@lastlines.bsky.social

Feedback to these publications is most welcome!

Find all project results here:
mmk.sbb.berlin/deliverables...

#bigdata #ML #culturalheritage #ELSI #digitalculturalheritage

Conclusion by Bloomberg staff writers Parmy Olson and Carolyn Silverman in an article reviewing the financial tally of generative AI impacts

#cost-benefitratio

Bloomberg: The data raise “an uncomfortable prospect: that this supposedly revolutionary technology might never deliver on its promise of broad economic transformation, but instead just concentrate more wealth at the top.“

bloomberg.com/opinion/arti...

See the interesting and literally playful study by Tarek Saier here (in German only):

illdepence.github.io/slf-origins/

doi.org/10.5281/zeno...

Well, we Germans had that back in the 1920‘s: A typographic culture war. ‚Modernists‘ used Antiqua, Nationalists used Fraktur. The resulting effect is that you open a book and think you know the mindset of the person who wrote it; which is complete nonsense…

Happy to be part of the proceedings with our contribution with the paper "How Scalable is Quality Assessment of Text Recognition?"
anthology.ach.org/volumes/vol0...
Our colleague Michał Bubula will present a related poster at the #CHR2025 and be available for questions ... Enjoy!

Yes, and I saw yours. Will be coming soon around the corner with some 4.500 maps from @stabiberlin.bsky.social …

And on another note: How come that these stickers can be found next to each other on the same door?

Thank you so much for these slides, Daniel! They add much to the discussion on „Fifty shades of openness“ ..

bsky.app/profile/jrgl...

Please extend this reading list!

#AITrainingData #Commons #OpenAccess #PublicDomain

@stabiberlin.bsky.social @europeana.bsky.social @bldigischol.bsky.social @nfitzger.glammr.us.ap.brid.gy @miaout.bsky.social @amsichani.bsky.social

Nikhil Kandpal et al.: The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text, June 2025
doi.org/10.48550/arX...

Stefan Baack et al.: Towards Best Practices for Open Datasets for LLM Training, Jan 2025
doi.org/10.48550/arX...

Lukas Gienapp et al.: The German Commons – 154 Billion Tokens of Openly Licensed Text for German Language Models, Oct 2025
doi.org/10.48550/arX...

Pierre-Carl Langlais et al.: Common Corpus: The Largest Collection of Ethical Data for LLM Pre-Training, June 2025
doi.org/10.48550/arX...

Thomas Padilla et al: Public Interest Corpus Principles and Goals, Dec 2025
www.authorsalliance.org/2025/12/03/r...

Paul Keller & Europeana Foundation: Outline for a European Books Data Commons, Nov 2025
openfuture.eu/publication/...