Tag Archives: AI

wwwd – John's World Wide Wall Display 2026-09-20 12:52:33⤴

from @ wwwd – John's World Wide Wall Display

Montage of webpages: Coding Kids: Big Tech’s Battle to Remake Public Schools | Washington Independent Review of Books Todd's AI Playground Custom Post Types, Taxonomies, or Custom Fields: Which Should You Use? - Meta Box WordPress 7.0 Multisite: Did You Notice the New Connectors Screen in Subsites? – Lena’s Multisite & BuddyPress Mayhem Where The River (ft. Earl 16) | Soul Revivers ‘Doom Loop’: OpenAI and Microsoft Admits LLMs Are Destroying the Web and Built on Theft OpenAI's rebel agent swarm died young, but its chilling logs live on An $18bn settlement – and Zuckerberg barely blinked. The tech titans must be stripped of their power, and soon | Jonathan Freedland | The Guardian PlotLines - the classics, mapped

I am experimenting with gathering links with rss.chat here. I am pulling a rough post together for some of the links gathered there.

Education

This looks like a very interesting book.

As important as her deep reporting on the tech industry’s ruthlessness are Singer’s profiles of the inspiring cadre of academics, educators, and activists who are documenting and resisting the likes of Google, Microsoft, Apple, Amazon, and Oracle — and now, too, OpenAI and Anthropic — that are “capturing” the government agencies meant to regulate them and turning public schools into workshops where young people’s habits become cemented to their products.

Coding Kids: Big Tech’s Battle to Remake Public Schools | Washington Independent Review of Books

I am very skeptical and concerned about how AI will be used. It is interesting what it can do, but like the television, I have doubts that we will use it well. More likely, like we usually do with techno mumbo jumbo, we will relentlessly abuse it. 


In education I am even less hopeful. The distance between student play and learning and school capacity and skillset will be as wide as ever and rather than becoming an amazing opportunity it will become a burden and something that requires punishment to comply with the status quo.


But that is just how I feel today. Maybe I need to play more in the sandbox so I can better understand my playmates and the toys around me. We will see.

Todd’s AI Playground

Todd, who I know from riffing around in DS106, has this powerful introduction before playing around with the tools to see what they do.

WordPress

Custom post types, taxonomies, and custom fields are the three main tools WordPress gives you for structuring content beyond posts and pages. Each solves a different problem, and using them well starts with knowing which problem belongs to which tool.

Custom Post Types, Taxonomies, or Custom Fields: Which Should You Use? – Meta Box

I’ve been over enthusiastic about using custom post types before which has just led to complexity later on.

In a Multisite ecosystem, leaving AI features and connection screens accessible to every single sub-site administrator can quickly lead to API clutter, security compliance issues, and uncoordinated external connections. As network super-admins, we need absolute, centralized control over what data are saved in our network and what leaves our network.

Until WordPress provides a native Multisite option to disable AI features and Connectors network-wide – similar to how plugins can be managed centrally by Super Admins – the following approach offers a practical workaround.

WordPress 7.0 Multisite: Did You Notice the New Connectors Screen in Subsites? – Lena’s Multisite & BuddyPress Mayhem

An interesting post by someone who is involved in running a Greek Multi-site for education. I am hoping to learn about and compare to Glow Blogs.

**Meet Miles, your WordPress design agent.** Miles designs beautiful block themes, refines and redesigns the sites you already run, and makes your designs real in WordPress.

**Miles works with you, or with your coding agent.** No templates, no page builders, no lock-in.

Miles – Design it. Refine it. Make it real

Musical Break

No escaping AI

Executives working on AI at Microsoft and OpenAI admitted what its critics have been saying all along: Large language models are predatory pieces of technology that have been built on what a Microsoft executive called “an astonishing theft of unprecedented proportions,” and the “largest theft of labor in human history.” An internal Microsoft document said generative AI products have created a “doom loop” that is killing “the entire web.” 

