I love …⤴

from

Inspired by today’s Daily Create based on Kim Douillard’s blog post.

I love. I love to sit and knit on a winter evenings, sitting on the sofa in front of the TV, with my legs wrapped up in a Shetland wool hap.

Shetland Lace Knitting
Shetland Lace Knitting flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

I love my cats. especially the way my Lacey cat will insist I put her favourite cushion on my lap so she can climb up and sit on my lap.

310 Lap Cat
310 Lap Cat flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

I love living in Scotland, with such beautiful scenery all around me. Having Mugdock Loch so close to where we live brings me peace when I need it.

69/365 Heron
69/365 Heron flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

I love the Daily Create, for making me think every day, and this blog, for giving me space to reflect.

What do you love?

 

A Glasgow Banksy⤴

from

Since I’ve been taking a photo each day to upload to Flickr (for each day in 2024 and every one so far in 2025) I’ve been looking at my surroundings more carefully and noticing things that I have not seen before. So when I walked up Great George Street recently for the first time in a long time I was pleased to see some street art on the wall.

78/365 Great George St Graffiti
78/365 Great George St Graffiti flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

When I used Google Lens to find out more about it I discovered that it was thought to be by a local artist called the Rebel Bear – otherwise known as the Scottish Banksy. It had only appeared in January 2025, so no wonder I’d not noticed it before.

There’s a parking sign in the other online photos though, which is not in mine. I wonder what happened to that.

wwwd – John's World Wide Wall Display 2025-03-24 15:28:18⤴

from @ wwwd – John's World Wide Wall Display

Likes “Wait, not like that”: Free and open access in the age of generative AI by Molly White.

Irresponsible AI companies are already imposing huge loads on Wikimedia infrastructure, which is costly both from a pure bandwidth perspective, but also because it requires dedicated engineers to maintain and improve systems to handle the massive automated traffic. And AI companies that do not attribute their responses or otherwise provide any pointers back to Wikipedia prevent users from knowing where that material came from, and do not encourage those users to go visit Wikipedia, where they might then sign up as an editor, or donate after seeing a request for support. (This is most AI companies, by the way. Many AI “visionaries” seem perfectly content to promise that artificial superintelligence is just around the corner, but claim that attribution is somehow a permanently unsolvable problem.)

A good post to read or listen to at the beginning of  Scottish AI in Schools Week article does not want the stable door closed.

Some AI Links & Things to Think About⤴

from @ wwwd – John's World Wide Wall Display

A montage of 4 webpage screenshots.

Bookmarked for future reading. AI in education is becoming increasingly confusing.

Education Scotland are running a week #ScotAI25: Scottish AI in Schools 2025 with live lessons for pupils & some cpd for staff. I might try to make some of those.

  • This week I’ve used:
    ChatGPT to make some questions up about a passage of text for an individual in my class; Write an example text about levers; create a formula for a number spreadsheet and create a regular expression.
  • Claude to make a fractions matching game and a trivia quiz.
  • I am occasionally using lovable.dev to play around making an alternative way of posting to WordPress.

I might have used ChatGPT a couple more times in school. Although it is accessible the login options didn’t seem to be so I’ve no history to check.

Quite a few teachers I know use it in some of these ways in a, like me, fairly causal way. This is a lot easier than thinking about any ethical and moral implication.

Lazy Sunday⤴

from

Soup
Soup flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

Since before we were married I’ve been going over to Niall’s parents for Sunday lunch. When I first went it was very much a family event with both N’s parents, usually his nephew Ewan who would be staying for the weekend, and his sister Shona would come over after her church service finished (usually very late, so she’d have to gulp down her food down to catch up). Lunch back then was a three course meal of soup, ‘meat’ and two veg, and a pudding with ice cream (always Mackie’s, with a choice of chocolate or vanilla). Sometimes Donald would come over with girlfriend Ruth, later with girlfriend/wife Kirsty. Shona would sometimes be accompanied by husband Nick. I started knitting again to counter the boredom of after lunch coffee – sitting in overstuffed armchairs in silence while Ian snoozed. Ewan and I would sometimes get the Brio trains out and play train crashes. Those were fun games!

Trains
Trains flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

After Ian died we continued with lunch at Morag’s, gradually moving from the three course meal to something lighter – soup with cheese and oatcakes became a regular. Pudding still came out as a treat, and the ice cream was always on offer. When Morag moved from the family home in Bearsden to a new flat in Milngavie the lunches continued. When we went on holiday to the Scottish highlands and islands we’d pick up local cheese and oatcakes – smoked cheese from Mull, cheese with mustard from Arran, mini oatcakes from Stornaway.

