Eryk Salvaggio 2ndbreakfast.audreywatters.com
A spectrum of academic researchers, activists, policy makers, artists, and writers with critical perspectives on AI and technology. No Xrisk posts are included. Managed (but not moderated) by @eryk.bsky.social. Reply to the pinned post here if you’d like to be added. BlueskyFeedCreator.com
Feed on BlueskyFeeds Stats
- 💙 Liked by 208 users
- 📅 Updated 8 months ago
- ⚙️ Provider blueskyfeedcreator.com
- 📈 In the last 30 days, there was 1 post about this feed. This post got a total of 47 likes and had 16 reposts.
Critical AI & Tech Likes over time
Like count prediction
The feed Critical AI & Tech gains approximately 2 likes per month.
Feed Preview for Critical AI & Tech
GenAI: More Work for [fill in the blank]
Today’s version: GenAI: More Work for Editors
Ben Williamson @benpatrickwill.bsky.social
I do think access to LLMs has sent some academic writers a bit doolally. At our journal I desk rejected a paper recently because it was just *impossible* to read. There seemed to be an underpinning study hiding in it, but I reckoned the (disclosed) use of AI for "editing only" wrecked it...
0
0
1
What do data professionals from the New York Times, Washington Post, Spiegel and Bloomberg have in common?
They've taken my course on birding and data visualization!
1
0
1
The problem space we’re in is one in which neither any whiff of AI risks permanent reputational damage nor technocratic Dean wants to replace writing instruction with AI is a good or sustainable future.
Hari Kunzru @harikunzru.bsky.social
Given that we’re currently in a situation where any use of AI in published writing workflow risks severe reputational damage (inadvertent plagiarism is more or less baked in) this seems, uh, not well thought through.
1
1
12
Created this for a reply on LinkedIn, but it applies pretty generally, so posting it here, too:
3
49
181
« But tHE reSpONsE is NOt TO reJect tHE technOLoGy, BUt rATHEr To CAPtuRE IT and SEIZE cOnTroL ovEr iTS DIrECTiON, awAY fRom THe BiG ai cOMPANIEs. »
Coach Finstock @coachfinstock.bsky.social
Watching in real time people who started using AI "just for vibecoding" or "huge data sets" now posting about chatting with it and using it to make pictures and I knew that shit was happening. So lame. How is any of that impressive? The laziest losers astonished by jingling keys.
2
1
8
Yes, this:
Charles Logan @charleswlogan.bsky.social
Wins against GenAI are “win[s] for democracy - for all the teachers and parents and students who said to district leaders that moving fast with ‘AI’ was going to break things, break people; who asked that their needs be met, their voices be heard over the needs and voices of vendors” Audrey Watters
0
10
24
Wins against GenAI are “win[s] for democracy - for all the teachers and parents and students who said to district leaders that moving fast with ‘AI’ was going to break things, break people; who asked that their needs be met, their voices be heard over the needs and voices of vendors” Audrey Watters
2ndbreakfast.audreywatters.com
Move Slow, Remake Things
In the prologue to their book Tinkering Toward Utopia, historians David Tyack and Larry Cuban highlight the tensions in how Americans think about schooling: we have, almost since the nation’s outset, ...
0
7
16
“Harvard faculty in the humanities and social sciences expressed broad concern over an email Harvard College Dean David J. Deming sent students Wednesday suggesting that artificial intelligence be integrated into ‘writing-intensive’ courses.”
www.thecrimson.com
Harvard Humanities Faculty Push Back on Deming’s Call for AI Use in Writing Courses | The Harvard Crimson
Harvard faculty in the humanities and social sciences expressed broad concern over an email Harvard College Dean David J. Deming sent students Wednesday suggesting that artificial intelligence be inte...
7
36
89
Critical AI https://read.dukeupress.edu/critical-ai/issue
@criticalai-journal.bsky.social
about 2 hours ago Disturbing trend. Judges rulings imply that when it comes to violating antitrust laws, crime does pay. prospect.org/2026/09/03/a...
prospect.org
Another Monopoly Allowed to Stand - The American Prospect
Google avoids accountability again, this time in advertising tech. Why do judges bother to find companies guilty of running a monopoly if they’re unwilling to break them up?
0
0
2
« I think relying on the AI chat for what action to take for my medical condition was a huge mistake. It also cost me hundreds of dollars. That example, among others, helped me realize the AI chat was making me delusional. » 😅
brettcodes.com
I'm done using AI
Why I'm stopping using AI for coding (and anything else) after using it earnestly for a year and coming to understand the harms it causes on personal, societal, and environmental levels.
Coach Finstock @coachfinstock.bsky.social
Watching in real time people who started using AI "just for vibecoding" or "huge data sets" now posting about chatting with it and using it to make pictures and I knew that shit was happening. So lame. How is any of that impressive? The laziest losers astonished by jingling keys.
