Stephen Downes

This is the sort of thing I like: "Centuries of copyright-free art, historical photography, illustration and archival footage. Find high-resolution files by describing what you're picturing, free to use forever. Every image and video is CC0 or public domain. Free to download and use, with no strings attached." How can they afford to do this? Decentralization: "Arkivism doesn't host the collections — the works belong with the museums and archives that keep them, listed with credit on the Sources page." There's also a search for video clips. Via Grant Potter.

This editorial (10 page PDF) introduces a special issue on AI-aided systemic reviews of research in online learning (also styled here as Open, Distance, and Digital Education (ODDE)). The authors respond to recent criticism of systemic reviews and survey some recent lapses, including "considerable variation in methodological approaches, reporting standards, and the transparency of review procedures." None of this is helped particularly by AI: "increased efficiency does not necessarily translate into improved methodological quality." So they propose standards for AI-assisted reviews that include stipulations that "methodological criteria must precede automation," transparency about the role of AI, systemic validation of AI decisions, and review process that are "sufficiently reproducible and auditable." There's nothing wrong with any of these, but they seem to me to set a pretty low bar. We're still in the era of (almost) any data counts, and (almost) any theory is a good one. 

I recognize that people will have a point of view on what it means when they share their online content, especially if they're institutions or commercial publishers. But I don't agree that "AI broke the open web's bargain" because there was never a 'bargain' to begin with. The idea that there is some sort of social contract describing a 'knowledge commons' is pure fiction. The authors here suggest that 'defensive enclosure' is the natural response to AI-based scraping, but that's only true if enclosure was always part of your 'business model' anyway. Most people don't work that way. They're not joining a 'knowledge commons', they're just talking to each other. The 'new internet bargain' being proposed will favour institutions and trample over people who just want to share their work without being drowned in legalese. Via Ben Werdmuller.

Just a day after I posted an article called Non-Cartesian Learning this article showing us just exactly where the Cartesian 'would' is coming undone appears in my inbox. It's a good read, accurate to my understanding of Descartes, and making a good argument. "LLMs are not programmed to carry out specific tasks, nor do they merely imitate; they learn, and like us, they do so by generalizing from individual examples and feedback to broader patterns... That doesn't mean these models work the same way our brains do (they do not), nor does it mean they are equivalent to people.28 But it does mean the dogma that the human mind and AI have nothing in common is increasingly tenuous." Quite right, and so is most of the rest of this article.

This is a link to a LinkedIn article, for which I apologize, but I wanted to highlight this. Csaba Szabo takes note of a paper called 'How contemporary academic structures constrain scientific creativity and hold back early-career researchers' and its 57 authors. Szabo writes, "What is remarkable is academia's ability to take a set of complaints that have already been repeated for years, repackage them as another publication, distribute the authorship over 57 CVs, generate another citable item, and thereby reproduce almost perfectly the very system the paper is criticizing." Quite right. Szabo should have also noted that this useless scientific output is behind a paywall. That's why I won't link to it directly (you'll have to search for it, like I did, because there's no link to it in the post, and I'm not going to click a dozen times to get through 88 comments to find the ubiquitous 'link in first comment').

Leaving aside the internal U.K. politics, which are beyond the scope of this newsletter, the contrast in the two approaches, and the option selected by most governments, is interesting. "In the UK's Growing Up in the Online World consultation, young people's most consistent demand was mandatory labelling of AI-generated content. The government response never mentions labelling, watermarking or provenance, and neither did 23 peers in the House of Lords debate on AI and children." There have been attempts to label AI-generated content, but in a murky "you can't see it" sort of way. After seeing the debates and policies around food labeling, I can't see anything clear and straightforward being developed. Nor can I see the social media bans enduring for any significant length of time.