Sign in

NON-TOXICNo ads · No shady algorithms

iiinova is part of the Atmosphere — an open, public network where you own your identity.

Dr Pen

@DrPen.mastodon.social.ap.brid.gy
77 followers 17 following 953 posts

Tech society posts and shares from Dr Pen, Digital Media & Humanities academic based in Malta, from London. Late bloomer, a bit pithy. #UserExperience #UGC […] 🌉 bridged from ⁂ mastodon.social/@DrPen, follow @ap.brid.gy to interact

PostsRepliesMedia
Dr Pen @drpen.mastodon.social.ap.brid.gy · 24/10/2026
I moved my Bluesky account over to Eurosky because its probably a much better choice if youre in Europe, and fancy not being on the main US hosted Bluesky servers. Im `@penworks.eurosky.social` #Bluesky #eurosky #fediverse #decentralization
the form that asks to confirm the migration to Eurosky from BlueskyMy existing account, exactly as it was originally, but now hosted in the Netherlands and with a eurosky server handleA brief explanation of Eurosky
0161 16
Dr Pen @drpen.mastodon.social.ap.brid.gy · 18/10/2026
Todays concern is how to get my Open Access APC fee paid to the journal I have my paper in resubmission with before my institution affiliation runs out. Because peer review takes so long (literally 3 months to review and then 3 months to approve/ask for more […] [Original post on mastodon.social]
A screenshot of the open access book  Peer review in an Era of Evaluation - Understanding the Practice of Gatekeeping in Academia

 The book is available here: https://doi.org/10.1007/978-3-030-75263-7
001 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 14/10/2026
The phenomenon known as "dark patterns," where user value is supplanted in favour of shareholder value... This paper from 2018 has some good stuff in it. #darkpatterns #ux #enshittification #academia dl.acm.org/doi/10.1145/3173574.3174…
001 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 14/10/2026
Academic question. Please read image text! Source suggestions on staff and student consent for use of edtech. #academia #academicchatter #edtech #research
Academic research question. Does anyone have any good sources on edtech and consent? Im particularly interested in staff consent as well as student consent - and issues beyond 'privacy', though this is part of it. Data tracking in learning analytics is not new and there's plenty of research on questionable (non existent) student privacy. But Im interested in any opt-out clauses for edtech platforms (Moodle, Canva, Blackboard, other), use of Turnitin, longtail shelflife storage of written content, use of genAI in course design and assessment, etc. Staff consent is almost never written about in any capacity as far as I can tell. Is this intentional? ie research is not funded if it questions the role of bigtech domination in education/academia. Any tips on good papers are very appreciated
102 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 14/10/2026
Terrifying levels of digital illiteracy out there: "A browser is a search engine." NO IT IS NOT. You can choose your search engine completely separately from your browser. This applies to phones, tablets and laptops/desktops. Some good alternatives to google […] [Original post on mastodon.social]
Wikipedia list of web search engines, with language and owners

