OLWeekly

Taking AI literacy seriously
Philip J. Kerr, Adaptive Learning in ELT, 2025/04/04

Philip J. Kerr raises five 'big issues' related to teaching critical literacy for AI, but in my view he is wrong in at least four of them. First, he says, "critical thinking is domain specific." No it's not. Second, he says "critical thinking can only be taught if learners have a disposition to think critically." Also not true; such a disposition, if it exists at all, is a result, not a precondition, of learning to think critically. Third, "a failure to think critically... is not primarily a problem of critical digital literacy." Leaving aside the mixed definitions, yes it is. People who have learned critical thinking aren't as likely to be fooled by disinformation. Fourth, "evidence for any effectiveness in critical digital literacy instruction is in very short supply." Maybe? Not going to check, but I have my doubts. Finally, fifth, "is it even possible to imagine a critically informed and ethical use of Generative AI?" Of course it is.

Web: [Direct Link] [This Post][Share]

View of Policies for Artificial Intelligence in Higher Education: A Call for Action | Canadian Journal of Learning and Technology
Mohamed Ally, Sanjaya Mishra, Canadian Journal of Learning and Technology, 2025/04/04

The argument is that "Higher education must prioritize AI policy development so that AI is trustworthy and used for good." I think we can agree we want these, but are they ensured (or even made more likely) by higher education policy development? Higher ed policy governs only higher ed, but AI development and use happens in the wider community, so even if effective, the policies are limited in their application. Anyhow, the short article lists 14 AI policy areas and 52 AI policy competencies.

Web: [Direct Link] [This Post][Share]

Rep. Marjorie Taylor Greene is wrong, ‘Sesame Street’ was not replaced by 'drag queen storytime'
Grace Abels, Poynter, 2025/04/04

This story reports that "Sesame Street continues airing on PBS and has not been replaced by drag queen storytime." What it misses is, I think, the main point, which is that it would be good if children were able to watch drag queen storytime and that early exposure to different and diverse cultures could do much to erase the intolerance and prejudice that exists today.

Web: [Direct Link] [This Post][Share]

The Mercury News put all its news on the web for free 30 years ago. Did it open Pandora’s box? - Poynter
Pete Croatto, Poynter, 2025/04/04

I've had a lot to say about this over the years, and there's more coming. But I want to set the record straight here. Pete Croatto writes, "The Mercury News would become widely regarded as the first newspaper to put its entire content online, initially for free. The decision established a lasting expectation that online news should be free." It did not. What created that expectation was that, on the internet (in contrast to closed commercial services like AOL) everything was free. You paid for your own computer, and you paid for your own internet access (or maybe your institution did), and then, you shared your content with everyone else. It's important to understand that you don't actually need commercial content to make the web work. You don't even need to pay for content production. People will create their own content, including news content, for free. 

Web: [Direct Link] [This Post][Share]

EUDI Wallet based Strong Customer Authentication and payment for Financial Services
Adrian Doerk, Lissi, 2025/04/04

You might not know about EUDI yet, but it - or something like it, if you're outside Europe - will be coming into your world soon. The European Digital Identity Wallet (EUDI Wallet) is a new method for Strong Customer Authentication (SCA) into digital services. Crucially, in Europe, "all private relying parties in regulated sectors - including banks - must support the European Digital Identity Wallet for SCA-related processes by 2027." The EUDI wallet supports core online functions: login, payment, data access, whitelisting, personal information, and consent. This article looks at EUDI in more detail, and EUDI in education is on my watchlist.

Web: [Direct Link] [This Post][Share]

MyNotes: Carnegie Learning’s State of AI in Education
Miguel Guhlin, Another Think Coming, 2025/04/04

Miguel Guhlin shares his summary of the "Carnegie Learning AI in Education webinar facilitated by Amanda Bickerstaff. The webinar is accompanied by a link to a report from Carnegie Learning." There's not a lot new here, but it's nicely packaged, and I mainly want to express appreciation for Guhlin's coverage of this and consistently good coverage of AI and critical thinking over the last few months. I'm not linking to everything, but it's all worth a read.

