were to be highlighted in the DCF, then I guess the expectation would be that both Copilot and Gemini would have to be opened up and that these products are then used across all curriculum areas, where appropriate, by the learner. As a slight aside on this, I was recently with a primary school teacher who told me that they wanted to use Copilot with their class and because they had no access through Hwb, decided to go 'outside' and the pupils used a free version of ChatGPT in class. I'm certain that this is something the Welsh Government wouldn't want to see replicated across our schools. So, if the DCF is updated then the expectations would be on teachers to provide generative AI opportunities for their learners. But, let's go back to the word 'appropriate'. What if a teacher, using their professional judgement, felt that it was actually inappropriate to use generative AI with their learners? What if they believed that by doing so they were actually harming learning or that by using these products offered nothing new that a non-tech solution could already do? What could happen, as has been reported in the many workplaces where AI has been mandated by bosses, is that some teachers may pay lip service to using generative AI with their learners. Addressing DCF descriptions of learning relating to generative AI, could then become just a tick box exercise for some teachers, whilst in reality it is rarely, if ever used, in a cross curricular context. The only time it maybe used would be if senior leaders are doing class observations or during an Estyn inspection. I personally believe that teachers have a strong professional identity and that could lead to resistance rather than genuine adoption.

Generative AI is not inevitable in our classrooms, even if it gets mandated by the government. It is a product, and like any other product being 'sold' to schools by a company or a government, teachers can decide whether it is suitable or not for our learners. Digital technologies have promised much and delivered little, with regards to improving learner standards, over at least the last 40+ years. Why should this latest tech hype cycle be any different to those others? Generative AI is just another 'shiny, new thing' aimed at seducing schools with its promises and potential, that are never realised. If it can be unequivocally proved that a particular generative AI led teaching and learning strategy will improve the learning standards of learners then of course we should take time to look at it. But until then, generative AI is certainly not necessary or inevitable, and whether mandated or not, it shouldn't be anywhere near our classrooms.

"The most obvious objection to my loathing of AI in the classroom and in childhood is this: AI is coming, and they need to learn how to use it or be left behind. But the tech of today looks nothing like it will in five or even two years’ time: it’s the equivalent of selling your child’s intellectual liberty in exchange for a penny-farthing bicycle. I asked a contact who is the president of a billion-dollar tech enterprise, what would you teach your sons? He said he wouldn’t teach his children to use AI, because it will prepare them for nothing. He would teach them the humanities. Teach them who we are, who we have been, what we can make. The FT’s Gillian Tett reported that a senior financier’s New York firm now seeks out humanities students, because “AI-native” Stem graduates are entering the job market with “alarmingly shallow ideas”." Katherine Rundell - The Guardian