Baidu 	Chinese 	Baidu
Brave Search 	Multilingual 	Brave
Dogpile 	English 	Metasearch engine
DuckDuckGo 	Multilingual 	Multiple
Ecosia 	Multilingual 	Google and Microsoft Bing
Excite† 	Multilingual 	Microsoft Bing
Google 	Multilingual 	Google
Kagi 	English 	Metasearch engine and Kagi
Lycos 	Multilingual 	Microsoft Bing
MetaCrawler 	English 	Metasearch engine
Microsoft Bing 	Multilingual 	Microsoft Bing
Mojeek 	Multilingual 	Mojeek
Petal 	Multilingual 	Huawei
Qwant 	Multilingual 	Microsoft Bing
Sogou 	Multilingual 	Tencent
Startpage 	English 	Google
Swisscows 	Multilingual 	Microsoft Bing
WebCrawler 	English 	Microsoft Bing
Yahoo! Search† 	Multilingual 	Microsoft Bing
Yandex 	Multilingual 	Yandex
Yep 	English 	Yep
You.com 	English 	Microsoft Bing
033 3
Dr Pen @drpen.mastodon.social.ap.brid.gy · 07/10/2026
I just bought Felicity, possibly the best local music player for Android. And I didn't use Playstore, I downloaded from F-Droid and paid via Gumroad. It was €8.56. #audio #music #fdroid #foss #gumroad #degoogle #android
Felicity music player for android. Interface for Plaid playlistFelicity music player for Android. The player interface.
012 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 26/09/2026
Good morning from London, Saturday 26th September 2026. I'm having a few days here to catch the last of London summer vibes before heading up to Scotland for a day or two. Photo: A view down the runway at Luqa Airport, Malta, taken last Thursday […] [Original post on mastodon.social]
A rare view down the runway shows a long straight tarmac road going to the horizon. Taken from my seat on the plane just before takeoff from Luqa airport, Malta, on Thursday 24th September, 2026
001 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 21/09/2026
Debates outside the Ivory Tower - ChatGpt (generic), and human cognition. #genai #chatgpt #ai #llm #academia #academicchatter
ChatGPT is People - it's Soylent Green
A post by Damien Walter in his FB page 'Science Fiction'. With comments.
012 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 18/09/2026
This morning it finally rained in my area, after more than 125 days. Weve had a very dry end of summer this year. #malta
A close-up photo of rain droplets almost solidly obliterating the view beyond a balcony window in Malta, September 2026.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 30/08/2026
An evening of a couple of cold beers and memories of old friends now gone followed by an early morning of Maltese biker columns sweeping through on the nearby bypass at high speed with very loud engines. It is already very hot. The thick cover of Sahara sand […] [Original post on mastodon.social]
A photo of dried flowers and leaves in dark orange and yellow with some dark read buds. A closed rose is at the centre.
010 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 20/08/2026
So it seems to me that Google can see your phone IP even if location is turned off and you are running a VPN. huh. I just went to see how to delete Google auto-generated passkeys and the only way seems to be to sign out of the device account (which I need) […] [Original post on mastodon.social]
google message on IP assigned to device from the IP provider.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 11/08/2026
An image of Arundel Street in Hull, north east UK, in the 70s. This woman and her son are the last residents of this street, before slum clearance. Via T Jewett, Hull Facebook page. #urbanism #cities #history #ukhistory #UK #TheNorth #academia
An image of a dilapidated m street in a northern British town in the 1970s. A young woman and her young son stand in the centre foreground of the photo.
015 5
Dr Pen @drpen.mastodon.social.ap.brid.gy · 04/08/2026
Today I got a post from Kirupa about how great #AI is at making things they dont know how to make. Kirupa thinks this is awesome. I know its a complex discussion. But I also know this is the root issue with using #GenAI in most cases. It removes thinking […] [Original post on mastodon.social]
Pasted screenshot of Kirupa text from https://www.kirupa.chat/p/ai-lets-you-take-bigger-swingsmore "AI Lets You Take Bigger Swings...More Often!"

Whenever I wanted to build something, I would ask: How much do I know? How long will this take? What can I realistically build myself?

This wasn’t pessimism. It was rooted in what I saw work. Building technology was expensive. Prototypes took time. Crossing into unfamiliar territory added a boatload of unknowns. The sensible approach was to start with the smallest version, validate it, and expand only when the results justified the next investment ... 

