Digitalhumanities, #librarylife and #NLP friends: has anyone bought a large historical dataset (books, newspapers, maps etc, images or ocr) from a publisher/digitiser to data mine? I’d like to compare notes… either on here or DM me 🙏

Brilliant piece just out on “Inappropriate AI” by Colin Milburn & @ritaraley.bsky.social. A critique of higher ed’s legitimation quest for policies of “appropriate” AI use despite fact that LLMs are by nature inappropriate. They "appropriate" all edu and other materials. uchri.org/foundry/inap...

What happens when you let AI write match reports? You end up with reports for games that never happened. Levante v Villarreal was postponed due to bad weather but according to this site it was played, finished 0-0 and was very dull. They even awarded a man of the match! www.bwin.com/en/news/post...

2025 in People embroidery is finished & on the wall! All the meaningful interactions I had with other adults, every day of last year. Different colors for people from different parts of my life, different stitches for different interactions. Guide here: quinndombrowski.com/textiles/202... #DHmakes

Last year I was asked to give the keynote at the Conference of European National Librarians (CENL) meeting on AI and Libraries, and here it is! "Be More Library: Upholding Library Values in a Tech Industry World". In short, we need to define what WE want, and go get it! www.cenl.org/wp-content/u...

We've been working on our new Digital Collections Platform - launching formally in 2026. Thanks @emmahorrell.bsky.social for all your UX guidance and support! blogs.ed.ac.uk/website-comm...

'Who decides what’s worth remembering, and how does that shape our shared story?' @andreakocsis.bsky.social and Dorsey Kaufmann's exhibition, Digital Ghosts, opens in November. Really happy that NLS could partially fund this work through our Digital Research Fellowship: data.nls.uk/projects/dig...

A really interesting piece on the Return/Enter key. aresluna.org/the-day-retu... I also did some work on the terminology of "return" in this context in www.sup.org/books/media-...

Customer feedback this week; ‘Please can you get some more diversity at the National Trust by encouraging more hot men to visit.’ Ideas for how we do this v welcome. Nothing too racy please - we are the National Trust not the National Tryst.

So @theturing.bsky.social has decided they no longer need Humanities (!) - so the The Alan Turing Institute’s Humanities and Data Science Interest Group are having one final event: "Network power: the humanities and data science in collaboration" In person and online, Oxford, 25th Sept. Pls share!

Spent the morning seeing if I could map some of London pornographers' addresses during the mid-19th century w/ the National Library of Scotland's georeferenced maps. Did I need to do this? Probably not. But very neat to see some of them laid out spatially.

Is this our youngest author on Transkribus Sites? Scottish writer Marjory Fleming was just seven years old when she starting writing her diary, which became popular in the late Victorian period. You can read the diary in full on Transkribus Sites 👇 app.transkribus.org/en/sites/mar...

GenAI FOMO has spurred businesses to light nearly $40 billion on fire www.theregister.com/2025/08/18/g...

Hello. I wrote a nice long essay about AI and this very strange moment where we're constantly told we're living in the dawn of a strange new future but the only thing that's actually clear is that everyone feels pretty unmoored and uncertain. I hope you'll read it

"Developing AI technologies in ways that truly benefit the public and treat common resources with care would constitute a more responsible practice, requiring further commitments to transparency and accountability around negative externalities." Noroozian et al on AI and the digital commons.

On Gaza’s killing fields, journalism faces its darkest hour – but that won’t stop us reporting | Asef Hamidi

The University of Warwick has paid a lot of money, I'm guessing, for their internet banner ads on the Guardian and elsewhere. But they can't spell "curiosity". This isn't a US-UK variant. They just, genuinely, have a massive spelling error in their banner ads.

🎪 The mayhem begins! Celebrate with us! Doors open at 10am at George IV Bridge (Edinburgh) and Kelvin Hall (Glasgow). Be among the first 500 at each site to get your special edition centenary comic, created by Beano! #NLS100 #Beano

Many VLM-based OCR models have been released recently. Are they useful for libraries and archives? I made a quick Space to compare VLM OCR with "traditional" OCR using 11k Scottish exam papers from @natlibscot.bsky.social huggingface.co/spaces/davanstrien/ocr-time-capsule

[Blogged] The AI powered Library Search That Refused to Search - From Clarivate's Summon to Primo Research Assistant, content‑moderation layers meant mostly for chatbots are "quietly" blocking controversial topics from being searched aarontay.substack.com/p/the-ai-pow...

Here's a write-up of my keynote from #DH2025, since academic publishing is slow and this was written for right now. roopikarisam.com/talks-cat/dh...

Unsure if that's an opposition to AI or an opposition to the opposition of AI. But the culture of backlash it has created is a real danger that could undo 20 years of utopian thinking, argumentation, and implementation.

The Library of Congress is looking for feedback on their amazing/unique "Selected Datasets" collection. You can see the collection here and the survey link is at the top of the page. www.loc.gov/collections/... Anyone interested in working with datasets check out the collection and take the survey.

