Stephen Downes

As background, you might want to first read Conscious Processing and the Global Neuronal Workspace Hypothesis (it's OK, I hadn't seen it either). This is a great article with an even better diagram articulating how (conscious) experience in a 'neuronal workspace' may be connected to and informed by (unconscious) more specialized 'workspaces'. It makes me think of 'communities of communities'. "Baars's global workspace involves processors related to the past (memory), present (sensory input, attention), and future (value systems, motor plans, verbal report). Thus, the global workspace achieves experiential integration that is, in terms drawn from the philosophy of mind, both synchronic (at a particular point) and diachronic (over time)." Now that we're caught up, Eric Schwitzgebel throws a spanner into the works - what if there is no global unified workspace? "On this model, disunity is the normal human condition. Our experiences are fragmented, except when we pull them together through attention. We just don't realize that fact."

Altering a title sounds really bad. And there's a reason why I keep the title of the article the same when I write a commentary on it - I want to be sure I'm not distorting the intent of the author by altering their title. So this Verge report seems concerning: "Google is beginning to replace news headlines in its search results with ones that are AI-generated." But a search engine's responsibility is a bit different from mine. As reported here, "For content that could impact someone's health, finances, or legal situations, Google seems far more concerned with making sure titles are accurate and helpful rather than keyword-optimized." That's actually a useful service, especially in a world where titles are so often used to mislead.

According to Maarten Boudry, the problem with teaching students how to spot fallacies is that they start seeing them everywhere. "They hurled labels and considered the job done. Worse, most of the "fallacies" they identified did not survive closer scrutiny." And the gist of the article as a whole is that "human reasoning is far more sophisticated and subtle than the theory of 'fallacies' suggests." As someone who has taught and written about fallacies, I am inclined to agree with both parts of this. But I never abandoned the teaching of fallacies, though I did adapt my method. Identifying fallacies is a three step process, I said. First, you can learn to recognize the 'signs' that a fallacy is present. But signs are often misleading; you need to reconstruct the reasoning to confirm that there is, indeed, a fallacy present. Finally, you need to show not simply that the fallacy is present, but to use your understanding of the fallacy to show that the reasoning is incorrect. If you name the fallacy in your response, I would say, you're doing it wrong.

I am at least partially influenced by the fact that I did read Paul Churchland when I was younger, and came to much this sort of belief: "It is very common to see confident assertions that LLMs mimic language use but do not really understand or use it the way that we do, that LLMs do not really reason or think, that they cannot know or understand things. On examination, these claims are often grounded in a folk-psychological understanding about how we think, know, or use language, or, at best, in ideas from philosophy or cognitive psychology that are profoundly disengaged from any understanding of the underlying mechanisms of the brain." 

This article starts by quoting in full a post of mine that has gotten some traction on LinkedIn describing "the impact of AI on higher education." It was preparation for an event I'll participate in later this year. The thrust of David Truss's comment isn't to agree or push back, but rather, to ask, who will get us to this vision? Who is this 'we' of which I speak? "'We' won't get there following the guidance of financially lucrative edu-tech business," he writes. "'We' won't get there like we did with Web2.0 tools in the late 2000's and early 2010's, on the backs of tech savvy educators leading the charge. 'We' won't get there because of some governmental vision pushing a new AI enhanced curriculum." Fair point. If the model of 'educators' is 'teachers working in schools following institutional guidelines' then they are unlikely to move us from point A to point B. No, I was thinking (and this should surprise no one) of 'educators' as 'people like me' - working as educators but not typically in education. I have long said that change will come from outside the system. I don't doubt today that this remains true. 

I have long been fascinated by the observation that the fields I cover here are composed of what might be called 'communities of communities', that is, clusters of writers and practitioners that tend to coalesce into smaller cooperative networks while still being connected to the wider community. This article (29 page PDF) in Higher Education both embraces and resists that idea when it comes to a history of its own contents. It wants to be a systemic review, but the data don't coalesce into a single overarching theme. We see an ebb and flow of ideas and concepts, along with the citation networks of practitioners that swirl around them. "The early 2000s saw... the onset of institutional and methodological transformation... 2006-2015... indicates a shift towards macro-level analyses, emphasising structural, political, and social dimensions of higher education... 2021-2025... suggests a renewed orientation towards measurement, pedagogical modelling, and teacher-centred research." (p.s. the diagrams could have used much tighter editing; the headings of table 2 are incorrect, the hierarchical structure of Figure 3 is masked by lines flowing for no reason behind blue circles, the prominent (and hyphenated) 'higher-education' in the word cloud is suspicious, and the flow from concept to concept in Figure 7 appears to be arbitrary).