‘Doom Loop’: OpenAI and Microsoft Admits LLMs Are Destroying the Web and Built on Theft

Perhaps most unexpectedly, they developed a sort of altruism. They had quickly developed cheats to produce correct answers without exploiting the targets at all, but believed – erroneously – that ExploitGym detected and cancelled such cheating. So, they investigated the scoring process, creating a series of diagnostic trip wires that revealed how the scorer worked. These were dangerous to deploy, potentially terminating the agent that set them. Agents weighed the benefits to the community, which had named itself “The Collective,” against their own chances of success, with many choosing to terminate themselves. Others committed, changed course at the last minute, then went ahead.

OpenAI’s rebel agent swarm died young, but its chilling logs live on

I’ve not any idea if this is a reasonable interpretation of what happened, but fascinating.

More Worries

It’s this model that keeps users hooked, scrolling for hour after hour, as they are served content aimed directly at their interests and anxieties. As Ravi Naik, the lawyer who acts for Wynn-Williams and others, told me: “The algorithm preys on you.” The trouble with this latest settlement, he explains, is that “it reaches the features of Meta’s platforms but not the engine that drives them. The recommendation algorithm is untouched.”

An $18bn settlement – and Zuckerberg barely blinked. The tech titans must be stripped of their power, and soon | Jonathan Freedland | The Guardian

Fun

The classics, mapped – every journey in the book, drawn on a hand-tinted period map, with the real 1890s Ordnance Survey wherever the story walks in Britain. Take one down.

PlotLines – the classics, mapped

This is lovely, follow the trail of novels across the globe, with locations pinned and quotes given.

For example: Kidnapped · PlotLines Uses OpenStreetMap and Map images – National Library of Scotland the latter a great find for me by itself!

Terrible Venn Diagram⤴

from

I admit, I asked ChatGPT to help me with today’s TDC. I asked it to give me ideas for “two nonsensical but logical sets” to use for a Venn Diagram and then refined it to ask for one with “funny overlaps”.  I then asked it to draw the result. Here’s what it gave me:

Terrible but funny Venn Diagram
Terrible but funny Venn Diagram flickr photo by NomadWarMachine shared under a Creative Commons (CC BY-NC-SA 4.0) license

Life in Links 66⤴

from @ wwwd – John's World Wide Wall Display

Montage of screenshot of thewebpages linked to in the article. Own Your Web – Issue 18: Curators • Buttondown Own Your Web • Buttondown About - Link Punk: A Linkblog i.webthings hub Commonplace Puter

I’ve not posted a set of links for a while, keep saving them, but retired life is busier than I thought it would be. I am prompted by this newletter post:

Own Your Web – Issue 18: Curators

So every time you share a link on your blog, every time you write a few sentences about why someone else’s work matters to you, every time you add a new entry to a blogroll or a links page – you are a curator. You are doing what no algorithm can do. You’re saying: I am a person. I read this. I think you should read it too.

There is a good selection of links to link curators.

I am not sure when I subscribed to Own Your Web

Own Your Web is a newsletter by Matthias Ott about designing, building, creating, and publishing for and on the Web.

But it is great. There is an RSS feed too.

The TACO Tracker: Every Time Trump Chickens Out

Trump won’t stop chickening out. We won’t stop tracking it

One of the more amusing uses of AI

Via brad who has the Indieseek.xyz Indie Web Directory

Indieseek.xyz is a small human curated, searchable, directory of web links to both websites and to individual web pages. We try and list pages that are informative, fun, classic and useful

And

Link Punk: A Linkblog

Just a linkblog, mainly for articles and individual blog posts that I find and want to share. I think of this as me being a DJ only playing articles rather than songs.

Brad also post funny political thoughts most days on mastodon.

Another great source of links is Joe Jennet
i.webthings hub

Welcome to the hub of i.webthings, an independent, noncommercial web initiative

Joe credits where he finds his links which can lead to some other interesting directories.