Web: [Direct Link] [This Post][Share]

Facing the Looming Threat of A.I., Publishers Turn to Decentralized Platforms
Ben Werdmuller, Werd I/O, 2025/04/04

Ben Werdmuller references "a lovely piece about Mike McCue, who, through Flipboard, Surf, and his general activities through the community, has become one of the open social web's most important figures." He points out that while Surf is different from them, it's not forcing people to make a technology choice. "This is the point that A New Social is making too: it's not about picking a protocol, because the protocols can easily be joined together." This is a refreshing change from silos like Facebook and X/Twitter. Closed access systems, like authoritarian governments, may seem to make things better for a while, but as Jonathan Last writes, "no empire can survive the degeneration of its people."

Web: [Direct Link] [This Post][Share]

Did an LLM help write Trump’s trade plan?
Gary Marcus, Marcus on AI, 2025/04/03

This is a textbook example of an 'inference to the best explanation', also known as abductive reasoning. Explanations are evaluated by their coverage (do they explain every case?) and by their simplicity(do they involve the fewest parameters possible). Good explanations are also testable. This meets all three criteria: all the tariff rates are explained, it's a very simple explanation (trade divided by trade deficit equals percentage), and it can be tested by putting the same question to AI engines. Sure, the explanation means that the American calculations are completely made up, but that's actually more plausible than any alternative. Kudos to James Surowiecki for working this out (though to be fair, the crowd probably helped him).

Web: [Direct Link] [This Post][Share]

MCP Security Notification: Tool Poisoning Attacks
InvariantLabs, 2025/04/03

It took this long to find a problem in model context protocol (MCP) servers. "A Tool Poisoning Attack occurs when malicious instructions are embedded within MCP tool descriptions that are invisible to users but visible to AI models." The page shows how a local MCP server could be prompted to extract private keys and send them to an attacker's email address.

Web: [Direct Link] [This Post][Share]

NaNoWriMo shut down after AI, content moderation scandals
Amanda Silberling, TechCrunch, 2025/04/03

National Novel Writing Month (NaNoWriMo) was a fun idea that started as a Yahoo discussion lost. Somewhere along the line it commercialized, forming a 501(c)(3) nonprofit (you might argue that this is non-commercial, but I struggle to understand why it needed financial organization at all). Anyhow, the inevitable happened: NaNoWriMo "announced on Monday evening that it is shutting down." The causes of the commercial failure included "longstanding financial issues that have made it difficult to operate", board member resignations over its stance on AI, and "controversies over content moderation." My suggestion: create and follow a #nanowrimo hashtag on Mastodon, write what you want this November, and share your stories using the hashtag. No commercial organization needed.

Web: [Direct Link] [This Post][Share]

Watch Global & Local Live TV Online for Free
2025/04/03

Remember I Crave TV? This is the same. I have no idea how legal it is, so enjoy free Angolan TV while you can. "Watch Free Live TV from Around the World on tv.garden! Stream live news, sports, and entertainment—no subscription needed. Start watching now!" Via Alan Levine.

Web: [Direct Link] [This Post][Share]

Protect authors’ livelihoods from the unlicensed use of their work in AI training
Anna Ganley, Change.org, 2025/04/03

Referencing an article in the Atlantic that is, naturally, behind a paywall, this petition launched by The Society of Authors demands "that Meta is held to account by the UK government following allegations in the U.S. that authors' works have been used without permission or remuneration to train its AI model." It would not surprise me at all that Meta would simply disregard copyright law in pursuit of the AI genie, just as it would not surprise me at all to see the same company vigorously prosecute third party scraping of its websites to collect data for their own purposes. Via Sheila MacNeill.

Web: [Direct Link] [This Post][Share]

AI In Education
Inclusive Design Research Centre, 2025/04/02

The Flexible Learning for Open Education (FLOE) project has put together a collection of articles on AI in education from a set of well-known education and edtech writers. From OpenStax (now in a partnership with Google) Richard Baraniuk writes, "to create Responsible AI for education, we must build it on an open, equitable foundation." Beth Rabbitt suggests AI should be a public good for education. Jutta Treviranus warns "Because AI is a powerful statistical pattern replicator, it is poised to make every other harm even worse for people who are already struggling," calling this phenomenon 'statistical discrimination'. Anne-Marie Scott points to legal issues and references the OSI definition of 'open AI'. Sarah Johnson focuses on he usability of AI products. And so on, with more authors. FLOE is a project of from The Inclusive Design Research Centre and OCAD University.