Each stage had to earn the next one. Taking a big swing too early felt irresponsible because the cost of being wrong was so high.
103 3
Dr Pen @drpen.mastodon.social.ap.brid.gy · 03/08/2026
It is hot here. I have just submitted a peer review of someone else's paper. Shall I give in to temptation and have a cold Cisk beer? #beer #summer #Mediterranean
A screenshot of the Maltese beer company Cisk's website asking if the user is of legal drinking age, with an image of Cisk beer bottles in the background.
010 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 24/07/2026
Ongoing electricity crisis in Malta. Im very lucky to not be affected (yet) from long power cuts but the vast majority of people have been without power for 30-50 hours and counting. Infrastructure is not fit for purpose. A big increase in resident population […] [Original post on mastodon.social]
An ice cols glass of water.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 20/07/2026
Google continue the policy of platform entrapment. They dominate the Android ecosphere, then trap you into paying for the vast data trawling they are doing, forcing you to pay for your own surveillance. Ive already closed my paid Drive storage, removed […] [Original post on mastodon.social]
We are getting in touch to notify you of an upcoming update to our storage policies, along with new controls designed to give you more control over your Android backups.

The policy update: Effective in 45 days, all data included in your Android device backups will begin counting towards your Google Account storage limit. Photos and videos in Google Photos and MMS data already count towards your Google Account storage today. With this change, all other backed-up data such as SMS messages, call history, device settings and app settings will also count towards your Google Account storage. The storage used by your device backup may increase once this policy takes effect. If your Google Account exceeds its storage limit, your automatic backups will pause until you free up space or upgrade your plan.
New controls for you: With this change, we are also introducing granular app backup toggles in your Android backup settings. If you are on Android 9 or higher, you can now select exactly which data types and individual apps you want to back up, ensuring that you only use storage for what matters most to you.
Your current storage status:
Your estimated increase in Android device backup will be 0.18 GB after this change. You've used 9.71 GB of your 15 GB storage.
001 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 14/07/2026
Waymo attack recently in LA. The New Luddite revolution is gaining momentum bigtime. If you search for video of waymo attacks youll find a lot, not only in the US. Anti data centre activism is expanding rapidly across the globe too. Brian Merchants blog […] [Original post on mastodon.social]
A man attacks a waymo in Los Angeles, vandalising it and throwing bits of it into the street.
010 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 04/07/2026
Ive been on vacation to the UK, and taking a break from socmed most of the time. Here's where I was some of the time: #photography #scotland #scottishborders #uk #family #holiday #vacation #photos #summer
a large stone house in the Scottish borders, with a vista on the right side of green fields, trees and bright blue sky
151 5
Dr Pen @drpen.mastodon.social.ap.brid.gy · 26/06/2026
The library in a Scottish borders house I'm staying in today. On vacation but doing a little academic work to meet a deadline. Its been very warm but today the weather broke again and its much cooler this evening after a thunderstorm and a downpour […] [Original post on mastodon.social]
A traditional country house library room with a cup of tea, a chess set, large arm chair,  crassula green succulent plant and pule of magazines. A sash window behind shows a glimpse of the countryside view beyond.
010 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 01/06/2026
So yeah. You might have seen the news about the fireworks factory going boooooom this morning in Malta. This photo is looking from my place, just after the really big explosion. A lot of nearby buildings had windows blown out, and farm animals and wildlife […] [Original post on mastodon.social]
A photograph of a huge plume of smoke in the sky, above some buildings in the foreground.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 28/04/2026
'Digital Transformation', so loved by JISC and UK HE is not mandatory. Other advanced countries are going in the opposite direction. > (my input) I spoke with a techbro guy who was on the digital policy board for UK HE. He knew zero about learning, and just […] [Original post on mastodon.social]
Sweden is championing a return to books and paper instead of screens in schools ... going back to basics, reintroducing physical textbooks, limiting screen time, and investing heavily in school libraries. "We jumped on that [digital] train very fast, ... but more people realise that we should not have done that," says education spokesman for Sweden's Liberal Party.
"The impact on students, on how schools work, has been very bad, so we agree that it was a mistake, but we're changing it back,". Sweden's international test scores are falling compared to other nations, and there's growing concern about the impact of screens on pupils' concentration.
In a nation where the vast majority of pupils have their own phone before they turn ten, scientists say there could be implications for young children's development. 
"Too much screen during this very critical developmental phase can essentially impact your language development, your cognitive development, impulse control," says Sissela Nutley, a neuroscientist who runs digital health programmes in schools.
But not everyone agrees with going back to books. 
"We can't go back wishing for a school system that was like the 1950s," says Linnéa Stenliden, professor of education at Linköping University.
She argues that the new policy could bring about a digital divide in a world that has moved away from books and paper, with some children understanding new technologies better than others. "Digital competence is actually needed," she says.
020 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 21/04/2026
There was a post on bsky that was exactly what happened to me recently - being asked to review a paper partially 'written' with chatGPT. I emailed the editor (Elsevier journal) to say the paper should not be accepted for peer review. I rejected the paper […] [Original post on mastodon.social]
I regret to have to inform you that I cannot review work that is partially processed by a product that in my opinion fails to meet standards of scientific integrity that I need to comply with. 

