Stop Talking About AI Potential. Provide The Evidence First

I don't know how on earth I missed this from the UK government. Back in January 2025, they announced, 'mainlining AI into in the veins of the UK'. Oh, good grief.  Keir Starmer saying that 

"Artificial Intelligence will drive incredible change in our country. From teachers personalising lessons, to supporting small businesses with their record-keeping, to speeding up planning applications, it has the potential to transform the lives of working people." 

I wrote in a recent post about a response I gave to a teacher asking me whether I believed AI was just a gimmick or novelty. In the post I highlighted that it was clear that both the UK and Welsh governments had invested too much time and effort in bringing in AI across all UK sectors, and in education we are going to see it heavily pushed as a panacea to help teacher workload and that I'm pretty sure we are about to see learners encouraged to use it in schools through an update to the Digital Competence Framework (DCF). So, based on that quote, I was very much justified in saying that generative AI is going to be embedded into our schools and into the curriculum, impacting all staff and learners.

Going back to Starmer's quote, the part that really annoys me is the word 'potential'. I see this so often when people refer to AI and generative AI in particular. That it has the potential to help teachers with workload or that it has the potential to improve learning. No proof that it will, but it might. Almost everything to do with AI tends to be prefixed with potential. If you are a business, it's up to you whether you want to heavily invest in AI technologies with the potential that it might provide you business with a good ROI (which is questionable that it actually does). But the bottom line is that's up to the business to decide whether it makes good business sense to them. If a business loses money, well that was just a poor business decision on their part. However, in education, I think potential positive impacts on learners is just not good enough and frankly wrong. Pupils are not guinea pigs, they are not an experiment. We could quite easily spend a couple of years with teachers and learners 'playing about' with AI tools only to find out there has been little impact on learner standards. 

When EdTech fails, the cost isn't measured in dollars - it's measured in the learning and development of our children. Dr. Horvath, 'The Digital Delusion' (2026), page 23-24

Forget potential, my job as a teacher is to know what proven teaching and learning strategies work with the learners I have before me. Look at the work of the Education Endowment Foundation (EEF) who specifically look at the evidence of teaching and learning strategies and compare the evidence, costs of implementing and the impact on learners. I really don't think it is the job of a classroom teacher to try and find out what these tools do for learners, especially learners in a primary school sector, unless it is part of some rigorous, academic study. Concentrate on what we know already works with young learners, not on potential! Provide us with the results of what studies have found. If we have any positive evidence in learner standards, then I'll be interested, but not before.