Stephen Downes

So the answer here, which is never clearly stated, is "no". That's not to say synchronous sessions are without value. "Live sessions, real-time conversations, orientations, guest speakers, and collaborative meetings can support connection, clarification, and engagement." But for many people, especially if they are a surprise addition to a course, or not tied directly to the course goals, or arbitrarily made mandatory, they can act as a barrier. My own approach has always been to include synchronous activities, because of the value, but to be sure to fully record them, so nobody misses out. 

The argument here, stated about half way through the post, is that "when content becomes abundant, judgment becomes scarce." So "that is where learning professionals can bring more value." Let's think about that. A while ago I proposed a type of AI, called "deontic analytics", would decide what is worthwhile, what is valuable, what should be done. I've seen nothing that suggests AI will be incapable of this. Why would we trust human judgement over AI, if AI can be demonstrably better at this? Consider the new decision models, like Jev and Clef and Decision 1.0 (I'll be a human named that one). It's not obvious that human judgement will always be better. Image: Shadhin.

EThis article describes the evolution of the Glossary of Digital Learning into a linked data resources that can be accessed by algorithms and data models. "Each concept now has a stable identity, source provenance, machine-readable JSON-LD and Turtle representations, and a place within a queryable semantic model." I get the intent of this - we want the model to depend on a reliable data set to inform its understanding of basic terms and concepts. What concerns me (and it is a theoretical concern only; I have no particular issues with this glossary) is how we come to define and formalize these terms and concepts. Do we leave it to expert bodies to define them a priori? Do we employ a mechanism that bases the definitions in actual usage? Is a semantical representation the best way to do this, or does it create sharp edges where fuzziness would be preferred?

I remember once reading everything I could find by Nicholas Rescher in the university library (he was a very prolific writer). He wasn't on any of my course lists, and wasn't recommended to me by anyone; he was just on the same shelf as some theories of truth I had been assigned. I liked his approach, and read his work with enthusiasm. That's the sort of thing being described in this post. We hear about it a lot, and it expresses the common concern that serendipity is being eliminated as we move from the physical to the digital. I disagree. Serendipity is a state of mind, not an environment. I encounter it in the digital world every day, because I take the time to explore the space around whatever I'm looking at. And I make sure my RSS feed contains everything from Satya Nadella writing on 'Super Intelligence' to Colossal covering the Timeless Swiss Masked Tradition of Silvesterchlausen. If you can't find serendipity on the web, you just aren't trying.

This post from a primary school teacher and Master's student from the Benemérita Escuela Normal de Coahuila (BENC) in Mexico recounts her experience attending the UNESCO's Digital Learning Week 2026 in France. "It was not a feeling of smallness as a synonym for inability, but rather an awareness of standing before an immense ocean. An ocean of knowledge, research, experiences, projects, questions, and possibilities that extended far beyond what I had previously been able to imagine." As she says, you don't need to know the entire ocean in order to begin exploring it. Really good reflection, worth the read.

This page contains Microsoft's Humanist AI Code of Conduct (38 page PDF), an analogy I guess to the Claude Soul document, and codifies Their "approach to developing and deploying Humanist AI, and contains the intended behaviors, values and guardrails for our models." It is motivated by the assertion that "humans must retain meaningful control over AI so that it can help people live healthier, happier, and more productive lives." That feels a bit loose to me - I would rather define for myself the purpose of my life, rather than have AI acting as though I should be productive. This looseness characterizes the document; there are numerous places where I would want to limit what AI should assume about people. I also found it interesting that Microsoft back in 2025 said it was "working towards Humanist Superintelligence (HSI)".