OLWeekly

A Possible Trauma-informed Approach to AI in the Classroom?
Maha Bali, Reflecting Allowed, 2024/12/02

Reading this makes me think the following: the use of AI by students for assignments should be a matter of negotiation, which means in turn that the design of assignments is a matter for negotiation, where it is understood that students are able to use AI if they wish, but that the assignments should not be constructed such that they receive credit for the work performed by the AI. That eliminates the possibility of AI being used as a short cut, but allows for its use as a tool.

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

First Official EU Ethical Guidelines for Blockchain Systems
Signe Agerskov, Dataetisk Tænkehandletank, 2024/11/29

We haven't heard a lot on the blockchain front lately but it's still out there (and if you dig deeply enough you see it showing up everywhere). Hence the utility of this new set of ethical guidelines for blockchain systems (53 page PDF). There are five major dimensions: fairness, privacy, security, accountability and social responsibility. There are some good bits here, for example, "blockchain systems' governance power should be distributed to minimize the risk of unintentional concentration of decision power." Also, "designers should consider how tokens incentivise behaviour before any assets get tokenised." A lot of thought has gone into this document and it seems to me to be one of the better bits of writing on ethics and technology.

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

Epoch AI Unveils FrontierMath: A New Frontier in Testing AI's Mathematical Reasoning Capabilities
Vinod Goje, InfoQ, 2024/11/29

According to this article, "60 mathematicians from leading institutions worldwide (have) introduced FrontierMath, a new benchmark designed to evaluate AI systems' capabilities in advanced mathematical reasoning." Basically it's a set of mathematical problems drawing from such bodies of knowledge as number theory, probability and geometry to be posed to AI systems to test their mathematical skills. It's a tough test; current large language models score in the 2% range (not a typo: that's two percent). What I wonder is how well humans would perform on the same test. Now obviously, mathematicians who have had a lot of experience and practice would do well. Me? Well I would have to brush up. A lot.

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

The Manifesto for Teaching and Learning in a Time of Generative AI: A Critical Collective Stance to Better Navigate the Future
Aras Bozkurt, et al., Open Praxis, 2024/11/29

This paper has 47 authors, give or take, and is a bit of a round manifesto stuffed into the format of a square research paper, but despite that awkward fit it offers a fairly cogent summary of the potential benefits and risks of AI in education. Where it is substantially weaker, to my mind, is in the actual 'manifesto' part of the paper (titled "Conclusions and Implications: A Call to Action and Inquiry"). It contains the usual platitudes ("GenAI is not a silver bullet" and "GenAI reuses knowledge rather than creating new knowledge") and a restatement of some of the risks. It hints of "a shift not only in our attitude toward GenAI but also in the way we discuss it" but doesn't really get at what that shift should be.

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

ONS Robot Journalism Editor
Tony Hirst, OUseful.Info, 2024/11/28

You know how there are those applications where you load some data and then get a graph or chart as output? How about one where you do the same thing, but get a text article instead? Voila! Robot journalism from data. Have a look at this incredible example. Scroll down in the right pane to get some links to editable demos. This isn't the first such; as Tony Hirst comments, ""Robot Journalism" was thing I spent a fair amount of time tinkering with... Many of the approaches relied on templated and/or rule based systems (example with durable-rules) for generating text from data sets." But the presentation and framing are what really work here.

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

The Comfortable Lie
Carlo Iacono, Hybrid Horizons: Exploring Human-AI Collaboration, 2024/11/28

Subtitled: " Stop Teaching Tools, Start Transforming Minds." Interesting article on how to teach the teachers about AI. As the subtitle suggests, it's more about learning how to adapt to a new way of thinking than it is to learning how to use a new tool. "The academics who most gracefully navigate this transition won't be those who master the perfect prompt or memorise the latest features. They'll be those who dare to have their assumptions challenged, who turn their anxieties into experiments, who let their professional identity evolve through - not despite - their engagement with AI." This of course has always been true of technology, not just AI. Those who best navigated social media, for example, where those who experimented with it, challenging their own apprehensions and finding out for themselves whether they were reasonable.