“ChatGPT-[version redacted]” is a non-transparent proprietary product created with corporate biases, through exploitative labour practices, violating authorship rights, and having known psychological, societal, and environmental harmful impact. Moreover, notwithstanding the hype that OpenAI has been fuelling, a product like ChatGPT cannot [redacted], because [redacted] is a cognitive capacity, and the product has none. All in all, the use of ChatGPT for scientific purposes, in my honest opinion, clashes with core scientific integrity principles such as honesty, scrupulousness, transparency, responsibility, and independence. 

Therefore, I need to decline this invitation to review. 

To better understand my position, I recommend the following papers from me and my collaborators (see my signature for more resources): 

.../
021 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 19/04/2026
Providing ideas for how to teach social media literacy, circa 2019 #socialmedia #digitalliteracy #academia #academicchatter
What is social media for?
I use social media to understand other people.
I use social media to understand myself.
I use social media to learn humility and enjoy other people’s posts even though I might think they’re boring.
I use social media to examine how I behave towards other people.
I use social media to learn the art of self control.
I use social media to appreciate the vast diversity of life and the world around me.
I use social media to be aware of dangers and potential wrong decisions I have made, or others have made.
I use social media to enjoy other people’s achievements without being jealous.
I use social media to reflect upon my place in the world.
I use social media to discover that I am not as great as I thought I was.
I use social media to discover that other people are more interesting than I thought they were.
I use social media to…
Derived from recent social media posts May 2019.
003 3
Dr Pen @drpen.mastodon.social.ap.brid.gy · 18/04/2026
**AI Assistance Reduces Persistence and Hurts Independent Performance. **A paper from researchers at Oxford, MIT, Carnegie Mellon and UCLA that needs to be seen by everyone in the education community. Full Abstract in alt text. #ai #GenAI #chatbots #edtech […] [Original post on mastodon.social]
People often optimize for long-term goals in collaboration: A mentor or companion doesn't just answer questions, but also scaffolds learning, tracks progress, and prioritizes the other person's growth over immediate results. In contrast, current AI systems are fundamentally short-sighted collaborators - optimized for providing instant and complete responses, without ever saying no (unless for safety reasons). What are the consequences of this dynamic? Here, through a series of randomized controlled trials on human-AI interactions (N = 1,222), we provide causal evidence for two key consequences of AI assistance: reduced persistence and impairment of unassisted performance. Across a variety of tasks, including mathematical reasoning and reading comprehension, we find that although AI assistance improves performance in the short-term, people perform significantly worse without AI and are more likely to give up. Notably, these effects emerge after only brief interactions with AI (approximately 10 minutes). These findings are particularly concerning because persistence is foundational to skill acquisition and is one of the strongest predictors of long-term learning. We posit that persistence is reduced because AI conditions people to expect immediate answers, thereby denying them the experience of working through challenges on their own. These results suggest the need for AI model development to prioritize scaffolding long-term competence alongside immediate task completion.
017 7
Dr Pen @drpen.mastodon.social.ap.brid.gy · 16/04/2026
ChatGPT circa 2022. I requested to delete my account using a "delete my chatgpt account request". I had used it 5 times. Chatbots are a political issue. #declutter #data #SpringCleaning #ai #ChatGPT #academia
a screenshot of chatgpt output cira 2022, with a prompt about a space droid and a totally non committal useless but human sounding response.
022 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 09/04/2026
