Libraries and Learning Links of the Week

Web links about libraries and learning, every week.

Librarians fear new penalties, even prison, as activists challenge books

It's not just Ron DeSantis' “War on Woke” in Florida that is interfering in core librarian work in the United States. Librarians in the state of Missouri are also working in a climate of fear after their far-right government passed vague laws that are ultimately aimed at criminalising queerness, and criminalise basic library work as collateral damage.

Putting “Nothing About Us Without Us” into Practice in Small and Rural Libraries

Something more positive, this time from rural North America. A few case studies in the use of funding to improve accessibility in libraries, with examples of how libraries changed their original plans after consulting with community members with disabilities.

Generative AI in Writing Research Papers: A New Type of Algorithmic Bias and Uncertainty in Scholarly Work

This is a pretty interesting preprint from December, looking at some of the more subtle problems with using large language models in the production of academic papers.

Due to the pressure put on scholars to publish, they may increasingly rely on automative solutions that are capable of producing fast yet potentially problematic papers

This is indeed a problem. I have thoughts on where those pressures come from but that's for another time.

We find 104 scholarly works using our criteria and query (Author: “ChatGPT” OR “GPT-2” OR “GPT-3” OR “Generative AI” OR “Generative Pretrained Transformer” OR “OpenAI”).

Nice, simple. But also, “some people say”...

We note that remnants of AI prompts left in manuscripts due to careless errors indicate that our systematic review leaves out articles that used a generative AI tool, but neglected to cite it as an author. We verify this claim via Google Scholar’s search engine, using the query as shown in Figure 2. We suspect that far more authors have used ChatGPT similarly, but will have edited out such apparent markers of direct AI use—which some may deem plagiarism, depending on journal policy.

A key point the authors make is that the norms of academia expect that any automation or computer model used in the research or the production of a paper should be openly available so that it can be verified and any suspected errors can be traced back to their source. But commercial LLMs are mostly black boxes:

if an AI-produced work is made public, yet the model used to generate that work is not, how do we judge the cause of the bias—is it attributable to the model or the user who provided a prompt to it?

Something I hadn't really considered before that is explored in the paper is the concept of “inadvertent red teaming”, which refers to what happens when an LLM is given a prompt without enough context to produce a coherent answer. Even with a good prompt, LLMs are prone to just make stuff up (or “hallucinate”, as AI boosters like to sanitise it as). With a prompt that provides inadequate context, they can easily go off the rails, however, and if researchers don't vet the output properly this can lead to highly problematic statements in papers.

A more subtle problem they identified is that if an LLM is given context about the prompter – even if this is inadvertent such as their name and position title in an ingested draft paper – it can bias the output in some obvious and some subtle ways.

If you're interested in this sort of stuff, you can also read this paper desperately trying to make it ok to make up your research data. Seems to fly in the face of the whole foundation principles of science to me, but what would I know?

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

Growing Up: A Maturity Model for Open Education

This model from Pressbooks is an interesting way to think about where your institution is at with open education efforts.

I found it particularly helpful as a reminder that whilst a lot of the ideas held within open education are very old, as a conceptual model unto itself it is still quite new. At my work we run one of the oldest open book publishing operations in Australian universities, yet it's only ten years old. No wonder we feel like we're still working out what it is we're even trying to do...

Choice overload: Finding the right tool for the job (conference)

A great piece from Sae Ra Germaine about choosing a platform for online or hybrid conferences. I've spoken with Sae Ra a lot about what matters in running conferences over the years, and I really like her focus on the three important things of people to speak, people to participate, and somewhere for them to come together. One might argue that this could be simplified to simply “a community that wants to meet, and a place and time for them to meet”. A lot of conferences have become huge behemoths with conference committees focussing on bells and whistles at great expense, but the basics are pretty simple.

Apple's Vision + The Cost of Forever

I share this not so much for the Apple part, but rather the “cost of forever”. Dan Cohen (Dean of the Northeastern University Library) shares some numbers and some provocations about how to spend a mythical USD$200 million. I guess all these figures are dependent on how we define “forever”.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

A Mismatched Group of Items That I Would Not Find Particularly Interesting: Challenges and Opportunities with Digital Exhibits and Collections Labels

This week's first article comes from Evidence based library and information practice and I admit I mostly chose to read this based on the amusing title. It turns out this is a great practical study of a problem that comes up a lot in my work – what to name menu items in a way that is neither too vague nor too specific and jargony. I'll be reading this one again to remind myself of the interesting combination of card sorts and surveys they used, and the simplicity of their study design (send out a link in regular newsletters that are already sent to the target audience).