As Morag’s memory started to become erratic Niall took to phoning at 12pm on Sunday to check that she had everything needed for lunch, and we’d drop into M&S to collect anything missing. We started picking up tins of ‘nice’ soup to have in the cupboard just in case, and I’d keep a spare pack of coffee in my Sunday bag as that was often forgotten.

Then I started making soup – every Saturday I’d throw together whichever veg needed used – often broccoli or cauliflower stalks and leaves – into a big pot and blitz it with a handful of stilton. M&S mini submarine rolls were perfect to accompany this – and Morag always mentioned how tasty they looked. I bought a ‘picnic basket’ to transport everything, and invested in a robust soup flask.

357 Sunday Lunch
357 Sunday Lunch flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

I also started making a pudding – maybe apple and bramble crumble, with fruit from the garden – or a chocolate, ginger and pear sponge. These were adaptations of puddings my mum had made and dad had loved. We bought Mackie’s ice cream and left it in Morag’s freezer, noticing that it would sometimes go down during the week as well when we were not there.

300 Chocolate upside down cakes
300 Chocolate upside down cakes flickr photo by NomadWarMachine shared under a Creative Commons (BY-NC-SA) license

During lockdown Shona set up a Zoom meeting for 2pm on a Sunday. We’d all join from wherever we were, and both Morag and her elder sister Lesley would join from their separate homes. This tradition continued once lockdown was over, and we’d join the meeting from Morag’s flat after lunch.  We’d usually join whenever we were on holiday, showing off the view from whichever lodge or house were were staying in.

A couple of weeks ago Morag was admitted to hospital with pneumonia, and it became apparent that she was not going to be able to go home to her flat. Today, for the first Sunday in a long, long time, the whole day is my own. It’s currently 11.45, and previously I’d be thinking about heating up the soup, checking my knitting was in my Sunday bag, and getting ready to travel over for lunch (in fact, my computer has just pinged to reminder me). As things are put in motion for Morag to move into a home (and these bureaucratic wheels move very, very slowly), today there is no need to leave home for lunch, and I have not made soup. (I have, however, made some rhubarb crumble for lunch). Today I can knit if I like, or wander out into the garden – the time is my own.

I cannot express how relieved I am that there is not a family Zoom at 2pm!

Malicious links⤴

from

20190425_Elephant at the Metal Zoo


20190425_Elephant at the Metal Zoo flickr photo by Damien Walmsley shared under a Creative Commons (BY-NC) license

A couple of weeks ago I had a feeling that my emails weren’t working properly. I couldn’t work out what was happening, but sometimes I sent an email and got no response at all. But I didn’t think much about it until I sent an email a couple of times about something time critical to somebody I knew was awaiting my email, and followed it up with a Teams message when there was no response.

Anyway, after this I put a request over the help desk with a couple of dates and times of emails I thought were not getting through and asked them to look into it. Later that day, as I was out and about, I got a Teams message from a guy who identified himself as “Threat and Vulnerability Lead” at UofG who told me what was happening.

Back in late 2023 I changed my Outlook signature to include a link to my Mastodon profile. I still had a link to my Twitter/X in the same signature, you might like to note. And, on Feb 27th 2025 Microsoft decided that this Mastodon link was “malicious”. Yes, that’s right – nearly 18 months after I added it to my signature.

I removed the link – as the IT guy said, it wasn’t something UofG were doing or within their control. But it annoys the hell out of me that MS are flagging up a link to Mastodon as malicious while ignoring my Twitter/X link. (I removed that as well, by the way).

Go figure!

Copyright and Cartoon Mice – Gen AI Images and the Public Domain⤴

from

(This post was previously published on the Open.Ed Blog.)

With many image and media applications now integrating AI tools, it’s easier than ever to generate all kinds of eye-catching graphical content for your presentations, blog posts, teaching materials, and publications. Want a picture of a cartoon mouse to liven up your slides?  No problem! Stable Diffusion, Midjourney, DALL-E, or Media Magic can create one for you. And if your AI generated rodent happens to bear a striking resemblance to another well known cartoon mouse, well that’s just a coincidence, no?

Copyright and AI

The relationship between ownership, copyright and AI is still highly contested both in terms of the works ingested by the data models driving these tools, and also the content they generate.  Many of these data models ingest content that has been scraped from the web, with scant regard for intellectual property, copyright and ownership. Whether this constitutes legal use of protected works is a moot point. Creative Commons position is that “training generative AI constitutes fair use under current U.S. law”.  Not everyone agrees;  several artists and media organisations are attempting to sue various AI companies that they claim have used their creative works without their consent. Creative Commons believe that preference signals could offer a way to enable creators to indicate how their works can be used above and beyond the terms of the licence, and are exploring the practicalities of this approach (Preference signals for AI training.) It remains to be seen whether this is likely to be an effective solution to an intractable problem.