Brief update on my LinkAce implementation. I run darkmode with custom CSS for look and feel, and themed for PWA design. Made a few tweaks of very small things in the template code. Now updated to 2.5.3 but next version is available. This can be a multi-user […] [Original post on mastodon.social]
The homepage in LinkAce, showing mid purple theme for darkThe links list in LinkAce, showing mid purple theme for dark, and zebra list with opaque purple hover roll over.A link page in LinkAce, showing mid purple theme for dark primary button, with thumbnail, and fields for each link.A link edit mode page in LinkAce, showing mid purple theme for dark primary button, with edit form fields for each link.
020 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 27/03/2026
The AI book Grift. Im sure we could make an image like this for funded AI research projects. #ai #books #academia #academicchatter
A screenshot of 21 different books advertised on Amazon
013 3
Dr Pen @drpen.mastodon.social.ap.brid.gy · 22/03/2026
The new T&F guidelines for editors when accepting paper submissions. #humor #funny #ai #fakepapers #academia #academicchatter #academicpublishing
An edited version of an academic journal AI policy statement, with additional truthful factors, mostly about making the cheating sound convincing if authors use AI.
025 5
Dr Pen @drpen.mastodon.social.ap.brid.gy · 16/03/2026
Monday, calm evening. #blackandwhitephotography #bwphotography #phonephotos
Are you other? A black and white photo of a small Taylor guitar hanging on the wall next to a narrow tall book shelf full of paperback books, with a jar of paintbrushes on the top shelf.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 05/03/2026
The PolyMarket Casino of Death (TM). A total dereliction of human morality, but don't let that stop you making humungous coin. After all, its powered by technologies of imperial scale human labour extraction, so why not their lives? Is anyone doping any […] [Original post on mastodon.social]
a screenshot of the polymarket homepage with a large display on the betting for next Iranian supreme leadera screenshot of the polymarket homepage with a large display on the betting for when the iranian regime will fall
001 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 05/02/2026
Why are we kidding ourselves? Or am I missing some key information about how great 'AI' is in L&T at university? How is this 'the future'? What is it about smart sounding chatbots (even if they were accurate and reliable) that is so good it replaces human […] [Original post on mastodon.social]
A screenshot of text from an anti 'AI' FB group, shared yesterday. 

Slightly edited text:

I'm currently taking up my Master's, so doing this anonymously for obvious reasons.
One of my groupmates is a 40+ year old college professor and we have a task to work on a critical analysis of some papers. I gave her the introduction to work on since it would be easiest and only needs ~500 words.
She basically told Chatgpt to do her work for her, but since it kept using business and technical jargon, she didn't understand it. So she kept telling Chatgpt to "simplify" but all it did was shorten sentences and use more convoluted words until 0% AI detection was reached.

● She's a college professor in the academe with decades of teaching experience but she's heavily reliant on AI.
● If she can't understand the introduction, how can she say she understood our reading?
● If she admittedly doesn't comprehend technical jargon and left it up to AI to do her schoolwork, is our MBA even worth anything and what's the point of pursuing it when you didn't learn?
● If these are the professors right now that my kids will have in the future, I might as well home school them.
● She blames Chatgpt for not understanding what she meant by "simplify" but it isn't the machine's fault that she didn't specifically say how she wanted it simplified.
● Chatgpt can now bypass AI detectors if prompted properly and result in 0% AI detected, while 100% human-written papers are flagged as at least 20% AI-generated.
011 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 31/01/2026
Good morning fellow Fediverse renegades. Good to be here today, and every day. #mastodon #fediverse #safespace #mypeople #socialweb #opensocialweb
An astronaut standing next to a swimming pool with a Googie steel frame chair at the side. Sunlight streams into the room from above, down a stairway. A large round window with slatted blinds is on the opposite wall.