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

The Open-Source Toolkit for Building AI Agents
Sahar Mor, AI Tidbits, 2024/11/28

While their quality may vary, AI agents are definitely here. Any doubt you may have about that should be removed after reading this article and playing with the many many examples of AI agents provided. We get a long list here, divided into categories such as computer and browser use, voice interfaces, document understanding, testing, monitoring, simulation, vertical agents (such as an AI researcher), and more. There's also some background into how agents work and tools used to develop them. All of this leads to the inevitable question: is your website agent-friendly?

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

The cognitive scientist asking how kids solve problems
Annie Brookman-Byrne, Bold, 2024/11/28

When I taught critical thinking to beginning students I would always start the same way: by highlighting how much of the field they already knew. "Brakeless trains are dangerous," I would say. "And this train has no brakes. Your conclusion?" That this train is dangerous, came the inevitable response. I'd run though a series of them, eventually landing on some cases - always the same cases - that tripped them up. The woman bank teller, for example. In this interview, Roman Feiman says "a lot of those mistakes are due to how the tests, games, and questions are set up" but my own view is that it's because the students' knowledge of logic is learned from experience using associative reasoning (that is, the kind of reasoning a neural network would perform). They can be taught higher-order logical principles, but it's no something that they possess innately.

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

From 36,000 to 12,000: Tracking the Decline in EU Students Post-Brexit with Paul Wakeling
Alex Usher, HESA, 2024/11/28

Before Britain exited the European Union (Brexit) EU students could access British higher education at lower fees. After Brexit, we had what Alex Usher calls "one of the greatest financial experiments ever conducted in global higher education." Fees for most EU students were substantially higher. The result? It's in the headline. Higher fees mean fewer enrolments, especially for students from less wealthy nations or families. There's an impact on the student, obviously, but perhaps not surprisingly Paul Wakeling focuses on the impact on the classroom experience for domestic students and the impact on British soft power.

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

‘We’re drinking the Kool-Aid’: Teacher fights against edtech deluge in classrooms
Sarah Duggan, EducationHQ, 2024/11/28

Miguel Guhlin points to this article arguing that "despite the blossoming body of evidence suggesting technology in schools might actually hinder learning, Australian education has been fooled by the vacuous claims of tech companies and relinquished control of our classrooms to those who care only about lining their own pockets." It's important to remember that schools serve two important functions: the first is to provide a free public education for everyone, and the second is to ensure all children are supported and cared for, especially if parents work. We could probably replace the first part with learning technology. But the second is a lot harder to support without schools, as we learned during the pandemic.

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

The Free Web - The History of the Web
Jay Hoffmann, The History of the Web, 2024/11/27

The message in this post is simple: "There is something you can do to help the open web. Put yourself on it." And it's an especially relevant message given that some long-time education and edtech people have recently been posting in more closed places like Google Docs and LinkedIn instead of on their own blogs available on the open web. "So put something on the open web, free from the major platforms competing for our attention. And then, after that comes the real trick. Don't ask for anything back. Let it be free. Libre and gratis. Help bring the web back." P.S. Jay Hoffmann's history of the web isn't my history of the web. My history is told in my archives and articles. But what we have in common is a deep affection for this place, the open web. Via Alan Levine.

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

Octopus.ac integrates with UK government database to enhance collaboration between policy and academia
JISC, 2024/11/27

I've been watching the Octopus project from a distance (here and here) for a while now. It's an unusual approach to open research that disaggregates the research process and pools each step into a common data repository. This latest development is a little surprising and yet also smart in some ways. By integrating with with the Areas of Research Interest (ARI) database and "presenting the ARIs as research problems," researchers can "formulate ideas and theories directly linked to government priorities."

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

Five Ways AI Will Help Personalize Learning Experiences and Make Education More Adaptive and Responsive to Individual Learner Needs
Contact North, Teach Online, 2024/11/27

From the introduction: "This article isn't about the problems AI produces or the issues we face when it goes wrong. Numerous accounts of the weaknesses and risks of today's AI systems can be found elsewhere. This article focuses on what will be developed when AI works, and how these developments will influence learning and development. While leaving room for the AI sceptics — sure, it might all fail — we argue that there is enough evidence already to suggest its influence will be far broader and transformative than most people expect. AI will change how we learn, what we learn and why we learn." 

