Stephen Downes

I know some people are completely tired of hearing about large language models (LLM) but they have had such an impact on the field of educational technology (edtech) that it would be irresponsible to ignore them. Fortunately, Simon Willison offers this set of 27 briefly described milestones for the year (it sounds like a lot but the article is actually a pretty quick read). Be familiar with this list and you will be able to speak and listen knowledgeably about cutting-edge edtech in the coming year.

The irony is that this page somehow managed to blast a video ad with super-loud sound as soon as I opened it (though UBlock seems to have blocked it the second time). Anyhow, the article describes the application Nobook, which is a reader for Facebook, which removes all the advertisements and recommended posts. As the author notes, "I could immediately view posts from friends and groups that I follow without being bombarded by the things that tend to get in the way." The app is interesting, first, because it's the private project of a single individual, and second, because it's basically a web browser that 'wraps' around Facebook and then delivers only the content users want to see. I think this will be a point of contention in the future, as services seek to clamp down on the freedom people have to use browsers - rather than proprietary apps - to view their content.

This article was written in 1970, when the author, Joseph Weizenbaum (the inventor of ELIZA) was 47 years old. Having escaped Germany before the war, he understood from experience the argument that there should be moral constraints on science, and the first part of this article puts it eloquently into text. People can and quite rightly should debate the ethics of science, because science is not removed from the rest of society. But on what basis? Here the argument is weaker. Weizenbaum offers two criteria: we should reject things "whose very contemplation ought to give rise to feelings of disgust in every civilized person" such as "the proposal that an animal's visual system and brain be coupled to computers"; and we should reject things "that which can easily be seen to have irreversible and not entirely foreseeable side effects," such as "automatic recognition of human speech." But he also says, "I am, of course, aware of the fact that these judgments of mine have themselves no moral force except on myself," and calls for people, and especially educators, "to be a whole person... (with) the courage to confront one's inner as well as one's outer world."

This is a fun read, especially since I remember all these obsolete tags that ought never have been created in the first place - though I did get a lot of mileage out of frames and tables. "These stumblings have made HTML what it is today. No, they weren't perfect, but it is mistakes that we learn from. These obsolete tags represent experiments in web development - some more successful than others. They're a reminder that the web platform we use today is the result of decades of trial, error, and evolution."

Part of my plan for this year's holidays was to migrate all my stuff to a single virtual private server (VPS) in an effort to reduce my spending on cloud services. I wrote an RSS reader instead and have been uploading Iceland photos. So much for planning :p Still, what I intend is something similar to Doug Belshaw's setup, which he describes here: "what follows is the stack I currently use, why I chose it, and some ideas for how you might borrow ideas for your own site. This, of course, is not the only way to do it, but it is a concrete plan and running in production on my blog." As I set up my own new stack I'll document it. My efforts are amateurish compared to people who know what they're doing, like Luc Belliveau and Dave Lane, but I think I'll get it closer to right this time.

By far the most interesting part of this post is in the second half where Bryan Alexander discusses how he used AI in authoring the post. To summarize a longer description, he wrote a draft, improved it with AI, and used AI for the image. I don't judge people according to whether or not they use AI. I focus on the results. The post is fine, so far as it goes, but as Alexander himself notes, it's still a bit long. It's a bit of judgement call, though, between preserving the original voice, which might be a bit wordy, and tightening it up with AI, which may push the article to the mediocre middle. As for the predictions: well, I think Alexander needs to read more broadly outside 'traditional' US-based education media. He says, "Not discussed very much (as far as I've seen) is the possible impact of agentic AI on higher education" and points to four very basic posts. I counted more than 60 references to AI Agents in OLDaily in 2025. The dearth of a wider background makes prognostication very difficult.