I’m glad to (finally) see some buyer’s remorse in the corporate ranks about their rapid deployment of #AI, often at the expense of jobs (or so they claim).

Costs are rapidly rising. Software quality is plummeting. Morale is tanking. The workers are getting angry. People are swamped with slop. About the only ones happy with this change are sociopaths who live increasingly solipsistic lives with a sycophantic word extruder (and maybe some jaded senior developers who want to get paid as much as possible while the going is good and no longer care what happens downstream).

I’m hopeful that the end is near. The bubble is deflating. The growth is flatlining. We are finally past the Peak of Inflated Expectations and are now descending into the Trough of Disillusionment. Deployments are getting walked back, people are getting rehired, AI budgets are reduced.

The best time for this to have happened was two years ago. The next best time is now.

https://finance.yahoo.com/technology/ai/articles/exclusive-zuckerberg-says-ai-agent-201123441.html

La litière pour chats en argile, résistante à l'incinération, remplit 7% des décharges suisses
https://www.rts.ch/info/suisse/2026/article/dechets-la-litiere-pour-chats-en-argile-sature-nos-decharges-29290608.html?rts_source=rss_t
La litière minérale pour chats en argile ne brûle pas. Jetée avec les ordures ménagères, elle se retrouve sous forme de résidus dans les usines d'incinération et doit être stockée en décharge. Il existe néanmoins une solution simple.

Videotex und WWW – 2 Schweizer Erfindungen mit unterschiedlichem Erfolg

Höhrenswerte Folge der SRF-Radiosendung «Geschichte» zu den Anfängen des Internets:

https://www.srf.ch/audio/geschichte/die-killer-anwendung-des-internets-wwwer-hat-sie-erfunden?id=AUDI20260630_NR_0005

Finding fake references every year when grading student theses probably shouldn't surprise us when academics do the same. 🥺 I am currently working on Python code (run locally) that helps me check if referencing is consistent throughout a document, and if bibliographic APIs can verify the existence of sources cited. I am tired of manual checks but also don't want to accept that citing non-existent research is the new normal.

https://www.nrc.nl/nieuws/2026/07/01/honderden-nederlandse-wetenschappers-citeren-onbestaande-bronnen-toename-hangt-samen-met-opkomst-kunstmatige-intelligentie-a4931564

In 2026, you cannot freely edit your own posts on neither Bluesky, Threads, or X.

But you can on Mastodon.

Spread the word! ♥️🐘

Specifically coding for type tools: AI has consumed the code and documentation that this type community developed and shared, supported and explained for free. For decades. I recognise the patterns that we designed in AI output. Now you must pay tokens to get to the same thing, and then you pay for a course to raise your professional profile? That is not how skills work, that is how scams work.

“… smart glasses are becoming slimmer, less noticeable […], raising concerns not only about exam integrity, but also about broader privacy risks.”

Maybe I missed it, but so far I haven’t seen much discussion of the risks posed by #smartglasses in academic settings in particular, for both students and instructors.
https://edition.cnn.com/2026/06/26/asia/ai-glasses-cheating-exams-intl-hnk

The program for #COMHUM2026 is out now:

https://wp.unil.ch/llist/en/event/workshop-on-computational-methods-in-the-humanities-2026-comhum-2026/

The Workshop on Computational Methods in the Humanities 2026 will take place September 9–10, 2026 at the University of Lausanne.

#DigitalHumanities #ComputationalHumanities #Unil

Frontiers forces AI onto academic journal editors

return of the rat dck

https://www.youtube.com/watch?v=vXOfALV8a68&list=UU9rJrMVgcXTfa8xuMnbhAEA - video
https://pivottoai.libsyn.com/20260702-frontiers-forces-ai-onto-academic-journal-editors - podcast

time: 4 min 58 sec

https://pivot-to-ai.com/2026/07/02/frontiers-forces-ai-onto-academic-journal-editors/ - blog post

Im Tessiner Bleniotal gibt es eine Reihe von Felsburgen, die «Case dei Pagani», «Heidenhäuser». Diese kaum zugänglichen Ruinen sind der Schauplatz vieler schauriger Geschichten, und sie geben der Forschung bis heute Rätsel auf. Mein neuer Beitrag im Blog des Schweizerischen Nationalmuseums.

https://blog.nationalmuseum.ch/2026/07/case-dei-pagani/

#geschichte #nationalmuseum #landesmuseum #tessin #ticino

New game for Sphere: Scrappy Bird, a mini flappy bird that runs on a 32x16 text screen on the 1975 Sphere system. I made this for the upcoming VCF West. You can play it in the emulator by clicking this link. Enjoy!

https://sphere.computer/emulator?run=bird

🎉 Herzlichen Glückwunsch an Thorsten Wübbena! @ThWuebbena@fedihum.org

Ab heute übernimmt er die erste gemeinsame Professur für „Digitales Kulturelles Erbe“ von Hochschule Mainz und IEG. Die Professur stärkt die Zusammenarbeit in den #DigitalHumanities.
www.ieg-mainz.de/news/thorste...

Thorsten Wübbena auf Professur...

Please help me get the word out to people who might be interested in reading Democracy Americana in German: I hope I can build a readership large enough to carry this experiment of offering my work in two languages (it’s a struggle!).

If you want to read the essay in English, you find it here:

https://steady.page/en/democracyamericana/posts/59b197fa-1a03-44e5-b155-2c221e2a3c0e

⁉️ In order to respond to growing demand and to clarify certain rules, the SNSF is adapting its Regulations on Scientific Exchanges.

🗓️ The new regulations will come into force on 1 January 2027.

A summary of the changes can be found here ➡️ https://sohub.io/apui

Pivot to AI brightens your life — send $5 to keep the stories coming daily

https://pivot-to-ai.com/2026/06/25/pivot-to-ai-brightens-your-life-send-5-to-keep-the-stories-coming-daily/

please forward!

Skatteverket (the Swedish Tax Agency) can definitely hold its own against the IRS when it comes to ridiculous bureaucratic demands.

Mind you, we’re talking about a small one-time honorarium, not a regular salary.

This is an interesting take on why LLMs won't take away all our jobs:

"Large language models are not reasoning machines. They’re plausibility engines."

Interestingly, the author excludes programming jobs, stating that coding "relies on a structured, formal language that can be tested in real time." I agree to some extent, but at the level of specifications and requirements (even if they are formal) this doesn't work anymore either.

https://www.nytimes.com/2026/06/30/opinion/ai-agents-steal-jobs-employment.html?unlocked_article_code=1.uFA.4iI_.o63z9fvtH1Cg&smid=url-share

#nytimes #princeton #llm

🥵

#impôt #taxes

Datenschützer warnen: «Smart Glasses sind ein Frontalangriff auf unsere Privatsphäre.» https://www.watson.ch/!796290848?utm_source=mastodon&utm_medium=social-auto&utm_campaign=auto-share