Stephen Downes

This is another update from Creative Commons signaling their shift from helping people access open content to helping copyright holders put conditions on access to their formerly open content. Specifically, they are "are advocating for the development and usage of carefully scoped AI opt-outs" and "doing research and development for a new tool designed to enable conditional access." If they're going to do this, I don't see why they don't just join ODRL or some other digital rights management (DRM) consortium. What they're doing has nothing to do with open access any more, and everything to do with locking down content. If Creative Commons were really interested in open access, they'd be thinking about how to enable free and open access language models everyone can use, so we don't have to pay multinationals to apply mathematics to content repositories.

Throughout my time as a student, from an early age to graduate school, my most powerful learning experiences occurred in what we call here "out-of-school time (OST) programs." I did a lot, some on my own, others based in the community: sports programs, Boy Scouts (and camp), Army Cadets, debating and public speaking, student newspaper. And so much more. For me, these programs were just 'there'. But they take time and a community to get right, and that's where this guide comes in. It describes four major principles: youth as network builders, interest-driven learning engagement, responsiveness to family contexts (addressing barriers like transportation, scheduling, and cost broadens), and grounding in long-term relationships. These are principles that I think should guide actual school-time programs as well. They create the real value that elite institutions provide students (as compared to training factories that emphasize job-ready skills). And long-term, I can envision the time we allocate to OST increasing and our emphasis on formal learning decreasing. 

If you develop software, you know what package managers are. They are services you can run in your development environment to find and install pre-packaged bundles of software designed for a specific purpose. For example, if I'm creating a web application, I might want to use React, which is a collection of web interface functions. Package developers load their software into the manager, and application develoipers use the managers to find and install that software. As this article notes, package managers "are now critical infrastructure for AI products. They're in the 'hot path', as we say in the biz." That's why AI labs like Anthropic + OpenAI buying them up? I can see the logic. But it's a fragile ecosystem, and I've run into issues many times, as one package is updated and creates a version conflict with the application or with other packages. Or one package develops a security issue that impacts a whole project. Or a package just doesn't do what you want it to do. I prefer to work without packages, if I can manage it.

If you always wanted to use Amazon Web Services (AWS) but didn't want to pay Amazon money, this is the tool for you: Floci, "A free, open-source local AWS emulator. No account. No feature gates." Now to be clear, I haven't tried to set this up, much less try to make it work in any serious way (and tbh I use Amazon pretty minimally as well - just S3 to store files and SES to send this newsletter by email). "Floci runs real Docker containers for services where in-process emulation would compromise fidelity - stateful databases, connection-heavy protocols, and runtimes that require native execution. The result is wire-compatible behavior against the actual engine, not a simplified approximation."

Though it recognizes that "our education and training systems were not designed for an AI-driven society," I regard the Castlereagh Statement - a declaration signed by an Australian 'Coalition of the willing' - to be at its heart a very conservative document (36 page PDF). It asserts three goals for Australian education and training: "a shared definition of what we value in human learners and educators, with aligned measurement systems; coherent learning pathways from early childhood to lifelong learning, aligned with societal needs; and every Australian being capable of confidently, critically, and creatively engaging with AI." It's always dangerous to begin a process of redefining what is value as a society, and always difficult to agree on a shared vision and collaboration, especially when it means that for most people the focus remains on jobs and skills, but the statement appears confident it has articulated that foundation, though Australians might not want their values to be redefined by some committee. Via Carlo Iacono. See also Learning Elearning with notes on the webinar to soft-launch the statement, and EdResearch Matters on how to implement the ideas it contains.

In my work developing CList I want to enable content and messaging federation - so that people can follow and talk to each other - without being dependent on a specific platform. So I want people to be able to create and use DIDs to help them find and follow each other. But how?