Stephen Downes

I participated in high school debates and learned a lot from it (not the least of which was that I was pretty good at argumentation). Here's an example of a high school debate from the City Club in Cleveland. It's pretty entertaining, though I still think the debaters focus too much on pumping out points of argument rather than crafting the whole for rhetorical force. But I quibble. This article is a fascinating look into that world. Stefan Bauschard offers a summary of this year's debate topic (hyperscale data center construction) and some general observations about the debate format, then gets into a deep dive into the topics and arguments raised across some 10,000 debates on the topic, resulting in a 2,100+ page scouting document. The article (happily) summarizes that. What can we learn from all this? A lot about the topic, obviously, including the sorts of considerations people bring to the fore around the issue. But also, we get a sense of the thinking around the issue, and especially the core points of contention.

I think we'll be seeing a lot more of this in the next year or two, though not always in manifesto form. This manifesto states, "The days of didactic instruction, where students are positioned as passive recipients of knowledge, should be long gone. We need to disrupt these approaches and reposition staff and students as active collaborators." Instead, students and educators should be thought of as co-creators of the learning experience. What does that mean? "Students and staff work together to decide on course content, such as course topics, learning activities, and assessment formats... Students via group work and peer learning activities bring their expertise and perspectives to the learning process... etc." There are more examples here.

Wendy Palmer teases out the tangled threads of a discussion about microcredentials. "If the problem is that someone needs to learn something quickly, then a short learning experience might be the answer. If the problem is that they already know it but can't demonstrate it, perhaps they need assessment or recognition of prior learning. If an employer doesn't understand what someone can do, perhaps we need better evidence and ways of communicating it. And if someone wants to build towards a larger qualification, we need pathways that allow smaller pieces of learning to connect and count. All of these might involve micro-credentials, but they don't necessarily require the same approach." This might all be too much to plan for in advance, especially if improvements to the education system are more incremental.

I keep an eye on all this, even if I don't report on it a lot (my focus is on learning technology, though personally I am very much in support of government-funded public education). Thomas Ultican writes and then documents, "In order to monetize education, deceptive slander of public schools became common place. People subjected to this slander still really liked their local schools but have become convinced schools in other areas are a disaster." Ultican suggests that "curriculum control was the goal" though of course the creation of a whole new private market was part of the motivation as well. Anyhow, the publications listed are mostly in my RSS reader (some are behind paywalls so I don't see them) and when I read or link to them I always keep this context in mind.

This document follows Jisc's updated learning analytics code of practice to add more on "the responsible use of artificial intelligence (AI) to enhance student engagement monitoring and support organisations to identify learners who may benefit from additional support." I think we should understand that the document refers to 'ethical' in the legal sense rather than the moral sense. That said, it contains a number of common-sense statements about AI in learning analytics, for example, "Where AI tools are provided by third parties, organisations remain responsible for how student data is used" and "'Understanding' a model means being confident in its behaviour through validation and monitoring, not necessarily being able to read its internal workings."

The author makes the very good point that AI isn't just one thing in education. AI in medicine, for example, is very different from AI for medicine. So Szymon Machajewski proposes a four-type model: AI for education, AI in education, AI of the profession, and AI literacy. Of these I think the third is a bit garbled, and the fourth seems superfluous. It seems weird to say "AI literacy is the only layer that can be taught without AI." But as a first take I think the model is useful - it separates two very distinct domains of AI: "Tasks that are formatting, layouts, under-the-hood, filing, and filtering (FLUFF) can be handed to AI because the work is not the learning. Work that is specific, persuasive, authentic, rigorous, and keen-insight (SPARK) is the learning, so AI should make the student think rather than think for the student."