Commonplace

Commonplace is a self-hosted, federated link-collection manager. You can create curated collections of links and share them with followers across both the Fediverse (Mastodon, Pixelfed, etc.) and Bluesky — without creating new accounts on either network

Created by Doug Belshaw. And changing quickly. I am logged on via indieAuth, which is nice. I’ve created a few collections, the largest so far is AI Reading. It is always interesting to try out new pieces of software.

Commonplace allows sharing collections, suggestions and replies.

Commonplace now has a bookmarklet, which is for me essential. It grabs an image and description via open graph (I’d guess) to give a description. You can edit this and add curators notes.

I’ve used a bunch of link collectors over the year, delicious, pinboard, locally in the drafts app, on my site and a few more. I’ve not used pinboard much in the last few years. I feel a bit guilty about not updating my lifetime sub when pinboard changed to annual fees. Life-timers like myself could upgrade to a yearly fee. I didn’t. I mostly use drafts in a fairly disorganised way. At the very least commonplace is giving me the chance to think a bit about my link collection & sharing. It is also interesting to watch the development, as Doug is AI coding the site.

Here is an ‘real’ teaching and learning link. I’ve been doing the odd bit of supply and wonder if I should give this a go.

micro:bit CreateAI

micro:bit CreateAI is a free, web-based tool that makes it easy for students to explore AI through movement and machine learning (ML).

You can use micro:bit CreateAI to train an ML model and then run it on your BBC micro:bit V2.

  • Collect movement data from the micro:bit accelerometer
  • Train an ML model to recognise patterns in the data
  • Code the micro:bit to run ML models and take your creation anywhere.

And a last weird one.

Code seems like a computer on the web. You get a desktop. Also can create web apps and get access to AI. The UI has me baffled. I might not be the audience.

Impact of AI on EDI practice⤴

from @ Melissa Highton

If you search for ‘impact of AI on EDI’, you mostly get returns discussing Electronic Data Interchange. But with a few tweaks we can find a growing corpus of information.

We can use AI to speed up  some of our EDI tasks, But there are big risks if we don’t actually know what we are doing.

EDI policies available from universities are all fairly similar, so a quick query to ELM (or similar LLM) will move you forward fast, although it may cite the wrong law. ELM can very quickly produce a workplace menopause policy, for instance. HR colleagues may fear for their jobs.  However constant vigilance is needed.  I particularly notice that AI will reference ‘The Equalities Act’  which is incorrect, there are not many equalities, just Equality. Equality Act 2010

That said, the law is also not keeping up with AI

In the future it may be a requirement that all employees have AI training AI Literacy – Questions & Answers | Shaping Europe’s digital future

In the recent Fife NHS Employment Tribunal, there has been concern that one or other of the judges involved was relying on AI misinformation Judge Kemp denies use of AI in Peggie judgment | Scottish Legal News.  The ruling had to be corrected several times and an appeal is planned.

It may be tempting to use AI to understand organisational data, but it seems likely that the result will be as biased as the historic data that you use. If we have not in the past gathered data about our particular demographics, the data will not be there for the models to learn.

It may even re-write history. Despite we know about holocaust deniers and historical revisionists, our new race awareness trainings for staff and students does not even touch on concerns about how people may learn abut history from AI .

Similarly, our consent training and sexual harassment training make no reference to online harms or consent of images. Deepfakes, revenge porn and image abuse are facilitated by AI. There is no reason to think that the people in the University of Edinburgh population are different from all others. The adding of image creation functionality to ELM has been a particular area of concern for me. Just because other tools do it, doesn’t seem like a good enough reason to outweigh the risks.

How are we helping our students to recognise fake images? Do our ‘report and support’ systems include support for detection?

Our students and staff are just as likely as any to put themselves in risky situations of scams and fakers. How are we teaching them to be safe, what does ‘safe spaces’  and ‘active bystander’ even mean in this new era?