Web: [Direct Link] [This Post][Share]

New SRAM Code hydraulic brake install - cut cable, Stealth-a-majig, install on bike, bleed brake
YouTube, Chasing My Freedom, 2025/04/02

OK, this is a 40 minute video on how to install a hydraulic brake on a bicycle. I just watched it beginning to end. It paced me nicely through the process, clearly showing me the tools used, step by step process, and what the result should look like. It's an example of direct instruction done the way it should be. Now I wouldn't say I have learned how to install the brake until I did one on my own, but I feel I could probably manage it after watching this. Now (with a nod to Kirschner) I wouldn't want to attempt this without prior instruction. It's too complex, there are too many steps that could go wrong, and the cost of failure is too high. Yet it has to be noted, there's no way I'm going to sit riveted with interest in this video without the prior experience of riding my bicycle and developing a very specific need for this knowledge. It also helps (a lot!) that this video and many others like it are available for free online.

Web: [Direct Link] [This Post][Share]

Markdown and the Slow Fade of the Formatting Fetish
iA, 2025/04/02

This is an interesting article though I don't exactly agree with it. It describes 'markdown', a system some text editors use to allow writers to create bold, italics, lists, headers and links without using complex HTML or even more complex .docx formats. Though the article attributes markdown to John Gruber and Aaron Swartz, it draws on conventions people used for decades in text-only editors, for example, **bold** for bold. What makes it different is you can run markdown text through an interpreter and it will produce nice HTML format for you. The article says "As a professional writer, you probably don't care much about bold, italics, and underline, while you write," but from my on perspective, I care quite a lot about formatting. Via Stephen Harlow.

Web: [Direct Link] [This Post][Share]

FediForum Has Been Canceled - We Distribute
Sean Tilley, We Distribute, 2025/04/01

The cancellation of FediForum as a result of two-year old anti-trans statements by one of the organizers is representative of the cultural conflict roiling the western world right now. It's particularly relevant here given the trans-positive origins of the fediverse. For my own part, it's hard for me to understand the hate some people feel for diverse and different people, including trans people, and I wish we could learn to live in an inclusive and equitable society. The whole concept of the fediverse is something like 'live and let live' and  'to each their own' and it's a concept I embrace, especially when compared with what we see in places like Twitter and Facebook.

Web: [Direct Link] [This Post][Share]

It’s time to examine neural coding from the message’s point of view
Daniel Graham, The Transmitter, 2025/04/01

By 'messages' we refer "not to individual spikes but rather to patterns of excitations - comprising spikes, their timing and their spread across the wider network." The animation at the top of this article offers a good example of a 'message' as the sequence of activations (signified by red coloured neurons) spreading out from the original signal. Characteristic input produces characteristic messages in a given connectome. "Message fingerprints—namely spiking patterns—showed up in specific neurons in a specific order, again and again." It's still early days - "we don't yet have an accepted definition of what constitutes a message." But I suggest that what's being called a 'message' here is sometimes a sensory experience - in other words, consciousness.

Web: [Direct Link] [This Post][Share]

The Mediocrity of Modern Google
Om Malik, On my Om, 2025/04/01

I've been researching freeze-dried food for my upcoming bikepacking trips. Because of recent events, I want to ensure I'm buying from Canadian suppliers. I got a list of providers from Google and then searched on each to check its web page. But Google clutters the results with lists of 'in stock' products at Amazon instead, so I added the '-ai' flag to my search request to knock out the AI. AI fail, right? But then I asked ChatGPT for a list of Canadian freeze-dried food companies and got a good list, with descriptions and links to the right websites. The lesson here is two-fold: Google search is becoming useless, and is not helped by AI, but if I adapt my own practices, AI generates better results. All of which sort of validates Om Malik's point here.

