Eryk Salvaggio
@eryk.bsky.social
Situationist Cybernetics. Gates Scholar researching AI’s impacts on the Humanities, University of Cambridge. Affiliated Researcher, Machine Visual Culture Research Group (Max Planck Institute, Rome). Aim to be kind. cyberneticforests.com
Pinned
Many critical AI resources are written for teachers, rather than students. I wrote this for undergrad students dealing with “AI literacy” in the humanities, hoping to raise more fundamental questions about what it means to write. People have added it to syllabi, so maybe you’ll find it useful, too.
Human Literacy
Something I Can Tell Students Now That I Am Not Teaching
You and I probably both keep hearing that students should be working toward AI literacy. That you should know what to type into prompt windo...
mail.cyberneticforests.com
Here’s a thread of three resources I’ve written that people have made good use of — esp. in undergrad and master’s curriculums linking humanities, media studies, and critical AI.
Finally, I will be free from the unwanted labor of leisure
Sony has patented an AI gaming ghost that will play PlayStation games for you when you get stuck www.eurogamer.net/sony-has-pat...
January 7, 2026 at 3:48 PM
Finally, I will be free from the unwanted labor of leisure
Who are the experts / journalists / researchers you turn to for critical perspectives on AI & the tech world? I created a feed of experts who watch AI developments with a skeptical & critical (but *informed*) POV. I'd love more people to follow it, but I'm also looking to add people - recs welcome!
January 7, 2026 at 3:32 PM
Who are the experts / journalists / researchers you turn to for critical perspectives on AI & the tech world? I created a feed of experts who watch AI developments with a skeptical & critical (but *informed*) POV. I'd love more people to follow it, but I'm also looking to add people - recs welcome!
Every year I come to understand that the issue of "anthropomorphizing" AI has confused the distinction between language and "intelligence" in ways that have lead people to deeply misplaced faith about what this tech is actually capable of.
Anthropomorphizing language can be cute when applied to your favorite car, but it helps to muddy the discourse when applied to tech sold as "AI". New from me & @nannainie.bsky.social on @techpolicypress.bsky.social -- how to spot & revise away from anthropomorphizing language applied to "AI"
We Need to Talk About How We Talk About 'AI' | TechPolicy.Press
We share a responsibility to create and use empowering metaphors rather than misleading language, write Emily M. Bender and Nanna Inie.
www.techpolicy.press
January 7, 2026 at 3:25 PM
Every year I come to understand that the issue of "anthropomorphizing" AI has confused the distinction between language and "intelligence" in ways that have lead people to deeply misplaced faith about what this tech is actually capable of.
Reposted by Eryk Salvaggio
Anthropomorphizing language can be cute when applied to your favorite car, but it helps to muddy the discourse when applied to tech sold as "AI". New from me & @nannainie.bsky.social on @techpolicypress.bsky.social -- how to spot & revise away from anthropomorphizing language applied to "AI"
We Need to Talk About How We Talk About 'AI' | TechPolicy.Press
We share a responsibility to create and use empowering metaphors rather than misleading language, write Emily M. Bender and Nanna Inie.
www.techpolicy.press
January 7, 2026 at 2:38 PM
Anthropomorphizing language can be cute when applied to your favorite car, but it helps to muddy the discourse when applied to tech sold as "AI". New from me & @nannainie.bsky.social on @techpolicypress.bsky.social -- how to spot & revise away from anthropomorphizing language applied to "AI"
Reposted by Eryk Salvaggio
"Anthropomorphizing language...portrays a world view in which the people responsible for developing the systems are not held accountable for the system’s inaccurate, inappropriate, and sometimes deadly output,” write @emilymbender.bsky.social and @nannainie.bsky.social.
We Need to Talk About How We Talk About 'AI' | TechPolicy.Press
We share a responsibility to create and use empowering metaphors rather than misleading language, write Emily M. Bender and Nanna Inie.
www.techpolicy.press
January 7, 2026 at 2:58 PM
"Anthropomorphizing language...portrays a world view in which the people responsible for developing the systems are not held accountable for the system’s inaccurate, inappropriate, and sometimes deadly output,” write @emilymbender.bsky.social and @nannainie.bsky.social.
Reposted by Eryk Salvaggio
I have started collecting AI-generated anti-anti-AI propaganda that gets posted to Reddit, this one is truly a masterwork
January 7, 2026 at 4:00 AM
I have started collecting AI-generated anti-anti-AI propaganda that gets posted to Reddit, this one is truly a masterwork
I have started collecting AI-generated anti-anti-AI propaganda that gets posted to Reddit, this one is truly a masterwork
January 7, 2026 at 4:00 AM
I have started collecting AI-generated anti-anti-AI propaganda that gets posted to Reddit, this one is truly a masterwork
Counterculture is associated w/ “nonconformist” but cool has always been about becoming the self that would be if you didn’t concern yourself with the conformist/nonconformist tension. There’s nothing less cool than trying to *prove* you don’t care.