The European Union have taken a slightly different approach to copyright and AI with their EU Artificial Intelligence Act. Broadly speaking, the Act permits GenAI providers to use copyright content to train data models under the terms of the text and data mining exceptions of the existing Directive on Copyright in the Digital Single Market (DSM Directive).  However, rights holders are able to reserve their rights to prevent their content from being used for text and data mining and training genAI. Furthermore, providers must keep detailed records and provide a public summary of the content used to train their data models. In short, it’s a compromise; Gen AI models can scrape the web, but they must keep a public record of all the content they use, and they must allow copyright holders to opt out.  How this will work in practice, remains to be seen.

The UK is one step behind the EU, the government is undertaking an open consultation on Copyright and Artificial Intelligence, which appears to be broadly following the EU’s approach.

Copyright of AI Generated Content

Then there’s the issue of who owns the copyright of AI generated content. One common assumption is that AI generated images are not subject to copyright because they are not creative works produced by humans. Creative Commons perspective is that “creative works produced with the assistance of generative AI tools should only be eligible for protection where they contain a significant enough degree of human creative input to justify protection.” (This is not a bicycle: Human creativity and generative AI.) The problems start when AI tools generate images that are almost indistinguishable from the content they have ingested.  Take that AI generated cartoon mouse for example.  The reason it’s so similar to Disney’s famous, and famously copyright mouse is that the AI data models are likely to have scraped millions of images of Mickey Mouse from the web, with little regard for Disney’s intellectual property.  Rights holders may be able to argue that an AI generated image infringes their copyright on the basis of substantial similarity (The complex world of style, copyright, and generative AI.) This represents a risk which AI application developers are keen to shift on to their users. It’s not uncommon for AI applications to explicitly make no copyright claim over the images generated by their tools.  For example with regards to the copyright of AI generated images, Canva states:

“The treatment of AI-generated images and other works under copyright law is an open question and the answer may vary depending on what country you live in.

For now, Canva does not make any copyright claim over the images you create with our free AI image generator app. As between you and Canva, you own the images you create with Text to Image (subject to you following our terms), and you give us the right to host them on our platform and to use them for marketing our products.”

So if Disney does happen to spot your AI generated cartoon mouse and decides to sue, it’s you, or your employer, that’s going to be liable, not the tool you used to generate the image.

OER Service Guidance

The University of Edinburgh’s OER Service currently provides the following advice and guidance on using AI generated images:

OER Service advice on using Gen AI Images: AI generative software cannot create “original” images out of nothing. ​ ​ Its models are created from pre-existing data to generate derivative ​ images for your prompts. ​ ​ The origin of the pre-existing data is crucial not only in determining copyright ownership for the generated images, but to avoid rights infringement of pre-existing works. ​ ​ When Generative AI is trained using works created and owned by a third party without permission, the results may infringe on the rights of the third-party creator or artist.​

OER Service advice on gen AI images: Questions to consider:​ Does the Generative AI provider confirm that the data used to train the model has been legally accessed or licensed?​ 2. Will the provider own any rights over the creations?​ ​ Don't use AI Generated images unless they confirm that the images being generated come from an ethically-trained model which respects copyright and creators.​

We also recommend consulting the University of Edinburgh’s Generative AI Guidance for Staff.

Public Domain Images

A more ethical, and environmentally friendly, alternative to using AI generated images is to use public domain images, of which there are millions, with more entering the commons every year.  Public domain works, are creative works that are no longer under copyright protection because copyright has expired and they have entered the public domain, or they have been dedicated to the public domain by creators who choose to give up their copyright.  This means that they can be used free of charge, by anyone, for any purpose, without any restrictions whatsoever.  You don’t even have to provide attribution to the creator, though we always recommend that you do.

There are many fabulous sources of easily discoverable public domain images on the web, including:

Public Domain Day is celebrated on the 1st of January each year. In many countries, this is the day that copyright expires on creative works, and they become part of the public domain. This year, on Public Domain Day, the Public Domain Review launched a new interface to their Image Archive to enable users to search and explore their collections. 

And if you do happen to be looking for a cartoon mouse to use in your slides you’ll find one in the public domain that you can use with no restrictions or risk of copyright infringement, either for you or your employer. The original version of Mickey Mouse from the 1928 cartoon Steamboat Willie entered the public domain in 2024.

Black and white image of Mickey Mouse from the 1928 cartoon Steamboat Willie.

Mickey Mouse by Walt Disney, public domain image from the 1928 cartoon Steamboat Willie.

Further Reading