Image from https://unsplash.com/@cashmacanaya
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 26/01/2026
Stupidity Delusion level off the scale. My email app is sad if I dont tell it I love it now and again. (Universities: we must prepare our students for empathy with agentic ai) #aibubble #aidelusions #TechFantasyLand #fomo #AIgrifting #academia #academicchatter
Screen shot of Business Insider article with headline "Anthropic's philosopher says we don't know for sure if AI can feel".
042 4
Dr Pen @drpen.mastodon.social.ap.brid.gy · 22/01/2026
Wikipedia for the people of the moment: 1. Carney en.wikipedia.org/wiki/Mark_Carney 2. Thucydides en.wikipedia.org/wiki/Thucydides #davos #ancientgreece #wikipedia #carney #thucydides
A photo of Mark CarneyA plaster cast bust of Thucydides, from a Roman copy.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 15/01/2026
Apple killed U2s career with this, they never managed to repair the damage of being in bed with a corporate platform. But outrage forced Apple to allow people to delete the album. It's a key issue of AI into OS and software, you cannot fully remove it. At […] [Original post on mastodon.social]
A meme saying:

Remember when Apple forced a U2 album onto everyone's itunes without making it possible to delete? Eventually the world bullied them into making a "delete" button. Lets get that delete button created in 2026 for a new two-lettered word plaguing the world.
104 4
Dr Pen @drpen.mastodon.social.ap.brid.gy · 11/01/2026
Video clip of the infamous Iranian girl with cigarette burning the image of the Iranian leader. www.facebook.com/reel/2461292674265… For those curious about GenAI, fake content and political (mis)messaging, the image shared in socmed does NOT […] [Original post on mastodon.social]
Screenshot of video still of the Iranian girl with cigarette burning the image of the Iranian leader

Link to FB reel video is in the post.
002 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 29/12/2025
I got notified by Google Scholar Alert about a new paper that is on topic for a paper Im writing: www.ijapas.ir/index.php/ijapas/arti… Its by Iranian university academics. The journal is not known to me. The language seems 'not right'. I […] [Original post on mastodon.social]
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 29/12/2025
I got notified by Google Scholar Alert about a new paper that is on topic for a paper Im writing: www.ijapas.ir/index.php/ijapas/arti… Its by Iranian university academics. The journal is not known to me. The language seems 'not right'. I […] [Original post on mastodon.social]
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 20/12/2025
This. Can academics continue their current AI (mostly GenAI) warm embrace? How can they not admit the car crash going on? Why would we want to contribute to the Great Slop? Does everyone have a kind of Stockholm Syndrome about bigtech Genai? In 2023 some of […] [Original post on mastodon.social]
Screenshot of text:

Anthony Moser, a software engineer and technologist, was among those who foresaw how chatbots could eventually hollow out educational institutions. “I’m imagining an instructor somewhere making a syllabus with ChatGPT, assigning reading from books that don’t exist,” he wrote in a post on Bluesky in 2023, less than a year after the AI model first came out. “But the students don’t notice, because they are asking ChatGPT to summarize the book or write the essay.” This month, Moser reshared that post, commenting: “I wish it had taken longer for this to become literally true.”