January 7, 2026 at 3:11 AM
Counterculture is associated w/ “nonconformist” but cool has always been about becoming the self that would be if you didn’t concern yourself with the conformist/nonconformist tension. There’s nothing less cool than trying to *prove* you don’t care.
Just glanced at Grok's system prompt. Grok is told to handle user requests as if 'teenage or girl does not necessarily imply underage.' While the instructions discourage "Creating or distributing child sexual abuse material," it's after "teenage" and "girl" are redefined as not being underage.
January 6, 2026 at 3:33 PM
Just glanced at Grok's system prompt. Grok is told to handle user requests as if 'teenage or girl does not necessarily imply underage.' While the instructions discourage "Creating or distributing child sexual abuse material," it's after "teenage" and "girl" are redefined as not being underage.
Reposted by Eryk Salvaggio
Something way more people need to internalize and help their friends understand is that the *only* politics and culture that you can reproduce using pile-ons, networked harassment, doxxing, death threats and the like is a fascist one. It doesn’t matter what you think your motivations or politics are
Losers decided to do gamergate but pretend wokely 🤦
January 6, 2026 at 3:57 AM
Something way more people need to internalize and help their friends understand is that the *only* politics and culture that you can reproduce using pile-ons, networked harassment, doxxing, death threats and the like is a fascist one. It doesn’t matter what you think your motivations or politics are
Reposted by Eryk Salvaggio
How I Built a Tool to Detect AI-Generated Fake References
dev.to/lidianycs/ho...
dev.to/lidianycs/ho...
How I Built a Tool to Detect AI-Generated Fake References
Large Language Models(LLMs) have become part of everyday academic and technical writing. But there is...
dev.to
January 5, 2026 at 6:24 PM
How I Built a Tool to Detect AI-Generated Fake References
dev.to/lidianycs/ho...
dev.to/lidianycs/ho...
I feel like BlueSky was seeded by the problem-solving activist set, for whom — to its credit — social media has always been imagined as a strategy room for the project of a better world, and so the site constantly feels like we are all huddled around a table with a giant map of Today’s Problem.
January 6, 2026 at 3:24 AM
I feel like BlueSky was seeded by the problem-solving activist set, for whom — to its credit — social media has always been imagined as a strategy room for the project of a better world, and so the site constantly feels like we are all huddled around a table with a giant map of Today’s Problem.
Reposted by Eryk Salvaggio
🚨Calling all new media scholars🚨
If (and I emphasize if) a 2nd edition of the New Media Reader happened to be in the works, what additions would you like to see? What do you think could be excised? @docmofo.bsky.social @n-w-f.bsky.social @psssssssss.bsky.social
mitpress.mit.edu/978026223227...
If (and I emphasize if) a 2nd edition of the New Media Reader happened to be in the works, what additions would you like to see? What do you think could be excised? @docmofo.bsky.social @n-w-f.bsky.social @psssssssss.bsky.social
mitpress.mit.edu/978026223227...
The New Media Reader
This reader collects the texts, videos, and computer programs—many of them now almost impossible to find—that chronicle the history and form the foundati...
mitpress.mit.edu
January 5, 2026 at 6:15 PM
🚨Calling all new media scholars🚨
If (and I emphasize if) a 2nd edition of the New Media Reader happened to be in the works, what additions would you like to see? What do you think could be excised? @docmofo.bsky.social @n-w-f.bsky.social @psssssssss.bsky.social
mitpress.mit.edu/978026223227...
If (and I emphasize if) a 2nd edition of the New Media Reader happened to be in the works, what additions would you like to see? What do you think could be excised? @docmofo.bsky.social @n-w-f.bsky.social @psssssssss.bsky.social
mitpress.mit.edu/978026223227...
Reposted by Eryk Salvaggio
Listen to warnings! www.eea.europa.eu/en/analysis/...
Experts were asked to list "the dates of early warnings, to analyse how this information was used, or not used, in reducing hazards, and to describe the resulting costs, benefits and lessons for the future." h/t @jackstilgoe.bsky.social
6/n
Experts were asked to list "the dates of early warnings, to analyse how this information was used, or not used, in reducing hazards, and to describe the resulting costs, benefits and lessons for the future." h/t @jackstilgoe.bsky.social
6/n
Late lessons from early warnings: the precautionary principle 1896-2000
Late lessons from early warnings is about the
gathering of information on the hazards of
human economic activities and its use in
taking action to better protect both the
environment and the health of...
www.eea.europa.eu
December 30, 2025 at 7:00 AM
Listen to warnings! www.eea.europa.eu/en/analysis/...