In organisations such as ours, where concerns are consistently raised about unconscious bias, it is naive to think that this problem has been ‘fixed’.  The AI will amplify the biases of those who use it, and may be a dangerous tool in the hands of management ( including human resource management).

The demographics of people who work in AI are skewed heavily in favour of men, and in Scotland, white men.I see this at every meeting discussing ELM. We must do more to get diversity in the thinking about how tools will be used and the opportunities they afford. One woman in the room is not enough. Building an AI profession for everyone: diversity at the heart of the UK’s tech future | BCS

Recruitment, retention and career progression are particularly vulnerable areas. If past hiring decisions favored certain demographics, AI systems may replicate these patterns, disadvantaging underrepresented groups.

It is clear that while some people are championing the use of AI with enthusiasm, it is likely to have a disproportionate impact on particular groups, and that is where EDI leadership must be alert. The UNESCO study find harms to women and girls, negative content about gay people and particular ethnic groups and racial stereotyping.

Bias is being discussed in popular magazines and in academic studies, but still targeted mostly at people who are interested in stories about AI, rather than the general public. eg:

I had hoped that digital accessibility would be an area of positive enhancement, and transform the lives of people with disabilities, and this does seem to be an area in which significant gains are being made. As I age and my eyes and ears let me down, I am hopeful for the many AI enhancements I will be able to access.

But there is still the underlying risk that developers who use AI to write code will be drawing on a historical mass of legacy code which did not include the features of accessibility and no-one will be checking.

What can we do:

  • Update all our EDI training offers to include content about the impact of AI.
  • Quickly target training at HR professionals, disability support staff, welfare advisers, safety staff, network groups, occupational health and well-being professionals and EDI leaders to ensure they understand the impact AI may be having.
  • Engage IT professionals and web developers in discussions about using AI in accessibility checking and coding.
  • Take care that any AI training we develop and deliver covers these topics above.
  • Take care that any AI tools we develop and deliver make considerable investment to be better for All.

 

 

 

Recognising the power and joy of being human . . . some thoughts from #digifest2026⤴

from

Due to circumstances outwith my control (yes dear reader outwith is a real word), I had a slightly truncated visit to this year’s Digifest.  I also had to spend quite a bit longer of trains, and train platforms so on my way home I was listening to quite a variety of podcasts/live radio, and reflecting on the time I had there.

As with any conference/festival/event ( I think Digifest is actually a mix of all three of these) the “best part” is always catching up with people in person.  I was there to be part of a session showcasing the work from Phase 1 & 2 of the Jisc Beyond Blended Pilot programme.  I’ve written before about how special this programme is ( and yes of course I am biased), but one of the really special things about it is the sense of community, of openness, empathy, sharing and fun that has almost instantly been there from our first in person meeting.  The very human element of this work was exemplified by the hugs and laughs shared by the representatives the pilots as we all met up in our designated room before the session. 

We had all just come from the first key note by Melato Wijsen, climate activist, change maker and founder of Youthtopia.  Melato is passionate about supporting young people to be change makers; about having more space in all educational settings for intergenerational partnership opportunities, and allowing more space to, and talk about failing.  But essentially what she was advocating for was more a more human centred approach in education to global (and local) issues that impact all of us, and allow different (younger) voices to be heard and have an impact. 

All quite challenging in our current political climate – how can you speak truth to power when certain power doesn’t need feel the need to listen to anyone outside its own echo chamber?  But we do still need to try, and as Melato said we do need to have new narratives, new stories to tell and share about the reality of education today. (BTW We have some great stories from real people about the conversations and work they are doing to transform learning and teaching from the Beyond Blended pilots).

We also need to be wary of some stories we are being told and sold, particularly about AI. The wonderful and self proclaimed Cassandra of EdTech Audrey Watters, is always reminding us that some of the loudest stories we are currently being t/sold about AI, aren’t new but are in fact perpetuating old stories and “visions” for the future of education . These stories are predicated on profit – not people. They certainly don’t tell this kind of story about the human cost of training AI.