Web: [Direct Link] [This Post][Share]

Parallel Universes and the Loss of Civil Discourse
Dean Shareski, Ideas and Thoughts, 2025/03/31

Even though I had a mainstream Anglophone education, from my perspective, everyone has a different point of view from me. So I expect people to disagree, and unless I want to be a practicing solopsist, I need to be able to interact with them on a relatively civil basis. It is, however, as Dean Shareski laments, becoming a lost art. "I remain desperate to find examples of thoughtful, intelligent people engaging across ideological divides—people who wrestle with difficult issues, challenge each other's thinking, and explore the real-world consequences of policy decisions," he writes. It's hard, because we grapple with each other not only with reason and logic, but with sentiment, emotion, passion, culture and history. Still, like Shareski, I remain "committed to diversity, critical thinking, media literacy, and civil discourse," even though these very foundations of civil society have recently come under attack.

Web: [Direct Link] [This Post][Share]

Writing by hand builds reading, writing and thinking skills
Joanne Jacobs, 2025/03/31

What's interesting about this article is the chart that accompanies it depicting the alphabet and numbers as written in cursive. Take a look at it. Now ask, why did we develop cursive writing when we had printed text as a perfectly good system of hand-printed text already in existence? It has to do with the quill pen and ink, and specifically, the fact that cursive allows you to write text without lifting your pen off the page, thus preventing drips and smudges. It was a shortcut. Anyhow, if we look at that chart, we can see that this original purpose has been lost, as many of the lowercase letters to not properly join with the letters preceding or following them. And all of that tells me that the whole idea that 'cursive writing builds x skills' is a rationalization created by people who do not understand why cursive exists (and why it is no longer necessary).

Web: [Direct Link] [This Post][Share]

‘What we see are wasted lives’
Rosa Furneaux, Schools Week, 2025/03/31

What attracted me to this item was not the tales of abuse in illegal schools but rather the philosophical question emerging from the headline: what counts as 'wasted lives'. Some cases are obvious, such as students who may perish in a fire when the doors are locked. Arguably, so are cases where the 'education' is nothing more than work in a sweatshop or when the teachers and staff are abusive. "Misogynistic behaviour doesn't just happen at religious schools." But where does the limit lie between a 'wasted life' level of abuse and merely 'undesirable conditions and outcomes' type of school? It's hard for me to imagine any life that isn't prematurely ended as 'wasted'. But that might just be my own bias. (p.s. please do not construe this as support for illegal or even private schooling; it is not).

Web: [Direct Link] [This Post][Share]

Organoids and assembloids offer a new window into human brain
Sergiu P. Pasca, The Transmitter, 2025/03/31

Today's new word is 'organoid', the outcome of "growing (neural) cells into more complex 3D structures... that mimic some of the structure and function of regions of the nervous system." The technique is useful because "neurons in these three- or four-part assembloids extend axons and connect with some specificity to other neurons (which) gives rise to emergent properties, such as the contraction of human muscle." They can be maintained long enough to emulate post-natal maturity and can be studied to find underlying causes of various conditions such as Timothy syndrome. An assembloid, meanwhile, is what researchers get when they "combined different types of brain organoids into structures... in which cells from different origins can intermingle."

Web: [Direct Link] [This Post][Share]

The 253 Most Cited Works in the Stanford Encyclopedia of Philosophy
Eric Schwitzgebel, Daily Nous - news for & about the philosophy profession, 2025/03/31

This list is a pretty good measure of "influence in mainstream Anglophone philosophy" and also a pretty good indication of a large part of my own philosophical background (which is supplemented by a lot of reading in religious studies and scientific works). I won't say I've read them all (*) but the first hundred or so are quite familiar to me and I can cite a good selection of the rest of them. (* I once said "I've read them all" to someone who was looking at the books on my office shelves. It was obviously untrue and I was obviously caught out and I've regretted it ever since. I don't know why I said it and I guess it reflects a character flew wherein I pretend to be more knowledgeable than I am. Anyhow, not making the same mistake twice.)