Experts were asked to list "the dates of early warnings, to analyse how this information was used, or not used, in reducing hazards, and to describe the resulting costs, benefits and lessons for the future." h/t @jackstilgoe.bsky.social
6/n
Experts were asked to list "the dates of early warnings, to analyse how this information was used, or not used, in reducing hazards, and to describe the resulting costs, benefits and lessons for the future." h/t @jackstilgoe.bsky.social
6/n
You really haven’t been able to tell a joke online since Obama left office
January 6, 2026 at 2:11 AM
You really haven’t been able to tell a joke online since Obama left office
Reposted by Eryk Salvaggio
US artist visas are now given based on your number of social media followers
January 5, 2026 at 11:22 PM
US artist visas are now given based on your number of social media followers
Word of the year candidate
Screaming + vomiting: ‘Scromiting’ is the mysterious cannabis syndrome becoming more common
Screaming + vomiting: ‘Scromiting’ is the mysterious cannabis syndrome becoming more common
The condition is rare but a University of Illinois Chicago found the number of diagnosed cases jumped seven-fold from 2016 to 2022.
trib.al
January 5, 2026 at 11:24 PM
Word of the year candidate
Reposted by Eryk Salvaggio
I have never seen the responses to an innocuous post go zero to 60 like this before
January 4, 2026 at 11:00 PM
I have never seen the responses to an innocuous post go zero to 60 like this before
Musk is basically implementing self-deportation policies over at X. He knew he was polarizing, he knew anyone left of a Klansman wasn’t going to pay him for a blue check. So he amplified blue check content, with paid accounts seeing 10x reach over free ones. That’s one piece.
January 5, 2026 at 10:49 PM
Musk is basically implementing self-deportation policies over at X. He knew he was polarizing, he knew anyone left of a Klansman wasn’t going to pay him for a blue check. So he amplified blue check content, with paid accounts seeing 10x reach over free ones. That’s one piece.
Some people are mad about this. I think it’s great. Let kids build stuff and understand sensors as a material that can be repurposed! It seems like a foundational bit of tech literacy that could spur creativity. I’d have used the hell out of this brick. As an artist I still might!
Lego announces Smart Brick, the ‘most significant evolution’ in 50 years
Starting with Lego Star Wars.
www.theverge.com
January 5, 2026 at 7:59 PM
Some people are mad about this. I think it’s great. Let kids build stuff and understand sensors as a material that can be repurposed! It seems like a foundational bit of tech literacy that could spur creativity. I’d have used the hell out of this brick. As an artist I still might!
Reposted by Eryk Salvaggio
If any media outlets want to talk about the closure of the Corporation for Public Broadcasting, I'm currently writing the official history of public meeting for the CPB and Current, and I'm available.
Corporation for Public Broadcasting to Shut Down After 58 Years Due to Trump Eliminating Funding
The Corporation for Public Broadcasting's board voted to close the organization after 58 years after the Trump Administration ended funding.
variety.com
January 5, 2026 at 7:38 PM
If any media outlets want to talk about the closure of the Corporation for Public Broadcasting, I'm currently writing the official history of public meeting for the CPB and Current, and I'm available.
US/UK tech policy in a nutshell: “cross our fingers and hope for the best”
Keir Starmer's spokesman asked if Government will still use X now it posts indecent images of children, replies "we always keep communication channels under review to reach public in most appropriate way possible”
"Operators are under obligation to act appropriately. We always hope they would do"
"Operators are under obligation to act appropriately. We always hope they would do"
January 5, 2026 at 3:08 PM
US/UK tech policy in a nutshell: “cross our fingers and hope for the best”
Reposted by Eryk Salvaggio
The end goal of Grok sexual harassment is to make women afraid to be visible. It’s a punishment for daring to exist in a way the perpetrators can’t control. There’s no one solution to this, because the problem isn’t just technological. It’s cultural. It’s misogyny. This is just one expression of it.
January 5, 2026 at 1:10 AM
The end goal of Grok sexual harassment is to make women afraid to be visible. It’s a punishment for daring to exist in a way the perpetrators can’t control. There’s no one solution to this, because the problem isn’t just technological. It’s cultural. It’s misogyny. This is just one expression of it.
Reposted by Eryk Salvaggio
Bluesky is 10,000 accounts waiting all day for someone to call a lectern a podium.
October 14, 2025 at 2:35 PM
Bluesky is 10,000 accounts waiting all day for someone to call a lectern a podium.
Reposted by Eryk Salvaggio
Here’s a thread of three resources I’ve written that people have made good use of — esp. in undergrad and master’s curriculums linking humanities, media studies, and critical AI.
Many critical AI resources are written for teachers, rather than students. I wrote this for undergrad students dealing with “AI literacy” in the humanities, hoping to raise more fundamental questions about what it means to write. People have added it to syllabi, so maybe you’ll find it useful, too.
Human Literacy
Something I Can Tell Students Now That I Am Not Teaching
You and I probably both keep hearing that students should be working toward AI literacy. That you should know what to type into prompt windo...
mail.cyberneticforests.com
January 4, 2026 at 3:42 PM
Here’s a thread of three resources I’ve written that people have made good use of — esp. in undergrad and master’s curriculums linking humanities, media studies, and critical AI.