Evening all,
I'm currently sittting on a train at Manchester Piccadilly station and I have half an hour until it moves, so why not clear my brain a little?
Precisely why I am sitting on a train in Manchester (surprisingly nice place despite its major football teams) is because I've left the AHE conference which I'm reliably informed is one of the pre-eminent assessment conferences around. It was a jam packed programme full of really interesting talks and ideas from caring and committed colleagues. I watched a talk this morning by Phill Dawson and Tom Corbyn from Deakin uni/CRADLE in sunny Melbourne. They were discussing generative AI and responses to it. It was a great talk that I enjoyed, but one part really resonated with me. They discussed (if I recall correctly) discursive and structural responses to gen AI, which nicely crystallised something that I'd felt throughout many of the talks, but also in many conversations and tweets and meetings and and emails since November 2022. What I thought in that moment is that so many of the well-meaning academics here, there and everywhere are lovingly focused on tending their garden. They are planting the seeds and pruning the petunias, and watering the seedlings with great care. That is, people are focused on the things within their locus of control or their speciality. However, many don't seem to look up very often.
I could be wrong, I often am, but while people notice there's a bad smell coming from the neighbour's yard (let's call this ChatGPT), if the camera draws back from this idyllic scene, their whole fucking city is ablaze.
Phill and Tom did a simple demo of how an online quiz could be answered with browser-based tools such as Copilot in about 7 nanoseconds, but the day before I watched a talk encouraging folks to embrace low stakes quizzes. WTF? These have been totally borked forever, but even if you started with Covid these are some of the most cheated assessments going. In the year of our Lord Twenty Twenty Four (I'm an atheist, roll with me here), why would anyone be heading toward that form of assessment with a straight face? Say it ain't so! But there it was. This is what I mean by tending the garden. By focusing on what works for you in your context, and not widening the lens, you may think the garden is lovely but basically you're ignoring the fire, which is student's ability (and willingness) to totally avoid any learning in your assessments.
Phill and Tom's talk encouraged people to embrace their inner Senor Speilbergo to consider broader structural issues. In my mind at least, this requires a massive culture shift.
I'm now somewhat convinced that as much as senior leaders may be seen as a problem (don't get me wrong here folks), practically speaking one of the things that must happen is that academics at the module/unit/subject level must start by relinquishing control. For example, in order for you all to agree on program/degree level learning outcomes, some of the freedom you currently enjoy (such as "I think I'll teach the students how to use chatgpt this semester") must go. If you have anything resembling programmatic assessment, this whimsical state of affairs is rightly impossible. It's not to say that outcomes can't change, but simply flipping a switch and teaching what you like shouldn't remain an option. On the flipside, you would be able to teach your elements of the program how you like. So there's choices.
But in my mind happily sitting in your garden and pretending everything around you isn't on fire isn't a professional or scholarly thing to do. It's daft and childish.
My last words are these. Whether you're fighting fires, battling Gozer the Gozerian,
or trying to stop a fire storm in higher education, crossing your streams is a good idea. No one of you will fix this, I assure you.
I'll probably post more thoughts as I make my way back to Sydney over the coming days, but thanks to everyone I met or chatted with at the conference, and thanks to the organising committee for a job well done.
Until next time.
KM
Generally speaking, I don't think academics will be convinced to act without 1) evidence that their own unit's assessments have been (not 'can be') compromised by GenAI, and 2) very clear instructions on what to do differently, with institutional/professional support.
Also, is a university's core function to produce graduates, or produce learning? (sure, ideally it would be both, but is it?)
Yep, there are no simple answers to a wicked systems problem. it requires a wholistic rethink of how and what is taught and assessed across the program where processes of learning must be evidenced to be able to see that it happened.