Eryk Salvaggio
banner
eryk.bsky.social
Eryk Salvaggio
@eryk.bsky.social
Situationist Cybernetics. Gates Scholar researching AI’s impacts on the Humanities at the University of Cambridge. Tech Policy Press Writing Fellow. Researcher, AI Pedagogies, metaLab (at) Harvard University. Aim to be kind. cyberneticforests.com
London Folks: I’ll be in town for a performance of “Human Movie” (offering real-time narration, as it is intended to be seen!) on December 4, 7-11pm. Additional films & performances by Tim Murray-Brown and Matt Spendlove. Details and tickets here: www.eventbrite.co.uk/e/deep-assig...
Deep Assignments #03
An evening of film screenings, multichannel performance and critical discussion.
www.eventbrite.co.uk
November 19, 2025 at 2:14 PM
Before my talk at Paris 8 tonight, I passed the Lenin apartment plaque. Wondering what Lenin’s apartment would get on AirBnB in 2025
November 17, 2025 at 11:13 PM
So the “pop charts” is complicated but I’ll stand by my point: AI makes some jpgs, everybody shouts about illustration being dead. It writes some text, we get books about what it means to prise words from thought. Code? It’s the end of computer science degrees. Music, though? Crickets.
being No. 1 on the Spotify Viral 50 is *very different* than being No. 1 on "Spotify's Top 50." It's never even cracked the top 200 most played songs in the country on the overall daily charts.
November 17, 2025 at 11:04 PM
AI-generated music on the pop charts seems like a much bigger deal than the collective lack of reaction seems to suggest?
November 17, 2025 at 4:05 PM
This sucks because now kids will have no way to have an imaginary conversation with a teddy bear
"Powered by OpenAI’s GPT-4o model by default...tests repeatedly showed that the AI toy dropped its guardrails the longer a conversation went on, until hitting rock bottom on incredibly disturbing topics."
AI-Powered Stuffed Animal Pulled From Market After Disturbing Interactions With Children
FoloToy says it's suspended sales of its AI-powered teddy bear after researchers found it gave wildly inappropriate and dangerous answers.
futurism.com
November 17, 2025 at 12:03 PM
Reposted by Eryk Salvaggio
"Powered by OpenAI’s GPT-4o model by default...tests repeatedly showed that the AI toy dropped its guardrails the longer a conversation went on, until hitting rock bottom on incredibly disturbing topics."
AI-Powered Stuffed Animal Pulled From Market After Disturbing Interactions With Children
FoloToy says it's suspended sales of its AI-powered teddy bear after researchers found it gave wildly inappropriate and dangerous answers.
futurism.com
November 17, 2025 at 4:00 AM
Reposted by Eryk Salvaggio
“Pluribus” takes the idealized dream of AI: all data linked, diffused agents coordinating and transferring skills, communication perfected as to render writing obsolete, the extreme automation of benevolent intelligence — and shows how, even if achieved perfectly, it’s still a horror story.
November 17, 2025 at 1:05 AM
Reposted by Eryk Salvaggio
> models give unsafe responses because that is not what they are designed to avoid. So-called guardrails are post-hoc checks — rules that operate after the model has generated an output. If a response isn't caught by these rules, it will slip through

www.forbes.com/sites/weskil...