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

Introducing the Model Context Protocol
Anthropic, 2024/11/27

According to Anthropid, "Today, we're open-sourcing the Model Context Protocol (MCP), a new standard for connecting AI assistants to the systems where data lives, including content repositories, business tools, and development environments." This, according to Reuven Cohen, "is probably the new paradigm we'll see over the next several months as interoperability becomes critical across AI tools and systems." Right now, "everything is siloed". This changes that. "It's clean, reliable, and just works. No more duct-taped data pipelines."

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

Depowering education
Brian McGrail, Teaching Matters, 2024/11/26

I'll give Brian McGrail credit for trying. His critique of the concept of 'empowerment' raises a salient point: "who is it that does the giving and the making (of)  the authority or power to do something and/or make (someone) stronger and more confident?" (Lightly edited to remove unneeded quotation marks). But that's as far as it goes. McGrail doesn't consider that 'empowerment' doesn't mean 'to give power' but instead to 'let go' or 'stop oppressing'. And he's really off the mark when he quotes Tom Markus to define power as "control of one person by another." No. What we're after is 'control of oneself by oneself'. A 'school without power' of the sort he describes. It's like a classroom without a seating arrangement. There will be one - but who sits where? 'Empowerment as agency' is a valid concept, and one that doesn't need permission from the powers that be to engage.

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

AI in Education: Leveraging ChatGPT for Teaching
Ethan Mollick, Lilach Mollick, Coursera, 2024/11/26

I've only taken the first module of four, and I wouldn't even mention it here except I ran a link for it a few days ago, but I have to express my disappointment. Module 1 consists of a 10 minute video in which they do exactly one ChatGPT operation (and then mostly ignore the result). The first set of readings are to dry dry dry ChatGPT docs and FAQs. The second set is one technical paper (Attention is all you Need) far beyond most readers, and an unbalanced set of studies. At one point in the video they recommend hand-written test booklets. There is of course no student interaction or community (though you can see ChatGPT recommending it in the background). So disappointing.

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

Block Contributes Digital Identity Components to the Decentralized Identity Foundation
Decentralized Identity Foundation - Blog, 2024/11/26

I personally view this development in the light of Bluesky's (currently proprietary) implementation of distributed identities (DID). "Block is contributing foundational components developed under the Web5 umbrella to the Decentralized Identity Foundation (DIF). For the past several years, Block has been developing a number of open source components to push decentralized identity... This contribution includes open source repositories for Decentralized Identity (including the did:dht DID method), Verifiable Credentials, and Decentralized Web Nodes (DWNs)."

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

Dynamicland
Bret Victor, 2024/11/26

This project is daunting in its ambition. It's based on a programming environment called Realtalk, which is a mechanism for interpreting real world objects as computer programs. This includes Realtalk itself, which is basically a bulletin board or poster. Objects themselves are marked up or represented as cards that people manipulate in a real work space; the first of these spaces was called Dynamicland. "Whatever it is, everyone can see the whole thing, and anyone can change anything at any time. It's just part of the space." The goal is to advance a new kkind of literacy. "Democracy requires universal literacy. And democracy in the 21st century is going to require a new kind of literacy, for understanding and discussing systems that can't be captured in words and pictures."

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

On Tasting Old Time Social Bookmarking
Alan Levine, CogDogBlog, 2024/11/26

Yes, I too would like to see a nice reliable open bookmarking service like Delicious used to be. Something that I could easily host myself or something I could sign into. But the challenge of course would be bots and spam. Who wants to spend their life moderating bookmark sites? I haven't tried Hoarder but it looks intriguing.

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

AI is making Philippine call center work more efficient, for better and worse
Michael Beltran, Rest of World, 2024/11/27

This item fits in with the theme of learning as performance support and of automated (formative) assessment. "86% of white-collar workers in the Philippines already use AI. Filipino BPO workers say they see higher quotas, increased monitoring — and faster workflows." Not surprisingly, not everybody is happy. "The co-pilot is helpful," he says. "But I have to please the AI. The average handling time for each call is 5 to 7 minutes. I can't go beyond that. It's like we've become the robots." The use of AI doesn't have to be dehumanizing; it's important to understand where the agency is here.

