Public Diffusion Private Beta, a core Stable Diffusion AI image-generator rival which is totally trained on “public domain and CC0 images”. Only. And it has also (according to the maker) never seen any manga or anime etc. Apparently to be fully open source.

The Prado museum in Spain freely shares 6,000 art history books and nearly as many magazines. Regrettably, their website is so painfully slow that I wasn’t able to test it.

From COAR, a handy new one-page Directory of Open Access Preprint Repositories (DOAPR). These are not being added en masse to JURN, I should perhaps add.

New in open source, Tagasuarus for annotating a locally-stored collection of media and documents. Images, audio, video, PDFs. The annotations are then stored with the files, and can then be searched across. No full AI to add rich image classifications, as yet. But there is a limited AI that can “automatically apply emotions [tags] when the images/videos have faces”.

New in open access, the article “Beyond borders: Examining the role of national learned societies in the social sciences and humanities”. Actually, not proper open access, as it appears you need to “log in” to Wiley to download or print (neither works for me without a log in). Thankfully Google Scholar is on the case, and having none of that nonsense — searching for the title there gets you a download link. The article is based on a spring 2020 survey of 194 learned societies in the UK and Europe. One finding is that…

Contrary to previous research, most SSH societies in our study have not undergone significant changes in the past 5 years, challenging expectations of their declining role.

Chuck Jones has updated his list of Ancient World Open Access Journals that are active and open to submissions.

Run private offline ‘chat’ AIs on Windows, without wasting a day on a frustratingly complex and fiddly Python-based setup and install. The free LM Studio makes discovery, downloading and running chat LLM AI’s as easy as installing any other Windows software. A good graphics card is required to run your selected AI chat. Note also that the LM Studio wrapper requires a PC with a processor that supports AVX2, if not one with a dedicated NPU. There’s also a beta release that supports older plain AVX processors.