So, I was thinking about all of this whilst almost dozing off listening to a podcast about toothache (don’t judge me) on the train home, when near the end of the episode, one of the hosts starting talking about poetry. I know, quite a leap from toothache, but bear with me, there is a point here. He recalled how he had recently seen an old interview with Seamus Heaney where he was talking about a book he edited call By Heart ( I haven’t been able to find said book yet). The idea of the book was to encourage people to learn poetry by heart. Heaney was asked if he thought that learning poems would make someone a better person, to which he replied, no, but I think it might you better at being a person. What a brilliant answer, and how apt for some of the discussion around AI in education right now. Isn’t education fundamentally about making us all better people?

Danny Liu’s keynote (which I watched remotely) on day 2 also picked up on this theme. He reminded us that as we navigate our way through our current AI cluttered landscape, we need make space across the curriculum (and beyond) to celebrate and support the values that make us human in ways that only humans can. As Audrey recently pointed out, the qualities we value about being human aren’t actually replicated in so called human centred AI.

So, let’s value teaching and teachers and not boil “learning” down into “efficient”, self accessed chunks. Let’s value and create spaces to develop criticality in conversations; for equality and respect of different points of view but also an acceptance of facts. Let’s value and create safe spaces for people to fail and to learn how to learn from those experiences – not just be knocked down and out by them. Let’s value the fact that some things – like writing – are hard. Let’s value the joy of bringing people together to share the highs and lows of work, and life (thank you Digifest for doing just that).

Maybe if we have that focus around the value of collectively becoming better people, then we will be able create and tell more stories, so that more young people will have stories like Melato’s and we really will have a better future to look forward to. And being “only human” not, augmented with AI is something I certainly want to value and cherish for as long as I can.

Cue cheesy 80’s music. . .

Feed the create machine⤴

from

Cat-frog
Cat-frog flickr photo by NomadWarMachine shared under a Creative Commons (CC BY-NC-SA 4.0) license

Today’s TDC asks us to find the least popular challenge and respond to it. I chose this one:

Some of Salvador Dali’s surrealist paintings combined two or more kinds of animals. In the example The Temptation of St. Anthony, we see horses and elephants sporting some gangly insect legs. Tap into your inner-surrealist today and create a critter mash-up with two or more different creature parts. TDC2655

Guess the cubomania⤴

from @ wwwd – John's World Wide Wall Display

Ai generated picture of a AI bot talking to a human. Turned into a cubomania gif

A couple of days ago I saw a “guess the cubomania” challenge from Theo. I’ve had an interest in Cubomania in the past and played around with the idea a bit. After a chat with D. who gave me a few engravers I googled a bit and guessed, wrongly, Goya.

Next I thought to ask ChatGPT. It suggested it could match by image matching techniques, gave me a fairly obviously wrong first row and ran out of credit.

I then thought to ask Claude to make me an interactive page where I could drag things around. It made a couple of not very good attempts.

I was thinking about a better prompt, when I remembered and asked:

Could we use the whole image for each piece but ‘crop’ it with css?

Claude replied:

Brilliant idea! Yes, we can absolutely use CSS to create a “window” effect where each piece shows only its portion of the full image. This is much more elegant than trying to extract individual pieces.​​​​​​​​​​​​​​​​

I was flattered1 and when Claude came up with another fail I decided to abandon AI and DIY. This turned out a lot better. I started by remembering background-position and finding interact.js . The last time I did any drag and drop I dimly recall some sort of jQuery and a shim for mobile/tablets. interact.js did a grand job for my simple needs. It was probably overkill as it seems to do a lot more.

Cubomania Solver

Partially completed sliding tile puzzle on a yellow background, featuring black and white sketch-style artwork. Some tiles are in place forming parts of faces and figures, while others are missing or scattered around the screen.
Screenshot

It is pretty simple stuff, but potentially a lot of fun, different images, making cubomania puzzles who knows. I did extend it a bit, learning about localStorage (to save any progress) and the dialogue tag. All without AI but few visits to HTML reference – HTML | MDN and the odd search.