Hidden Inequities of Access: Document Accessibility in an Aggregated Database

Next up, a horrifying article with an unfortunately vague title. By “horrifying” I mean the results, rather than the paper itself. The study is limited and uses random sampling, but it seems very likely that it would be simple to replicate the findings in other corpora.

This is a study of the accessibility (in the disability sense) of a selection of highly-cited journals from EBSCO’s Library & Information Source database. It's absolutely damning of a profession that claims to champion equitable access to information.

Fewer than half (48%) of the articles overall included an HTML format option.

Which is a pretty big problem since well-marked-up HTML is about the most accessible format you can use when it comes to screen readers. But it somehow gets worse:

Findings from the audit of PDF accessibility showed that 100% of the PDF articles (N= 120) from this study’s original sample failed the minimum standard of PDF/UA accessibility of containing a tagged structure.

That is, the entire sample of article PDFs was inaccessible without running them through third party software to attempt remediation. As the authors state, this is not the fault of EBSCO but rather of the journal publishers. The irony of this article is that it is only available in PDF format – ITAL doesn't publish articles in HTML.

Taking Control of Our Data: A Discussion Paper on Indigenous Data Governance for Aboriginal and Torres Strait Islander People and Communities

Finally, an interesting paper from the Lowitja Institute. This helped clarify for me what is actually meant by “Indigenous Data Sovereignty”, “Indigenous Data Governance”, and how they are different. I really like the way this discussion paper is framed in part as a practical guide for local Aboriginal and Torres Strait Islander communities and organisations. The case studies and planning guides make the theory more tangible.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

Ending profiteering from publicly-funded research

This is a mildly interesting and somewhat confused report from the Australia Institute. It's interesting in that the absurdity of scholarly publishing is now becoming a salient public policy debating point in Australia. Unfortunately the report itself offers up a fairly standard grab-bag of “solutions” that don't address the problems it accurately identifies, and don't address the incentives and pressures on researchers that drive the current system. It's all very well to identify that RELX and friends are “greedy” (they would say “maximising shareholder value”), but “greedy publishers” isn't much of an analysis of what drives the system.

Producing more but understanding less: The risks of AI for scientific research

May contain traces of the infamous “Giant rat dick” illustration. You have been warned.

Many AI tools reawaken the myth that there can be an objective standpoint-free science in the form of the “objective” AI. But these AI tools don't come from nowhere. They're not a view from nowhere. They're a view from a very particular somewhere. And that somewhere embeds the standpoint of those who create these AI tools: a very narrow set of disciplinary expertise—computer scientists, machine learning experts. Any knowledge we ask from these tools is reinforcing that single standpoint, but it's pretending as if that standpoint doesn't exist.

Indexing the information age / The birth of our system for describing web content

Over a weekend in 1995, a small group gathered in Ohio to unleash the power of the internet by making it navigable

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

Digital Scholarly Journals Are Poorly Preserved: A Study of 7 Million Articles

From the Journal of Librarianship and Scholarly Communication. There are some limitations to this article, but on the other hand these limitations mostly serve to identify the difficulties of discovery when it comes to archiving scholarly literature: the study didn't consider institutional archives, presumably because there's no easy way to know whether any given article is stored in one. This would be the same problem should someone want to find an archived copy of something when the DOI no longer resolves. The tl;dr is right there in the title, (for which the authors should be congratulated).

Some things to consider when deciding whether to start building with “AI” in libraries and archives

Ed Summers said elsewhere that he got some pushback from colleagues for this short talk, but I'm personally very grateful that he published it, as it will probably form some of the basis for work I will be doing in the first half of this year to work out a framework for how we assess the various “AI” discovery tools that will increasingly infest librarianship and academia.

Call for Proposals for JLSC Special Issue: Open Access: Diverse Experiences and Expectations

I didn't mean to have two links related to JLSC but it just kind of happened. Anyway, proposals are due by 5 April.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

LEARNING LESSONS FROM THE CYBER-ATTACK; British Library cyber incident review

On Friday/Saturday the British Library released a “lessons learned” report into the October ransomware attack that severely damaged their ability to function and is still affecting operations. Whilst understandably a lot of detail is missing, the report is a useful document for anyone interested in cultural memory institutions, government services, organisational cyber security, or public policy.

I've seen a few takes on this on Mastodon, mostly along the lines of blaming the BL for poor security practices and culture. The report admits that their culture was not as security-focussed as it should have been, however I have a different view on this. I'll possibly provide my own lukewarm take in blog form at some point, but in essence I think this highlights a significant problem within all cultural institutions where there is a clash of organisational cultures and values between a short-term focussed future-looking information technology industry and a primarily long-term focussed, past-looking knowledge management profession. I'm probably betraying my bias with how I describe the two, but one doesn't have to view one side of the relationship as inherently better/smarter to understand how this difference in outlook causes problems.