2/n
November 17, 2025 at 5:56 AM
“Pluribus” takes the idealized dream of AI: all data linked, diffused agents coordinating and transferring skills, communication perfected as to render writing obsolete, the extreme automation of benevolent intelligence — and shows how, even if achieved perfectly, it’s still a horror story.
November 17, 2025 at 1:05 AM
I’m speaking tomorrow at Paris VIII University and based on the Wikipedia article, its early years haunt JD Vance’s present-day imagination of, like, Johns Hopkins University
November 16, 2025 at 11:35 PM
Objet Petit AGI
November 16, 2025 at 11:17 PM
Reposted by Eryk Salvaggio
"Ultimately, the collective strategy of AI companies threatens to deskill precisely those people who are essential for society to function. What value, after all, is there in automating art, thinking, or reading – especially in educational and academic settings? None."
“While the AI industry claims its models can “think,” “reason,” and “learn,” their supposed achievements rest on marketing hype and stolen intellectual labor. In reality, AI erodes academic freedom, weakens critical reading, and subordinates the pursuit of knowledge to corporate interests.”
AI Is Hollowing Out Higher Education
Olivia Guest & Iris van Rooij urge teachers and scholars to reject tools that commodify learning, deskill students, and promote illiteracy.
www.project-syndicate.org
November 16, 2025 at 8:15 PM
Reposted by Eryk Salvaggio
The emails have Summers reporting to Epstein about his attempts to date a Harvard economics student & to hit on her during a seminar she was giving.
November 15, 2025 at 9:48 PM
Reposted by Eryk Salvaggio
Not a unique observation here, but yet another AI ad marketed toward lonely dudes
just saw the worst gaming ad ive ever seen
November 16, 2025 at 12:16 PM
Reposted by Eryk Salvaggio
In this week's post I make the case for an #AI #metaphor observatory tracking how we think and talk about and act on GenAI and LLMs makingsciencepublic.com/2025/11/14/m...
Making the case for an AI metaphor observatory
Between 2023 and 2025 I have written various posts on GenAI, Large Language Models and metaphors: one where I went out hunting for metaphors; one where I had a chat with ChatGTP about metaphors for…
makingsciencepublic.com
November 14, 2025 at 7:18 AM
Well put
Musk is a parasite on the future; he preys upon the imagination of others, taking any vision of a better world and depreciating it by hawking options on a transparently fake facsimile.

He is capitalism's full-throated assault on the utopian imaginary.
November 16, 2025 at 8:10 AM
Reposted by Eryk Salvaggio
🚨 New tell someone used ChatGPT
No em dashes
AGI any day now
November 15, 2025 at 11:06 PM
Reposted by Eryk Salvaggio
AGI any day now
November 14, 2025 at 11:48 PM
Reposted by Eryk Salvaggio
Some inspiring words on cinema in the streaming era from the Pope.

Yes, THAT Pope.
November 15, 2025 at 4:26 PM
Reposted by Eryk Salvaggio
“Without human literacy, you might assume that words and their order are all that matter.” I happened on this @eryk.bsky.social post from way back in August and love the move to put human literacies up against AI literacies. mail.cyberneticforests.com/human-litera...
a man in a suit and bow tie asks how did i get here
Alt: Clip from the Talking Heads music video for Once In A Lifetime, showing David Byrne in a suit and bow tie on his knees asking to the heavens with open arms: “How did i get here?”
media.tenor.com
November 15, 2025 at 9:22 PM
Reposted by Eryk Salvaggio
You were incredible, Alice Wong, and it was a blessing to have collaborated with you on projects, to have read your words and to have heard your voice, as full of life and passion and joy and rage as it was.

You made a difference—so much of a difference.

Rest always in peace, and always in power.
November 15, 2025 at 7:34 AM
AGI any day now
November 14, 2025 at 11:48 PM
Reposted by Eryk Salvaggio
Just to say, I am a supervisor here:

UKRI AI Centre for Doctoral Training (CDT) in Responsible and Trustworthy in-the-world Natural Language Processing
www.responsiblenlp.org/2026-student...

If you are reading things like this and planning PhD project, do reach out:
www.ft.com/content/e5b7...
2026 Studentships
www.responsiblenlp.org
November 14, 2025 at 11:31 AM
Reposted by Eryk Salvaggio
I'm attending a workshop hosted by @techpolicypress.bsky.social tomorrow. They invited me to write a "provocation" about the state of technology and democracy right now.

I think I wrote something appropriately dark and foreboding.

www.techpolicy.press/the-dance-wi...
The Dance with Big Tech is Different under Trump 2.0 | TechPolicy.Press
If we are going to repair democratic institutions, we are going to have to do it ourselves, writes Dave Karpf.
www.techpolicy.press
November 13, 2025 at 9:09 PM
Reposted by Eryk Salvaggio
"Archival Images of AI," a guide to making hype-free images of AI, wins Silver & People's Lovie in Art, Culture & Heritage! Congrats @aixdesign.bsky.social, @nadiapiet.bsky.social, @taniaduarte.bsky.social, Ploipailin Flynn, @thedominica.bsky.social! (Delighted to be a contributor & advisor!) ⬇️
Details
winners.lovieawards.com
November 13, 2025 at 3:04 PM