Web: [Direct Link] [This Post][Share]

Privacy died last century, the only way to go is off-grid
Steven J. Vaughan-Nichols, The Register, 2025/03/31

Here's the tag from this item: "I was going to write a story about how Amazon is no longer even pretending to respect your privacy. But, really, why bother?" The gist here is that while we still think we have privacy, we really don't; it has long ago been lost in a world of data breaches, government programs and (I would add) Equifax. It's a useful perspective. The concept of privacy itself, I think, is relatively recent and limited to urban environments in the more developed world. Growing up less wealthy and rural, I was raised in an environment where everybody in town knew your business and you knew theirs. That was often useful - and for some people, vital - information. Privacy is, ultimately, the right to lie about yourself in public, generally to protect yourself from the consequences. I get that there are reasons to do this, but it's something that should be talked about, not assumed as an automatic - and existing - good.

Web: [Direct Link] [This Post][Share]

Just a metatool? Some thoughts why generative AIs are not tools
Jon Dron, Jon Dron's home page, 2025/03/31

Jon Dron prefaces his argument with a discussion of what counts as a tool ("something that an intelligent agent does something with in order to do something to something else") which seems not quite right (are elephants 'intelligent agents'? is a 'pen and paper', thought of as a system, not a tool?) but which does the job, which is to get us to the essence of the argument, which is this: "The big problem with treating generative AIs as tools is that it overplays our own agency and underplays the creative agency of the AI." Specifically, "It encourages us to think of them, like actual tools, as, cognitive prostheses, ways of augmenting and amplifying but still using and preserving human cognitive capabilities, when what we are actually doing is using theirs." Yes, non-humans - including animals and now machines - have cognitive capacities.

Web: [Direct Link] [This Post][Share]

OpenAI Academy
OpenAI Academy, 2025/03/28

OpenAI has launched something called the OpenAI Academy, which Siya Raj Purohit writes is "our community-first approach to making AI literacy accessible, inclusive, and global." Sign-up is free and open to all (within, I assume, the context of U.S. restrictions on people from certain countries) though you don't appear to need to register to access the content, which consists of fairly basic introductory videos. Some events are also on offer. OpenAI may be good at AI, but I don't think they've mastered online learning yet.

Web: [Direct Link] [This Post][Share]

Learning Design in the Era of Agentic AI
Philippa Hardman, Dr Phil's Newsletter, Powered by DOMS™️ AI Dr Phil's Newsletter, Powered by DOMS AI, 2025/03/28

According to Philippa Hardman, "The rapid emergence of agentic AI has forced the learning and development field to confront a long-standing truth: most asynchronous online learning is not well designed." This article three major shifts that will be required to adapt: first, from organizational to learner-centered learning goals; second, from passive to active information consumption; and third, from measuring learning activity (like clicks) to learning outcomes. It's worth noting that all of these were needed long before we got agentic AI. It's only now we're seeing how urgent they are.

Web: [Direct Link] [This Post][Share]

AI is transforming university teaching, but are we ready for it? - University Affairs
Loleen Berdahl, University Affairs, 2025/03/28

The questions here are what's interesting and significant: "Are faculty going to be leaders – or, barring that, engaged participants – in determining how AI shapes the future of higher education? Or will they leave this to those outside the sector?" Also, "AI is already changing how we teach... But will it also change what we teach?" 

Web: [Direct Link] [This Post][Share]

Model context protocol (MCP) - OpenAI Agents SDK
OpenAI, 2025/03/28

The model context protocol (MCP) was introduced last November by Anthropic and has spread across the large language model community. This page is OpenAI's documentation describing how it too now supports MCP, which essentially cements the importance of the protocol for developers. What it does is to allow an LLM to access a 'context' in the form of information and services from local or remote systems; an MPC server would allow chatGPT to access a database, for example. Image: Norah Sakal.

Web: [Direct Link] [This Post][Share]

Artificial neurons organize themselves
Abdullah Makkeh, Marcel Graetz, Andreas C. Schneider, Michael Wibral, IDW, 2025/03/28

Interesting discussion of new types of artificial neural networks, which would build on the simpler networks in use today to produce such things as chatGPT. "The new artificial neurons, known as infomorphic neurons, are capable of learning independently and self-organized among their neighboring neurons. This means that the smallest unit in the network has to be controlled no longer from the outside, but decides itself which input is relevant and which is not." Here's the full paper on PNAS. The neurons in these new networks pursue more specifically defined learning objectives, resulting in "infomorphic" neural networks.

Web: [Direct Link] [This Post][Share]