from a Blusesky thread: https://bsky.app/profile/milesklee.bsky.social/post/3ma7gerotjc2z
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 16/12/2025
LinkedIn shitty emails achieve new low with "puzzle games". I never use Linkedin. I have unsubscribed from every email list. I never engage with the platform, unless a colleague sends me some post they shared. I then use cookie blockers. But now they think my […] [Original post on mastodon.social]
a screenshot of an email from linkedin showing two low brow "puzzle games", sent to a phd academic.
002 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 15/12/2025
Ive been digging around for text analysis OS apps and found AntConc via a Reddit thread. This app is very good from what I can see in early quick testing. Im looking at term frequency across relevant papers, and some 'concordance' context but AntConc will do […] [Original post on mastodon.social]
A wordcloud output from AntConc text analysis software showing the frequency of words for a group of PDFs about big tech and edtech governance. Governance, edtech, design, system, time, media and language are large words in the cloud.
002 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 10/12/2025
Good morning the great wonderful Mastodon and wider fediverse. Its good to be here, amongst the sane and lovely people of the world. Have a super academic and otherwise day Image is my view this morning around 6.45am CET, Malta sunrise. #morning #europe […] [Original post on mastodon.social]
Sunrise with low glowing cloud and silhouetted buildings, Malta.
011 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 04/12/2025
Always remember, as you attend or give presentations on the use of GenAI in T&L, many people in each session will be sceptical, critical and not agree with the widespread adoption of text or image generators in their practice. But they wont speak out openly […] [Original post on mastodon.social]
Screenshot of post in the Artists Against Gen AI group on Facebook. It reads:

"We had a training at my school today about using AI in the classroom. The presenter was using it for everything from generating warm up questions to grading and giving feedback to students. I don’t know what’s even the point of people anymore.
Well, after his spiel, I raised my concerns about its use. I pointed out that AI LLMs don’t produce actual content, but approximations of what you want, resulting in made up citations and factual errors. I pointed out that the creators of the content used to train the models were not compensated or even asked if their works could be used, so it’s built on theft. I pointed out that studies have shown that repeated use of AI is shown to reduce critical thinking skills and empathy. And I pointed out that AI is incredibly bad for the environment, and data centers are causing water supplies to dry up in this country. I concluded by saying that I felt disheartened that our district and admin were pushing for its use.
I felt really nervous speaking my mind about it, but people clapped at what I said. I had numerous teachers find me today and tell me they appreciated what I said. It’s always good to know that those of us against AI aren’t alone, even if it feels that way sometimes. I know this group has given me a lot of strength and hope. Keep fighting the good fight!"
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 02/12/2025
An interesting online advent calendar from the Ashmolean. Here are Dec 1 and Dec 2... #christmas #advent #calendar #adventcalendar #museums #Ashmolean #academia #academicchatter www.ashmolean.org/advent
View of snow on Benten Hill at Kinryūzan Temple in Asakusa by Utagawa Kuniyoshi

Kuniyoshi was one of the last great masters of the Japanese ukiyo-e style of woodblock printing. He is probably best known for his prints of warriors and battles of legendary heroes, but also created designs of ghosts, cats, beautiful women, actor prints and landscapes.

This print shows snow at the Kinryūzan Temple in Asakusa, Tokyo and is from Kuniyoshi's series of Famous Places in Edo (modern Tokyo). The temple sits across a lake with snow-laden trees on the right. Small figures in front of the temple hold umbrellas against the snow.Bronze brooch, 100–200 CE, from Amiens, northern France

This Roman bronze brooch in the shape of a stag, once decorated with bright multi-coloured enamel, would have been used to fasten a shawl or cloak. The stag is looking backwards and in the centre of the body are aged pieces of coloured enamel.

Enamelled jewellery was very popular in Britain and France during the Roman period.

The vast Roman empire encouraged the movement of people, ideas and fashions.