Artificial Intelligence Blog Series: Introducing Our AI Metadata Generator

Wait, did somebody say “tech bros”? In hindsight we probably should be surprised that Ex Libris have taken this long to release an “AI-generated catalogue records” product. They're being cautious with this initially, pitching it as an “enhancement” tool, but I'd say it's pretty clear where they want this to go.

I think it's significant that this product is aimed at the “Alma Community Zone” (i.e. library-created records) rather than vendor-supplied records. The latter are by far the most complained-about and likely to be completely borked “at scale” as they like to say, but there's more long-term money in convincing libraries that they don't need to hire cataloguers any more.

Australia’s chief scientist takes on the journal publishers gatekeeping knowledge

Rounding out our week of depressing libraries and learning news, Chief Scientist Cathy Foley has come up with a great idea for increasing the profits of the North-Atlantic investment companies holding scientific knowledge ransom. According to the Grauniad she will “take on” the journal publishers by the cunning wheeze of offering them a big cheque to keep operating in exactly the same way but with more of the Australian public's money, paid to them directly instead of being laundered via universities. Apparently Elsevier couldn't say “yes” fast enough. Read to the end to get some good takes from people who actually understand what's at stake here and try to ignore Dr Foley's patronising comments about it being “threatening for some”. The entrenched interests here are the people whose pockets she wants to fill with your money.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

Risks to knowledge economies

A long piece from Helen Beetham on Generative AI and higher education. There's a lot to unpack here but this part in particular struck me as important for academic librarians to grapple with:

Guidance from UNESCO, the UK Government, every university, and even OpenAI itself, is for students to ‘check with other sources of information’ before relying on synthetic text. But as synthetic text become the interface of choice for search, and as the results of search become more and more likely to be synthetic, it becomes difficult to see how this is going to work. Or at least, how students are supposed to operationalise this advice using their own tools and resources. What I think students may be hearing, from all the contradictory advice they are getting, is ‘keep trying until you get something that sounds right’.

Utah teachers could be criminally liable if banned books are found in their classrooms, new bill proposes

It's all there in the headline. The Fascist States of America continue to unravel.

Australian Universities Accord Final Report

It remains to be seen whether all or any of the recommendations here will be implemented. But if they are, the Universities Accord has the potential to fix a large number of structural problems in the Australian university system.

There's nothing here about the rampant casualisation and wage-theft that characterises higher education labour relations, nor anything to fix the absurdities of scholarly publishing, but it's a good start.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

I remember thinking around this time last year that once everyone stopped freaking out, the “pivot to AI” might just be seen as the beginning of a new era for librarianship. Or rather, the end of a period that has lasted most of my career, where the profession seemed to have given up on ever having anything to say on its own behalf, with professional bodies and institutional leaders retreating to a position of merely claiming that whatever the powerbrokers of the world wanted, we could help them either deliver their goals or clean up the mess they left behind them. I'm still feeling confident about that prediction, but maybe ask me in another year and we'll see where we landed.

Whose ethics? Whose AI?

Helen Beetham has a great newsletter called Imperfect Offerings, and she's delivered some real doozies lately. This is the first offering for 2024, and is a kind of reply to the replies to a keynote speech she gave. Whilst Beetham is speaking to a presumed audience of university educators, she has a lot of interesting things to say to librarians.

I feel we should probably avoid ‘prompt engineering’ as a term, and definitely stop selling it as an important skill. I think it will be about as relevant to graduate employment as writing html code, and for the same reasons. Alongside all the ‘100 best GPT prompts’ you can cut and paste from the internet, the ability to call up ChatGPT (or another model) is already being integrated into search engines and browser extensions and thousands of intermediary apps. They offer drop-down lists or push-button choices, or helpfully assume what it is you need to know. What I think we probably should do, working with our colleagues in libraries and study skills centres, is to update our support for search skills. Help students to understand what the algorithms are hiding as well as what they are revealing, how to search when you know what you are looking for as well as when you don’t, the business models as well as the algorithms of search, and how search online is being systematically degraded both by commercial interests and by these new synthetic capabilities.

