Post

Jed Brown
@jedbrown.org

Our labor makes the companies more powerful as they pursue epistemicide and fascist goals including sabotage of higher ed, degrading our working conditions, and the education of students. Explicit refusal sets a norm and helps organize in pursuit of a liberatory future not owned by them.

Jed Brown@jedbrown.org

Yes, AI companies using residential proxies incurs a political price. Evasion creates a paper trail. Linking their behavior to that of malware is useful when talking to university administrators, lawmakers, and regulators. And people feel more violated when they have affirmatively said no.

Jed Brown
@jedbrown.org

It also makes it just a bit harder for the companies to ship features that impersonate us without consent. This specific product was taken offline after a class action lawsuit by @juliaangwin.com, but there are other attempts, and every one of these companies is thirsty to further erase consent.

nilay patel@reckless.bsky.social

Here’s my interview with Shishir Mehotra, the CEO behind Grammarly’s “expert review” feature which attributed writing advice to people - including me lol - without permission. Or, as you will hear us talk about a lot, compensation. www.theverge.com/podcast/8987...

Kristina Lerman
@kristinalerman.bsky.social

I agree that impersonating specific (living) individuals without permission is not ok. But I don’t speak your language. “Epistemicide”? “Liberatory future”? What do these words even mean? I have a hypothesis: the people most anti-AI are those privileged to grow up speaking English. You fit the bill

Jed Brown
@jedbrown.org

"Epistemicide" is widely used in sociology and information science, and has a Wikipedia page, etc. It was coined by a Portuguese sociologist and came into my lexicon via a Cypriot colleague. Two further resources with more depth on the subject:

nyslibrary.libguides.comResources: DEI Toolkit: EpistemicideNew York State Office of Cultural Education
Iris van Rooij 💭@irisvanrooij.bsky.social

AI slop and the destruction of knowledge irisvanrooijcogsci.com/2025/08/12/a...

irisvanrooijcogsci.comAI slop and the destruction of knowledgeThis week I was looking for info on what cognitive scientists mean when they speak of ‘domain-general’ cognition. I was curious, because the nuances are relevant for something I am researching at t…
Caede
@gasparin.bsky.social

Yeah, I dont speak english as a mother tongue but I understand epistemos and I understand -cidium so maybe your hipothesis should change to: the people most anti-AI are those who are highly trained/educated. Do you fit the bill?

Sylvia Wenmackers
@sylviawenmackers.be

I like the initiative. It's a pity the webpage looks like it was made with AI though. (Generic look and things like "set this value before going live" in the source. 😬)

Johan Schuijt
@resolve.works

The problems with these sort of protections is that it's impossible to see a difference between somebody that uses a local coding agent to fetch a paper and somebody that is scraping the whole. the whole internet for all the information there.

Thomas Dietterich
@tdietterich.bsky.social

3. A substantial fraction of papers posted to ArXiv are eventually published in journals. Most of them have sold access to AI companies. Will restricting ArXiv access make any difference?

debbie-leung-q.bsky.social
@debbie-leung-q.bsky.social

Problem is LLMs not citing our work when using them, not the arXiv giving them access. We write papers, want them read and used! Not a bad thing if our work benefits other researchers, even though they first learn about it through AI. Of course AI / downstream users should cite properly.

Yash Jain
@earanyash.bsky.social

Worth signing, though arXiv papers are already public preprints, so blocking crawlers mostly stops training on the raw text/PDF, not citation or metadata scraping. Still a reasonable opt-in ask.