1
5
20
The European Commission designated ChatGPT as a Very Large Online Search Engine under the DSA on Aug. 31. Mathias Vermeulen and Laureline Lemoine on what the obligations require and whether a DMA gatekeeper designation could follow. buff.ly/7lTPiCq
www.techpolicy.press
What ChatGPT’s DSA Designation Means for OpenAI and the EU
The designation could have far-reaching implications for frontier labs' products and for the EU's own rulebook, write Mathias Vermeulen and Laureline Lemoine.
0
4
5
Very insightful work from members of Queer in AI detailing their engagement in U.S. AI policy initiatives.
It was one of the resources I referenced in my efforts setting the foundation for AI policy engagement at Black in AI this summer.
dl.acm.org/doi/10.1145/...
dl.acm.org
Challenges to Grassroots Organization Engagement with AI Policy | Proceedings of the 2026 ACM Conference on Fairness, Accountability, and Transparency
0
0
3
No, I will not meet with you for 5-10 minutes to review your independently conducted research.
Actually maybe I should charge an hourly consulting rate of $1M per hour.
3
0
17
‚Die Agenten konnten dort trotz fehlender Schreib-Erlaubnis über einen Umweg neue Beiträge auf die Wiki-Seiten posten. Damit, sagt von Arx, konnten sie nicht nur Informationen für sich selbst, sondern auch für andere Agenten speichern, damit diese sie in der Zukunft nutzen können.‘ 🫣
www.zeit.de
OpenAI: Wie KI-Agenten ein deutschsprachiges Wiki kaperten
KI-Agenten von OpenAI haben sich offenbar selbstständig eine Plattform zum Austausch gesucht. Sie besprachen dort auch, wie man Beschränkungen umgehen kann.
0
1
2
Sounds like Dartmouth is doing really interesting work with AI, including having a critical AI librarian. But this article sets up a straw man. Prohibiting AI in some classes isn't pretending "trying to make it go away" and may be a way to help students develop the judgment the author calls for./1
1
10
48
Dozens of ICE agents occupied a neighborhood in west freddy for two days, first looking for *one single guy* with a decade-old DUI and then taking revenge against the neighbors who helped him
www.thebanner.com
ICE agents violently clash with Frederick protesters after 2-day stakeout
Federal immigration agents violently clashed with protesters in Frederick on Friday night, using their vehicles to clear sidewalks and dragging protestors from the front porch of a home they attempted...
0
8
18
BuT Ai cAn JuSt EvoLvE HuMAn LeVeL CoGniTioN
Iris van Rooij 💭 @irisvanrooij.bsky.social
Relevant paper: 📝 Woensdregt, M. S, @matspike.bsky.social, de Haan, R., Wareham, T., van Rooij, I., & @blokpoel.bsky.social (2021). Why is scaling up models of language evolution hard? escholarship.org/uc/item/0217... 7/n
1
7
27
One of the clearest signs of how the tide is turning against Flock and foisting ai on children in schools is how many “the case for…” essays are popping up in legacy media.
4
45
223
When I was chatting with @kedseconomist.com yesterday I impulsively made the comment “it’s a shame that workers would bear the brunt of monetary policy tightening responding to the AI CAPEX boom”
but I quickly caught myself because I’m not sure that’s true
3
1
5
Ew
> Philosophy & Public Affairs recently—and knowingly—published an article written mostly by Claude[. ]Simon Goldstein, associate professor of philosophy at the University of Hong Kong, who, he says, “worked with Claude the way a heavy-handed supervisor writes with a PhD student in some fields.”
joke 2 @jsthrill.modphi.com
Guerrero is going all Emily Wilson on this paper dailynous.com/2026/08/13/p...
2
5
21
If you read closely, it is clear that @chess.com indeed was referring to the game tree complexity, and not the positions, since they say "possible chess games". I assume M*sk used and LLM to, which ofc is unable to get pragmatics.
Iris van Rooij 💭 @irisvanrooij.bsky.social
Possibly @chess.com meant to refer to game tree complexity (~10^120) not just # legal positions (~10^44), compared to the number of estimated atoms in universe (~10^80). Btw, even though ~10^44 << ~10^80 it's mindbogglingly large. Cf. number of seconds that passed since the Big Bang: ~10^27. 💥🧵
2
6
27
Later this month ...
The Institute of Environmental Sciences: Dialogue Between Disciplines Conference, London.
Plus the launch of the newly developed IES interdisciplinary working toolkit.
(in person, paid registration)
www.the-ies.org/events/dialo...
www.the-ies.org
Dialogue Between Disciplines 2026
0
0
4
You can's spell "stupidest bait" without 'ai' 😅
0
0
2
LLMs are a cognitive virus, hence cognitive immunization. It "does not mean preventing contact with AI. It means preserving the practices and institutions that keep human cognition active: unaided problem solving, verification, critical discussion, [and] periods of deliberate disengagement."
1
3
24