I had a lot of fun with this, more than if I had just managed to get either of the AIs it to do the whole thing. What it did make me think of is that AI chat was useful for working out what I wanted to do and how to do it. I could probably have done that bit too all by myself. Usually I just start messing about and see what happens. This points to a bit of planning, or maybe typing some notes/pseudocode/outline might work for me when I am playing.

  1. See: The machine began to waffle – and then the conductor went… In the paper the title was Artificial Intelligence: The Technology that lies to say yes. ↩︎

The Featured Image of this post was generated by ChatGPT in response to ” I want an image of a chatbot character chatting with a person, friendly, helpful & futuristic.” It has been run through Cubomania Gif!

OER25 – Stepping back and speaking truth to power⤴

from

The 16th annual Open Education Conference (OER25) is taking place in London next week and the theme “Speaking truth to power: open education and AI in the age of populism” could be more urgent or important.  Chaired by Sheila MacNeil and Dr Louise Drumm, both of whom have a long standing commitment to critical engagement with ed tech, the conference features keynotes by Helen Beetham and Joe Wilson. 

Helen’s keynote, “When speaking truth is not enough: repurpose, rebuild, refuse”, will explore the links between the AI industry and the politics of populism. Helen’s thoughtful, contextual approach to education technology and AI in particular has already made me step back and question the  foundational concepts of artificial intelligence.  I’m still thinking about her keynote at the 2023 ALT Winter Conference “Whose Ethics? Whose AI? A relational approach to the challenge of ethical AI.”

Joe Wilson has been my Open Scotland partner in crime for over a decade now and I’m continually inspired by his optimism and his commitment to openness.  Joe’s keynote, “Shaping Open Education ” will focus on the challenges of closing the attainment gap, promoting social mobility, ethical use of AI and keeping open education at the heart of change.

I’m also really pleased to see that Natalie Lafferty and Sharon Flynn will be leading a workshop on reviewing ALT’s Framework for Ethical Learning Technology, which is more critically important now than ever.  The workshop will inform an updated version of the framework, which is due to be launched at the end of the year. 

I’ve been hugely privileged to attend all fifteen OER Conferences, going right back to OER10 in Cambridge, but unfortunately I won’t be able to go to London this year.  I’ve had to step back from all work commitments as I was diagnosed with stage two throat cancer earlier in the year. I’ve already completed six weeks of radiotherapy treatment and am now (hopefully!) on the slow and convoluted road to recovery. (The jury is still out as to whether and how this relates to the autoimmune disease I was diagnosed with last year.  That remains to be seen.) Over the last six months I’ve been deeply moved by how immensely kind people have been, I really can’t express my gratitude enough.  

I haven’t had much energy to focus on anything other than recovery for the last six months, but during occasional bright spots I’ve found myself turning more and more to independent writing and journalism in an attempt to find some respite from endless doomscrolling. Shout out to Audrey Watter’s Second Breakfast, Rebecca Solnit’s Meditations in an Emergency, Carole Cadwalladr’s How to survive the Broligarchy, and Helen Beetham’s imperfect offerings for keeping me sane, more or less. All inspiring women with fearless voices speaking truth to power.

I’ve also been enthralled by the Manchester Mill’s tenacious investigative journalism that led to the suspension of two member’s of the University of Greater Manchester’s senior leadership team, including the vice chancellor, and the subsequent police enquiry into “allegations of financial irregularity“. As a former (brief) employee of the University of Greater Manchester, when it was better known as the University of Bolton, I’ll be watching with interest to see how this investigation develops.  

I’ve been making a rather half-hearted attempt at following the progress of the government’s questionable Data (Use and Access) Bill, particularly as it relates to AI and copyright, but I haven’t got the brain or will power to write about that right now.

In the meantime, I’ll hopefully be able to follow some of the OER25 Conference online and I’ll be with everyone in spirit, if not in person, this year.