Stephen Downes

The most important paragraph in this article on the (generalized and widely construed) science of learning (SoL) is the last, and though it takes a lot of work to get to this point (some of which I would disagree with) the bottom line is sound: "Although there is a wealth of research exploring "what works best" in education (eg, John Hattie's Visible Learning), it's important not to confuse this work with SoL. Whenever experiments deal in strategy rather than mechanism, they are considered 'educational' research rather than 'learning' research. Without a doubt, educational research is highly important but, like all practical issues, it will forever be contextualised, naturalistic and value-laden. This means 'what works best' will necessarily change depending upon context and goal." That's why, for example, "teachers well versed in the science of learning are more likely to employ student-centered, constructivist approaches within the classroom." But let's also be clear: progressive educators don't dismiss the science of learning (lower case), but they do dismiss politically-motivated education research movements with the title 'Science of Learning' (upper case). And that's fine.

There's a long discussion of whether 'digital nomad' is the right term, or maybe 'migrant', or even 'hostage'. The authors also "consider digital identity formation as a complex set of practices whereby users of social media platforms form an outward presentation of themselves." There is also a "feeling of digital nostos, a longing to return to a digital place where we felt at home." Carving out a space for academics on social media platforms is difficult. "We asked – and we still ponder – whether these social media spaces could be something that academics could re-territorialize, differently." Finally, "As some of the established social media like the social network formerly known as Twitter 'go down' and others like Mastodon or Bluesky 'rise' we must finally realize that these or indeed any social media platforms will always continue to ebb and flow – they are out of our control." This, of course, has always been true. Image: Gondy Leroy.

This item cites research describing the tendency of children to offer personal information over VR. "In my experience," says one participant, "it doesn't take much to know a kid's private information. Sometimes when I go into rooms, I see kids share all the information publicly and voluntarily...Kids don't understand that it's just like standing on the street corner and shouting it." All true, but let's not blame the kids. We should be clear, I think, that the problem is not the VR technology, the problem is the companies collecting information using VR. (Also, I found the 'race' data collected a bit weird - leaving aside whether race is even a scientific category, the table lists only one 'race' for both parent and children, which seems rather prejudicial to me).

This is a shortish and matter-of-fact article describing authors' recommendations for integrating generative AI into higher education. It begins with some definitions of key terms, a brief overview of relevant literature, and then offers a set of experts' recommentations, including: generating awareness of the coming disruptive change and faculty training; changing teaching and assessment practices and importing learning libraries adapted to GenAI; and applied research. They also offer a four-level scale of a GenAI adoption and finally suggest directions for future research. Image: Integrating Generative AI into Higher Education: Considerations, Educause.

Perplexity is a generative AI (GAI) interface that promises accuracy and cites real sources in its answer. "It has a conversational interface, contextual awareness and personalization to learn your interests and preferences over time." Wikipedia notes, "the paid version Perplexity Pro has access to GPT-4, Claude 3, Mistral Large, Llama 3 and an Experimental Perplexity Model. It has garnered about 10 million monthly users." The current link advertises a seven-lesson course for teachers to introduce them to Perplexity. So is it any good? I asked it, "What is connectivism" and received a reasonably accurate response with a questionable set of citations. Via Miguel Guhlin.

This is a bit of an update on the topic of fake news and misinformation. The use of the term 'psyop' is a clever way to externalize (and militarize) the use of language as a weapon. Reading through the three steps, however, I was first struck by it being centered on U.S. politics (the first step is "pay attention to media messages that paint groups of American people as foreign, or somehow not quite "real" Americans"). This is known more generally as 'othering' and can be used anywhere. But I was also struck by how dated and clumsy the mechanisms are. Misinformation today is far more sophisticated, personalized and targeted, generated by AI, used to create not just othering and criminalization but also loyalty, affinity, identification and many more attitudes, and used widely for commercial, political and military purposes. Via Ben Werdmuller.