Own Goals and the Howling Public

Evening all,

Yesterday Australian media published three (here, here and here) articles about another article, since retracted by the newspaper, which was itself a response to an earlier article- the "industrial-scale fraud" one. The retracted article was removed on the basis that it was written by/with AI. Now, the point of this post is not to go over the articles, and even less to talk about individuals involved. But there are a few pertinent lessons which I think we can take from this saga.

First, the thought has occurred to me, more and more often recently, that there's not a lot of sense in bemoaning the views of the media, or the public they speak to. I've often seen posts (especially on Linkedin) to the effect of "oh that article missed all the nuance", or "they're out to get us" (nasty media). Us being the education sector. Perhaps it's just my sense of it, but this particular refrain is simply not...fraining. Perhaps, and perhaps most likely, the fact that the public see us one way and we ourselves another is more a function of us being waste deep in all the stuff than any kind of nonsense ivory tower narrative out there. But that doesn't mean they are stupid (well, some of them are), and it doesn't mean what the public is concerned about has no substance. AI may be penetrating their conciousness far more slowly than it penetrated ours, and they don't see it as a research project, road to success or many of the varied things higher ed might see it as. However, they do seem to have cottoned on that LLMs are a case of mass avoidance of learning, and they believe we're not doing much about it.

Of course, in many, many, ways the public are wrong. LLMs and their impact on higher have caused an explosion of activity and attention on assessment that dwarfs that which came before re: concerns about cheating/avoidance of learning. However, they haven't forgotten the handwaving unis did on cheating for the past decade. And I get the feeling that they're looking at our current efforts with a very jaundiced eye. So when an academic uses copilot (copilot, really?) to draft an article, it confirms in their minds everything they already believe about us. Even though in our walls Nunya (nunya business whether someone writes with AI or not) might be the norm, it seems to me from afar that if we want to convince the public that we are authentically human, we should communicate with the public that way too. We don't need more own goals with a public that, truth be told, would pull out the pitchforks if this government even dreamed of giving higher ed more money.

Our problem with the public isn't simply that they may not understand what we are saying about assessment, it's also that they are not listening to us at all. I can only suggest that we start reading the room and start sending signals that will garner public support, rather than melting it.

Secondly, we need to acknowledge what's happening, and let that guide our immediate actions. The public belief that every student is cheating (which is not the case) is not diminished at all by higher ed folks saying it's a "small minority", a phrase i heard AGAIN today, which is also not the case. Blind Freddy can see that students are passing subjects without learning much at all, when they want to. Pretending that's not the case, and focusing on the "chronic problem" I mentioned last week is fuelling the fire of public anger. A friend of the blog insisted that we can't get to secure assessment unless we got to the educational researcher's promised land (I'm paraphrasing), wherever thay may be. He and I want the same thing, good teaching, quality learning and secure/well designed programs of study, but what I said to him was words to the effect of "if you imagine we're going to get more funding for gold star assessment without throwing the public a bone, you're mad." That is, being prepared to fail students who can't demonstrate their learning without the crutch of an LLM (or another person for that matter).

I agree completely that exams are an imperfect form of assessment, but they're less labour intensive, and therefore doable right now. So are interactive orals, but the time needs to be taken to explain them to the public, and otherwise explain how we are assuring learning. So I ask why we make no attempt to make them better, rather than dreaming up assessments we already know we can't afford, at least not here and now? For better or worse, invigilated exams are a standard the public recognises and, I'd suggest, respects. Most people have sat one. They know it's difficult. They know they're designed to prevent people passing who shouldn't. And rather than going on flights of rhetorical fancy, it might be helpful if we set about regaining public trust, "social license", with less words, more actions. Less talk, more rock.

Until next time,

KM