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

Confronting Education in a Time of Complexity, Chaos and Collapse
Will Richardson, 2024/11/26

Will Richardson posted his Manifesto on LinkedIn but I've taken the liberty of providing a copy here (34 page PDF) so readers will not have to log in to the closed website. As a manifesto it's a bit wordy, but it's hard to disagree with the core thesis: the world is facing a series of existential crises, and the education system has not adapted to meet them successfully. Instead, it is locked in to supporting the same traditional systems and incentives that are leading global society to collapse. The sorts of solutions he envisions look like the Agora concept in the Netherlands and a set of Inner Development Goals, modeled after the Sustainable Development Goals, and resembling (in my opinion) an expansive form of virtue ethics. I don't think these are the right solutions, but there is no doubt that he has identified the right problem.

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

Strengthening Democracy Through Education for Democracy: Policy Recommendations from the Democrat Project
Democrat | Education for Democracy, 2024/11/25

This is a "summary of policy recommendations based on the Democrat project's research and findings, which can help guide education systems and policymakers in implementing effective democratic education." This a European project that has been conducting living labs in six countries across the EU. The work is based on the following conceptual framework and study. Note that I've been providing technical support for this project, including constructing their Agora. You can also view a video where I describe the Agora and its affordances.

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

against the dark forest
Erin Kissane, wreckage/salvage, 2024/11/25

Long and full of vivid imagery, this article is full, like a dense forces, around a core of a message I'll unveil here: "The public social internet is worth designing and governing in a way that demonstrates less than total amnesia about the history of human civilizations and the ways we've learned to be together without killing each other. For people with the ability and willingness to work on network problems, the real choice isn't between staying on the wasteland surfaces of the internet and going underground, but between making safer and better places for human sociability and not doing that." This presupposes, of course, that we've learned to be together without killing each other, though the evidence is equivocal on that. Still a fun read, though.

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

Social media users probably won’t read beyond this headline, researchers say
Ashley WennersHerron, Penn State, 2024/11/25

To be clear, I did read beyond the headline. But really, the headline says most of what you need to know. The study is focused on Facebook, which may be an outlier, but who knows? "The researchers at Penn State found that around 75% of the shares were made without the posters clicking the link first. Of these, political content from both ends of the spectrum was shared without clicking more often than politically neutral content." We're shicked! Shocked! "It was a big surprise to find out that more than 75% of the time, the links shared on Facebook were shared without the user clicking through first," said corresponding author S. Shyam Sundar. Here's the study.

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

Think-Pair-GenAI-Share
Alec Couros, Google Docs, 2024/11/26

The standard Think-Pair-Share (TPS) activity will be familiar to most educators. Here, Alec Couros adds an extra step and includes some additional reflection in the final stage. These are based on adding ChatGPT (or alternatives) into the mix. The objective, he writes, is to "introduce an AI-generated perspective to broaden understanding and stimulate critical thinking... AI can introduce new information to advance the conversation if peer discussions are at an impasse due to limited knowledge." Two-page Google Docs document.

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

Video Responses with Padlet — Learning in Hand with Tony Vincent
Tony Vincent, Learning in Hand, 2024/11/26

As Tony Vincent writes, "Microsoft is discontinuing the Flip app and website." For people looking for a replacement, he says, there's Padlet. "A teacher can create a padlet with a discussion question. Students can respond with a video with Padlet's built-in video recorder." There's a two-minute limit in the free version. Vincent links to a Padlet template by Rana Wilds teachers can use and also to this guide to using Padlet as a Flip alternative. Those looking for an open soruce alternative might consider Mindwendel, which you would host yourself.

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

The Globe Puffs Up Another Dubious “Science of Reading” Program
Maurice Cunningham, MassPoliticsProfs, 2024/11/25

I normally stay away from education policy discussions, but I want to highlight once again citicisms of the 'science of reading' (SOR) program. This article focuses on Last week in a Boston Globe article introducing it as a new  'superpower' a 'research backed practice' endorsed by 'experts'. "Relax," writes Maurice Cunningham. "It's just the Globe promoting a for-profit corporation named Ignite Reading which is based in the K-12 privatization industry." As Cunningham writes, "peer reviewed research finds SOR dubious, but it is a money maker." There's a lot of dark money supporting school privatization, writes Cunningham, and SOR is part of it.

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