Happy to hear any pointers to publications / blogs / podcasts / videos etc. of institutions in the cultural heritage/GLAM/information science/information infrastrucutre sphere speaking publically about the risks associated.
I'm sure they are out there.
I understand that AI-technology can significantly improve some (not most/all) GLAM processes, allowing us to focus on other things. And I do not deny that there are areas where AI makes sense. I do criticize the fact that from what I can see there are two approaches that (public) information infrastructure institutions / cultural heritage institutions take. EIther:
(a) they're either all over AI research
(b) total silence
Where is the healthy, public ciriticism?
This is the conversation I want to see around AI - in general everywhere, but in particular also in information infrastructure organizations like my own - instead of think-about-how-AI-can-be-included-in-your-daily-work :
"We, the undersigned scientists, contend that both more AI safety research and more emphasis on containment technologies and procedures are needed."
*digipres job opportunity in Australia*
The National Archives of Australia are looking for a new "Director, Digital Archives Preservation and Research". Full-time, permament position, AUS $147,889 - $160,006 pa + 15.4% superannuation. Task are all managerial / strategic. Application closes Nov 1st 2026. Unfortunately only open to Australian citizens. All info here:
https://careers.naa.gov.au/jobs/director-digital-archives-preservation-and-research-various-act-australia
Judging just by the toots here, it looks like there’s a lot of AI mentioning in #nttw10 talks ?
another bi-lingual add to #zenodo #digipres community by Dig. Reserach Alliance of Canada
Slidedeck to a Love Data Week panel, moderated by Alyssa Conlon and feat. Lucia Constanzo, Corey Davis, Mary C. Shea, Julie Shi, @amanda , Adele Torrance, Natalie Vielfaure.
English:
Research Data Preservation: A Love Letter to Our Future Research Communities
https://doi.org/10.5281/zenodo.18714976
French:
Préservation des données de recherche : une lettre d'amour aux communautés de recherche
https://doi.org/10.5281/zenodo.18715052
Auf einer deutschsprachigen Mailingliste ist aktuell die Migration von Friedhofsverwaltungssoftware ein Thema. Die Softwaretools heißen WinFried und HADES X ... und ich erfreue mich sehr wahrscheinlich viel zu sehr an diesen Namen und entwickle Ideen für neue. OSiris. RIPdesk, CHARON2, ...
Very honored that EOSC EDEN Core Preservation Processes is one of 3 finalists for the Digital Preservation Awards 2026 in the "Research and Innovation" Category!
The other two candidates are "AI-Enabled Transcription & Dissemination of Handwritten Documents from the Municipal Archive of La Nucía" & "An OAIS-Compliant Archiving Platform with DNA Connector within OLOS".
If you're curious to find out more and see us pitch the CPPs, check out the video here:
https://www.dpconline.org/news/dpa2026-finalists-research-and-innovation
#DPA2026
As NDSA is moving their publications out of OSF and into Zenodo, they have set up a dedicate NDSA community but are also adding their reports to the #Zenodo #digipres community!
Latest additions to the digipres community by NDSA are:
"2025 NDSA Fixity Survey" by the NDSA Fixity Survey Working Group, available here:
https://doi.org/10.5281/zenodo.22099729
"2025 NDSA Staffing Survey" by the NDSA Staffing Survey Working Group, available here:
https://doi.org/10.5281/zenodo.22003450
Any recommendations for curation/submission policies with good AI usage statements!?
the zenodo digipres community is seeing more submissions of AI generated papers. I'm hesitant to include entirely AI written papers, even if disclosed in authorship note (e.g. "This draft was generated by ChatGPT (OpenAI) in response to a brief supplied by <author name>").
However, I don't want to totally rule out AI usage, as I know it can help non-native speakers with e.g. grammar enhancement.
new @pdfcpu version v0.16.1 out!
Version is a security & stability update with no public API changes.
Some fixes of potential interest to the digipres crowd:
- improves structural validation of embedded files, 3D annotations & multimedia dictionaries
- fixes strict validation of PDF dates ending in z
Release announcement here:
https://github.com/pdfcpu/pdfcpu/releases/tag/v0.16.1
If you're new to pdfcpu and want to know how to get started, see here:
https://pdfcpu.io/getting_started/getting_started/
@Mareike2405 points to the (2019) presentation "The Final Death(s) of Digital Scholarship—An Ongoing Case Study of DH2005 Projects" by Robin Camille Davis to illustrate the short-livedness of project websites post-project
#digiculturebbaw
https://robincamille.com/presentations/death_of_digital_scholarship/
next session at #digiculturebbaw is on "Sustaining / securing reserach data"
with the two presentations:
- @Mareike2405 (Deutsches Historisches Institut Paris) "Fit für das Ende. Strategien für die nachhaltige Transformation digitaler Forschungsprojekte" /
- Alexander Czmiel (BBAW). Wie nachhaltig sind digitale Editionen?
.
Paraskevi is sharing some figures on BnF's digital archive "SPAR = Scalable Preservation and Archiving Repository":
- 8 PB data
- 16.0000 * 2 LTO tapes + disc
- 30 agents at variable capactiy and roles (incl. developers, digital preservation staff, etc.)
#digicultureBBAW
back from lunch break at #digicultureBBAW with Papadopoulou Paraskevi on "Navigating the channels. Orchestrating the digitisation in the french national library (BnF)". You'd figure with @BertrandCaron currently working for TIB, I'd know more about what SPAR actually looks like. but no. i think that's a gap i'll have to work on closing 😆
last presentation of the session is Martin Breuer talking about the German Digital Library (DDB): having been around for 15 years, the federal/state/local imitative includes metadata records for over 66 million digital objects of significance to german cultural heritage. 26 million of these include access to a digital object. 824 German institutions contribute to DDB.
I'm surprised by the rate of ~40% of publically available digital objects. I would've expected that rate to be higher? (...)
i'm also very impressed with the clear reflection of and vison for HathiTrust that Furlough is giving. IMO leadership like this an essential part of making networks successful #digicultureBBAW .
Furlough describes how HathiTrust
membership increased by 40% during COViD (2020-2022) in response to emergency temporary access service to address library closures during the time. He says "This was really a moment where a lot of libraries [who had to close their reading rooms due to the pandemic] realized 'Oh! This is what preservation and digitization is for!'"
Really appreciate all the discussions here at #digicultureBBAW about how access helps make a case for preservation funding.
Mike Furlough is introducing the
#digicultureBBAW audience to HathiTrust: content is digitized books in the volume of 19.5 million items (6 billion pages), saved in 4 copies (1.5 PB per copy). 57% of the content is still in copyright, 94% has been scanned by google in partnership with libraries.
HathiTrust provides preservation and access to the collection to its 215 members in 9 countries. Annual budget for 2027 is 5.1 million USD, 93.1% of that coming from returning membership fees