Hjalmar Fors

@hjalmarfors.bsky.social

Historian of science, medicine & books & lover of all things weird and wonderful. Posts in English and Swedish.

Beheading from a physiological point of view by Frithiof Holmgren (1831-1897). Holmgren is more well known for his discovery of colour blindness. But this here was also a somewhat peculiar research interest of his.

Bild

Here’s a nice detail of the frontispiece of W. H. Ryff, New Kochbuech fuer die Krancken (1545). Hagströmer library, collection of the Swedish pharmaceutical society. Cooking for the sick and infirm in the 16th C!

Bild

During his talk at RBS, James Malin—not on Bsky—showed this title page from a 19th-century catalog for Alfred Nobel’s Nitro-Glycerin Company I’d like to nominate this image for the most metal cut in print history—we need to make something with it for Halloween @skeuomorphpress.org

BildBild

”I went to a French restaurant accompanied by four very well dressed poets. We all ordered steaks with pepper. When the steaks arrived, we rubbed them all over our clothes, soaking ourselves in the sauce. Once this was accomplished, we ordered the same thing again and repeated the act. … ”

By the way, tomorrow #Finland will have been independent for longer than we were a part of the Russian Empire. Finland had been part of Russia for a total of 39,526 days. It will be 39,526 days since independence on Monday, February 23, 2026. suomenkuvalehti.fi/mielipide/ol...

Oletko huomannut? Suomi on pian ollut kauemmin itsenäinen kuin osa Venäjää

Suomen itsenäistymisestä tulee kuluneeksi 39 526 päivää maanantaina 23. helmikuuta 2026, kirjoittaa Ilkka Hemmilä.

suomenkuvalehti.fi

OpenAI ”acknowledged in its own research that LLMs will always produce hallucinations due to fundamental mathematical constraints that cannot be solved through better engineering, marking a significant admission from one of the AI industry’s leading companies.” You can’t trust chatbots.

OpenAI admits AI hallucinations are mathematically inevitable, not just engineering flaws

In a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limi...

computerworld.com