Certain Roman styles of dress and jewellery became widespread with some personal items travelling long distances with their owners, though some distinctive local fashions did continue to thrive.
022 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 27/11/2025
See my earlier post today about the journal mill grifters all coming to the AI yard. #academicpublishing #ai #journalmills
Screen capture of a Ben Williamson tweet referring to the huge amount of research papers all about AI.
011 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 27/11/2025
Submitted another paper. Its interesting, but may be unsuited to the journal. Journal specs are narrow, my papers cross discipline boundaries. In the words of the great and very interesting Peter Morville, I'm a boundary object. Here's one of his books on IA […] [Original post on mastodon.social]
A screenshot of Intertwingled by Peter Morville, on the Internet Archive.
002 2
Dr Pen @drpen.mastodon.social.ap.brid.gy · 21/11/2025
Ive been tussling with this all day, its from a high prof new paper. I find it a very misleading description of 'GenAI'. 1. #GenAI isnt only text (!). 2. It is not human chatter, its billions of pieces of written output - books, reports, academic work […] [Original post on mastodon.social]
screenshot of a section of text from an academic paper. GenAI is different. It is not merely an application of the logic of print literacy. It emergedfrom the internet era, relying upon massive datasets generated through the continuous,collective ‘chatter’ of online human communication. Unlike its predecessors, generative AIdoes not operate strictly through predefined rules. Instead, it learns patterns from vast un-structured data, offering context-sensitive, open-ended responses. From an epistemologicalperspective, this brings new challenges.  

(PDF) A dialogic theoretical foundation for integrating generative AI into pedagogical design. Available from: https://www.researchgate.net/publication/397662591_A_dialogic_theoretical_foundation_for_integrating_generative_AI_into_pedagogical_design [accessed Nov 21 2025].
011 1
Dr Pen @drpen.mastodon.social.ap.brid.gy · 18/11/2025
Democracy in society resource alert. This is a fantastic resource on the quality of and threats to information systems in our digital world. Also, free book - a comprehensive and very relevant report. There is also a visualisation tool to see all issues […] [Original post on mastodon.social]
screenshot. 
Information ecosystem and troubled democracy – Interactive Map
Developed in partnership with CNRS Institute for Complex Systems.
This map represents a statistical summary of the thematic content of the report. The network graph represents relations between the words in the report, placing them closer to each other the more they are related. The bigger the node, the more present the word is, signalling its role in defining what the report or chapter is about. The colors represent words that are closely related to each other and can be interpreted as a topic.

The map was generated by Giovanni Maggi using the GarganText tool – developed by the CNRS Institute of Complex Systems – on the basis of the report’s text. A Global Synthesis of the State of Knowledge on News Media, AI and Data Governance
This inaugural meta-analysis provides a critical assessment of the role of information ecosystems in the Global North and Global Majority World, focusing on their relationship with information integrity (the quality of public discourse), the fairness of political processes, the protection of media freedoms, and the resilience of public institutions.

The report addresses three thematic areas with a cross-cutting theme of mis- and disinformation:

Media, Politics and Trust;
Artificial Intelligence, Information Ecosystems and Democracy;
and Data Governance and Democracy.
The analysis is based mainly on academic publications supplemented by reports and other materials from different disciplines and regions (1,664 citations selected among a total corpus of over +2700 resources aggregated). The report showcases what we can learn from landmark research on often intractable challenges posed by rapid changes in information and communication spaces.
013 3
Dr Pen @drpen.mastodon.social.ap.brid.gy · 06/11/2025
Im cleaning out boxes and going through old computer kit. Ive got 2 old Acer Revos, (win 7 net tops), 2010 and 2012. Is there any market for this stuff? Resale on eBay isnt great but maybe computer museums might want them. They're in fantastic condition (with […] [Original post on mastodon.social]
A box for the Acer Revo 2012A box for the Acer Revo 2010.
000 0
Dr Pen @drpen.mastodon.social.ap.brid.gy · 30/10/2025
I love it when recommender systems are so chronically off that it just confirms the coming automated dystopian future we have built will be 90% Brazil and 10% LOTF. #recommendersystems #researchgate #academia #academicchatter
A screenshot of an email from ResearchGate that says, We found a free webinar that matches your interests. The body of the message says, A reliable CNS safety assessment method. CNS safety risks often go undetected in the GLP talk studies because animal models fail to predict human neurological effects. CNS-3D brain organoids are 7.4 times more accurate than animal models offering a more reliable CNS safety assessment. 

It has to be said that I am not a neurological or biological scientist. I am in fact a technology in society, in urban and learning contexts academic. So in fact have absolutely nothing to do with this field of study.
000 0