Eryk Salvaggio

Live-read of the Magnifica Humanitas part two: This section starts by covering truth, work and freedom (and unspoken: communication). Figured it will probably need its own thread. www.vatican.va/content/leo-...

Encyclical Letter of His Holiness Leo XIV Magnifica Humanitas (15 May 2026) vatican.va

Anthropic’s response and influence on the encyclical is most evident in the claim that “even its creators don’t understand how it works.” In June, I wrote about what that actually means & why Anthropic says it. (Written while I was in Rome, FWIW). www.techpolicy.press/the-blac…

The Black Box Myth: What the Industry Pretends Not to Know About AI | TechPolicy.Press techpolicy.press

“Whether AGI will arrive matters less than what waiting for it already organizes.” link.springer.com/article/10.1...

Waiting for AGI | AI & SOCIETY | Springer Nature Link link.springer.com

Save the dates: “Noisy Systems: Aesthetics, Epistemology & Computation” examines noise as a position of reference across technical, social & cultural domains, including machine learning, critical AI, media theory / archaeology, art history, philosophy & practice. www.biblhertz…

biblhertz.it

The Machine Visual Culture Research Group has some exciting predoctoral fellowships (a PhD, basically) through UZH. Come live in Zurich, meet in Rome and do some cool critical AI & digital art history with us! Deadline is VERY SOON.

biblhertz.it

After a newsletter hiatus, sharing some questions that have come up around video diffusion models and what seems to be a phenomenon of a "dead frame": images that offer no further possible movement to the machine, which, as a result, cannot animate them. mail.cyberneticforests…

mail.cyberneticforests.com

From the LA Times, 1986, an article on expert systems (the kids were calling it "AI"). Skeptics dismiss it as "so much industry hyperbole." www.latimes.com/archives/la-...

latimes.com

The reason people like the Pope’s AI missive is because there is simply no other institution has taken the side of humanity in the humanities sense. I’m not Catholic, but I am a human who cares about the human mind and the poetics Catholics call a soul.

Lots of people concerned about data theft use AI detection tools like Pangram. I’m curious, does anyone know where Pangram’s training data comes from? Its white paper says “public” data, which sounds to me like “publicly available,” which OpenAI said when it meant “anything we can get our hands on.”

Cardinal says "the encyclical is not about AI, it's about the human condition." I see, "it's not x, it's y," sparks a 5-second psychosis: cannot tell if he's exhaling the AI language virus. Moral panic in my body. Covid memories of animated air particle graphics. Bad news: this is already a disease

A frustrating aspect of writing about LLMs is that any interesting question will be framed as stating that I am for or against AI consciousness; for / against intelligence; or for / against utility. Sometimes I just have a question!

We gotta start talking about the use of LLMs in the peer review as well as in the submissions.

There is a line of posthuman critique (even the sympathetic “decenter rather than replace the human” strand) that holds a weird view of AI as if it is from nowhere: emphasizing that the images and text are unconstrained as if they aren’t specifically optimized to produce a specific form of object

It's striking to go from the conservative institution of the Catholic church discussing real matters of human dignity in the face of AI, and then see a call from a major cultural org seeking artists to develop new metrics that facilitate an equitable value distribution across sectors

The demands on my year-old 1200-word tech policy press piece so far: a) it should have fully explained all of the technical details of interpretability research at Anthropic b) I should have produced a universally agreed-upon definition of morality and reason