Forget AI cheating. Cognitive offloading is more of a worry

The biggest threat posed to education by generative AI is not academic integrity, it is cognitive offloading.

In FE and the wider education sector, concerns around AI tend to largely focus on how it is providing learners with the tools to “cheat”. This is an argument which ignores the reality that education and modern work practices now exist on a continuum of AI partnership. It isn’t the case that all AI use is bad and no AI use is good, but clearly the process of AI replacing learning is something that we need to safeguard against.

The optimal way to ensure that learners cannot use AI to replace learning is through assessment redesign. Even ahead of widespread curriculum reform, assessments at Middlesbrough College are being enhanced to focus on process over product, with an increasing variety of authentic assessment throughout all programmes.

When evaluating assessment methods prior to the academic year, curriculum staff would ask themselves the question, “could a learner use an AI tool to pass this assessment with no learning taking place”? If the answer was yes, that assessment would be redesigned. We are committed to ensuring our learners leave us with the knowledge, skills and behaviours to thrive in their chosen career path, we are not teaching them how to be meat proxies.

Cognitive offloading is not a new concept, but it has been supercharged by AI. It is now possible to outsource the entire process of generating ideas through to creating solutions, with significant negative implications. The result of chronic cognitive offloading is that learners find it harder to concentrate, harder to formulate arguments, harder to think critically and in short harder to learn. This of course not only affects them whilst they are with us, but in their next steps in education and the workplace.

Taken one step further, cognitive offloading has the potential to lead to intellectual obesity in our learners. Still only a novel hypothesis, the gist of intellectual obesity is as follows. Imagine, you spend 90 minutes consuming easily digestible, dopamine-inducing mental “junk food” in the form of TikTok videos or Instagram reels for example. Following that you sit down and try to design an engaging, well-researched presentation. You will probably find it difficult to focus as it takes time for your brain to reset following such a flood of stimulation. But don’t worry, AI can help take away the discomfort of looking at a blank screen trying to think about how you will approach the problem, it will do the cognitive ‘exercise’ for you. By failing to do that mental ‘rep’, the next time this situation occurs, it is even harder to do your own thinking.

We are striving to tackle AI-induced cognitive offloading with our learners at Middlesbrough College from the beginning of their journey with us and throughout their programme. As I write this, we are in the midst of delivering face-to-face digital inductions to all of our first-year full-time learners. During these inductions, which take place over the first two weeks of term, not only are learners provided with guidance on the essential digital skills that employers are seeking, we also cover the effective and ethical use of AI.

The key message we relay is how AI can be used effectively as a study partner, but not as a means to replace learning. Checking for understanding, revision aids, breaking down tasks into actionable steps and preparing for interviews are just a handful of the beneficial uses of AI that we promote to our learners, whilst steering them away from using it to provide answers and results which circumvent the learning process. In addition to this, our Thrive tutorial AI content, which is delivered by personal tutors throughout the year, has been redesigned around these same principles and is delivered in a way which is contextualised to the level and individual needs of learners.

It is not only incumbent upon us as FE professionals to instil the principles of effective and ethical use of AI in our learners to protect their cognitive health, it is essential that we practise these behaviours ourselves. Let us spare our colleagues from the bane of AI workslop and remember that in the age of AI, authenticity has more currency than ever.