Stanford HAI AI Index: 'training compute doubles every five months, datasets every eight, and power use annually' but AI models 'often fail to reliably solve logic tasks even when provably correct solutions exist, limiting their effectiveness in high-stakes settings where precision is critical'

For 30 years I've been experimenting with ways to see & use the online collections of libraries, archives & museums. I've made useful things, playful things, & weird things. Here's a big list of them for you to explore! https://wraggelabs.com #glam #histodons #digitalhumanities

Is it human to make a huge profit from knowledge? asks Hilde Van Wijngaarden from Vrije Universiteit Amsterdam in this afternoon's session on Knowledge as a Catalyst. #LKFS2025

here's my presentation to the super #LKFS2025; super important conversations. Many thanks to @masudk.com and colleagues for showing the way to knowledge futures. and here's Allan Ginsberg on the library dancefloor! anatomiesofpower.wordpress.com/2025/06/27/k...

Happy to see this published today, and it's been a real privilege to work on with Paul Gooding, @melissaterras.bsky.social and the @facetpublishing.bsky.social team. Thanks to all the wonderful contributors! Here's the link: www.facetpublishing.co.uk/page/detail/...

Hurrah! My next book is instantiated, and in-the-world! Library Catalogues As Data: Research, Practice, Usage. Edited with Paul Gooding and @semames.bsky.social. Proud of this one! Official publication date soon… www.facetpublishing.co.uk/page/detail/...

Today we launched a small service on JSTOR to help expert researchers get bulk access to bibliographic data and full text content. A fun part: you can go download 12M+ metadata records bulk in one file, right now: www.jstor.org/ta-support

'“I don't think that people appreciate how few people are working to keep these collections online, even at huge institutions,” Weinberg told me. “It's usually an incredibly small team, one person, half a person, half a person, plus, like their web person who is sympathetic to what's going on.'

Fancy Edinburgh Uni as a visiting fellow? Global Majority Fellowships in NLP 2025-2026, for early careers researchers. For this 1st year, we welcome applications on “Responsible NLP and Digital Cultural Heritage”, working with me! 2-3 months in duration, Visa fees, stipend &travel. Please share!

Today we released Institutional Books 1.0, a 242B token dataset from Harvard Library's collections, refined for accuracy and usability. 🧵

Brilliant definition outline of the ALA defn of sustainability from @rebekkahsa.bsky.social Sustainability comes from meeting the triple bottom lines: • environmentally sound • socially equitable • economically feasible And a good reminder that every job is a climate job. #LCSeminar2025

The creators of Myst series (@cyan.com) saved hundreds of video tapes from the making of their iconic adventure game franchose. We helped preserve them. Here's what we found.

So if I am getting this right: Thursday the librarian of Congress was fired. Friday the copyright office rushed out a draft report suggesting genAI training on copyrighted works shouldn’t be legal. Saturday the head of the copyright office is fired.

Got suggestions of priority #LoC datasets or collections to be saved? Please let the excellent @datarescueproject.org folks know!

Very pleased to say Paul Gooding, @semames.bsky.social and I are looking for a PhD student, funded by AHRC @sgsah.bsky.social. Topic: '“Mainlining AI into the Veins”? Library collections as data in the age of Extractive AI Capitalism'. Closing 5pm, Friday 30th May 2025. Please share!

Data is the foundation of AI. Poor quality data drives up costs and can lead to hidden problems for AI, while biased data negatively affects the performance of AI models. Our latest research seeks to define AI readiness and its underlying principles. Find out more.

Seeking PhD student to work on '“Mainlining AI into the Veins”? Library collections as data in the age of Extractive AI Capitalism'. AHRC-funded @sgsah.bsky.social & supervised by Paul Gooding, @melissaterras.bsky.social , @semames.bsky.social and me. Please share! melissaterras.org/2025/05/06/s...

Here's the video of my BRAID seminar on 'Decomputing' - countering AI's scaling of structural violence and environmental nihilism via degrowth, de-automatisation and pushing for convivial tech through people's councils vimeo.com/1079775335

🔊 Applications are open for our new MSc in Digital Humanities (Data Science). Tell your students! pixelfed.social/p/dh/8227564...

Incredible satellite image of the 33 (!!!!) operational fossil fuel combustion engines at the site of Elon Musk's XAI data centre “appalling that xAI would operate more than 30 methane gas turbines without any permits or any public oversight" @selc.bsky.social www.theguardian.com/technology/2...

An imitation of Ukraine's Zorya Luhansk has just started up in the Russian pyramid. It's not the first example of teams from sovereign Ukrainian territory appearing in Russia, and may not be the last. Football is being used to normalise violent denial of the past. www.theguardian.com/football/202...

Cognitive dissonance is the humanities being at the same time both a useless waste of time and an imminent threat to the government.

My Slate piece on the widespread NEH grant cancellations is live. So much gratitude to everyone who took the time to talk with me these past few days while fighting fires on every side. #humanities #highered #neh slate.com/news-and-pol...

'Our content is free, our infrastructure is not'. Important statement from @wikimediafoundation.org. People creating crawlers to feed AI should behave better. diff.wikimedia.org/2025/04/01/h...

Dissolution of the Office of Digital Humanities is confirmed. #DigitalHumanities