Generative Artificial Intelligence: 8 Critical Questions for Libraries

A team from Oregon State University have put together this great paper which is to some extent an example of the sort of thing Beetham is talking about: finding the right questions to ask about machine learning and automation, rather than a series of “one weird trick”. In this case the questions are for librarians:

  1. Why AI?
  2. How can libraries make informed decisions without undue urgency?
  3. How can libraries engage in ways that align with library values?
  4. What policies are in place?
  5. How can libraries support the exploration of AI tools by staff and users?
  6. How does AI fit into library information literacy instruction?
  7. Moving forward, how can libraries lead conversations and collaborate?
  8. So what?

A Genealogy of Open

Betsy Yoon provides a history lesson, an interrogation, and ultimately a challenge regarding what we call Open Education. I hadn't realised until I read this that this is not just a rhetorical link between Free & Open Source Software, and Open Educational Practice, but also a more direct connection. And as Yoon shows, all the arguments between and dynamics of the Free Software movement and Open Source advocates are applicable to Open Education. It's definitely worth a read for anyone working in the Open Education space.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

This week we are leaning heavily into the technology side of libraries, and learning when to use the term “learning”.

I will dropkick you if you refer to an LLM as a Librarian

I don't love Mita Williams' rather violent headline, but this is a pretty interesting piece that I definitely need to go back to so I can read all her references. Williams' makes the case for specificity of language when talking about computer automation, large language models and so on. Can a machine really “learn”? Can intelligence be artificial? She's sceptical.

On the matter of the British Library cyber incident

Ciaran Martin is the former head of the UK National Cyber Security Centre. This is a medium-depth dive into what happened at the British Library when they were hit by a “cyber incident”, and what can be learned from it. There are a lot of lessons for institutions like national and state libraries, and universities – like the British Library the effect of them being taken offline by a cyber attack is not obviously life threatening in the short term, but would have both a wide impact and cause a significant number of people to simply be unable to do their jobs.

No data? No problem! Undisclosed tinkering in Excel behind economics paper

Finally, from Retraction Watch the most bonkers story of the week. Swedish economics professor Almas Heshmati straight up just admitted to a PhD candidate who asked him that he had completely made up 10% of the figures in an international comparative study, including simply copying the data for Netherlands across to the next row and pretending New Zealand had recorded the same figures. His excuse? This is totally normal behaviour in the field of Economics 🙃.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

The Art and Science of Speaking (into Microphones) at Conferences

John Dewees has a great post on a single topic that he helpfully explains in his blog title:

Look at this person talking into the microphone at a distance of 6 to 12 inches pointed directly at their mouth. 10 out of friggin 10, absolutely no notes, utterly nailing it.

That's about 20cm for those of us who don't walk around saying things like “I walked a few leagues to the drug store to buy a stick of milk and a quart of butter”. Anyway, this is a great post that stays on message and has a strong message for Bob who refuses to work out how to switch the mic on:

If you refuse to do the above, or worse yet, refuse to use a microphone at all while speaking to a group, you are creating an inaccessible environment where almost certainly some members of the audience won’t be able to hear what you are trying to impart. ...If you aren’t interested in making sure everyone in the room can hear you by leveraging basic principles of accessibility and universal design than I’ll just assume what you have to say isn’t that important.


Ok this is kind of self-indulgent. I've rewritten by “Aus GLAM Blogs” app and now you can also register events, calls for papers, newsletters (like this one), and groups (elists, Discord groups etc), in addition to blogs.

Newsletters and blogs run on RSS in order to fetch the latest posts and editions, and there's also an RSS feed for each type (and combined) in addition to a Mastodon bot. Check it out!

Oh, also my apologies to those whose blogs weren't announced by the bot whilst I ironed out a few bugs. People like Emilia Bell...

Finding purpose & potential in library impact assessment

This piece from Emilia Bell is interesting on the perennial battle in libraries to actually assess what we're doing rather than simply collecting data hoping to prove that what we're doing is great:

Without a purpose for impact assessment, evidence collected on library services with a relational and connected quality (especially in teaching and learning) risks becoming transactional. The questions that drive the process of evidence-based practice start becoming more extractive, fixated on what evidence we can get (ahem, collect), rather than the community this evidence and impact benefits.

I have seen this dynamic everywhere I've ever worked. Often it seems to happen because the purpose of a program is assumed to be so obvious as to not require definition. Sometimes I've wondered whether it's because the librarians didn't really want to know whether their work was effective. And sometimes it seems to come from a genuine belief – as Emilia gestures to here – that there are aspects of our work that are simply intrinsically valuable to society and library users, and they can't be adequately captured in the form of measurable data.

Libraries and Learning Links of the Week is published every week by Hugh Rundle.

Subscribe by following on the fediverse or sign up for email below.

Enter your email to subscribe to updates.