Stephen Downes

Fresh off the press. Open educational practices (OEP) include open educational resources (OER), open pedagogy, "where students are creators of learning materials and not just consumers", open access publishing, and open research and open data. But (depending on who you ask) OEP can be much more than that, entailing as well commitments to such things as global justice, decolonization, diversity, equity and inclusion (addressed directly by Jade Hupé in section 1.5 and Roberta Campbell-Chudoba and Heather M. Ross in 1.6). The chapters are short, often frustratingly so; this brevity often comes at the expense of clarity (a slightly longer section on types of open access, for example, could properly define each of the different terms instead of just mentioning them in passing). Section 6.3 brushes by the First Nations OCAP principles, the CARE principles, and the Tri-Council Policy Statement without really telling the reader why these policies are in place. Positioning them as 'exceptions' feels wrong to me.

This article (advertorial) announces Google's partnership with Khan Academy to integrate its AI, Gemini, with Khan's reading and writing coach. Khan has been moving in this direction since forever, so an announcement is surprising only because it's with Google and not Microsoft. But even that is only the mildest of surprises. 

I have nothing like the deep fascination Jim Groom has for film and video culture but I do appreciate as Groom does the work Mike Caulfield is doing to understand what AI is doing when it creates ("a great recent example," says Groom, "is his "book chapter" on My Cousin Vinny (and) he even built a film-aware fact-checking tool). That's the genre Groom is actually working in here as he has ChatGPT take a look at an essay he wrote in 1999, analyze it for originality ("Verdict! Bloggable!") and produce a post that would be relevant in 2026. The post stands up and I'm sure film-Groom's audience would be interested. For my part, tech-Groom's post about the post is also a pretty good read, one in which I am interested. And to return to his main subtheme, the question of assimilation, well, that is a theme that has been very well worked over in film, literature, and reality. Sure, the discussion is topical ("the ways in which our writing and thinking is being mapped, grafted, and commodified in a tool like ChatGPT is some next-level horror," writes Groom) but compared to what might be it's also hyperbolic.

I have mixed feelings about this article. It recommends 'early career researchers' reach out to media to gain the traction needed to attract the 'lightning strike' in the form of a big grant or scholarship. My feelings are mixed because while I definitely encourage engagement with multiple media, I don't see career advancement as the primary purpose of it. At the risk of overgeneralizing, I think there are two types of researcher: careerists, and scholars. Careerists are worried about publicity for their own advancement. Scholars are interested in exploring and disseminating ideas. In my world the careerists are disposable, while the scholars are the people worth following and listening to. It is a problem if grants and scholarships are based on publicity rather than scholarship, and not something people should be encouraging.

The core tenets of using video in learning haven't changed since 2013, says David Hopkins. "Then, as now, video works best when it is framed by purpose, it draws attention to key ideas, it is embedded in activity, (and) it leads somewhere (discussion, reflection, application)." What has changed is that purpose is now critical. "Students need to know why they are watching, what to listen for, (and) what they will be expected to do with it afterwards." He adds that AI isn't ready to design learning with and around such videos. "If a task can be completed by passively watching or automatically summarising, it probably wasn't a learning task in the first place." To me this misses the point a bit. The utility of video is that it shows what can't be easily summarized. It then (if well designed) invites and enables physical replication of the task or process being demonstrated - simple reflection or discussion don't really count as 'activities' in this scenario. And yeah - AI can't yet produce such videos. But I'm watching for them.

If you publish something online and want to declare that it is, say, open access, what do you do? You could just slap a Creative Commons logo on it, but these offer no guarantees - there are many cases where people have applied the license to content they don't own. This project addresses that problem. "CommonsDB uses a combination of cryptographic signatures and structured metadata to ensure the integrity and authenticity of content declarations." This post reports on the second phase of the feasibility study (49 page PDF) (the first was reported on last July). This phase "implemented the trust model in production, deployed three public APIs with a Developer Portal, enabled Data Suppliers to make live declarations, and launched the CommonsDB Explorer to expose registry content." It would be interesting to work with the APIs to both declare and use content. Related: Distributed Identity Foundation (DIF) Creator Assertions working group's user experience guidance document.