Lightnews — Scholar-powered news

Reposted by The Distributed AI Research (DAIR) Institute

nathan kim @nathanckim.bsky.social · 3h

Did you know that the university of Michigan is building the world’s largest supercomputer for Los Alamos National Labs’ nuclear weapons production programs, and paying for $830 million of it despite not being able to use >90% of it?

I wrote about it here:

www.michigandaily.com/opinion/umic...

UMich says the Ypsi Data Center won’t “manufacture nuclear weapons.” What does Los Alamos think?

University administration must make a choice: Will it side with the public and live up to its stated ideals of serving the public good and public interest? Or will it side with LANL and spend hundreds...

www.michigandaily.com

1 10 18

Reposted by The Distributed AI Research (DAIR) Institute

BRAID UK @braiduk.bsky.social · 13h

Congratulations to artists Identity 2.0, selected as the first guest zine to be included in the @dairinstitute.bsky.social Zine Library. 'Conversations about resistance and generative AI' can be downloaded here: zines.dair-institute.org/ai-z Image courtesy of the DAIR website.

1 2 3

Reposted by The Distributed AI Research (DAIR) Institute

dylan @dylnbkr.bsky.social · 1d

Saw this article linked by @wonkish.bsky.social in the replies here and wanted to expand on it a little, because this really does feel like a place where a tech solution *is* warranted to help address a major problem. 🧵

www.theverge.com/2024/8/21/24...

A screenshot of the title of an article from The Verge. It reads:

This system can sort real pictures from AI fakes — why aren’t platforms using it?
Big tech companies are backing the C2PA’s authentication standard, but they’re taking too long to put it to use.

by Jess Weatherbed
Aug 21, 2024, 6:00 AM PDT

A screenshot from the linked The Verge article. It reads:

Step one: the industry adopts a standard
A body like C2PA develops an authentication and attribution standard.
Parties across photography, content hosting, and image editing industries agree to the standard.
Step two: creators add credentials
Camera hardware makers offer to embed the credentials.
Editing apps offer to embed the credentials.
Both hardware and software solutions work in tandem to ensure creators can confirm the origins of an image and how / if it’s been altered during edits.
Step three: platforms and viewers check credentials
Online platforms scan for image credentials and visibly flag key information to their users.
Viewers can also access a database to independently check if an image carries credentials.

2 9 20

Reposted by The Distributed AI Research (DAIR) Institute

Hypervisible @hypervisible.blacksky.app · 2d

“…that Sora is being used for stalking and harassment will likely not be an edge case, because deepfaking yourself and others into videos is one of its core selling points.”

Far from an edge case, it’s the primary use case.

Stalker Already Using OpenAI's Sora 2 to Harass Victim

A journalist claims that her stalker used Sora 2, the latest video app from OpenAI, to churn out videos of her.

futurism.com

9 190 370

Reposted by The Distributed AI Research (DAIR) Institute

Anne (Fox Rainbow) @foxrainbow.bsky.social · 2d

My question is in their! I like the answer.

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 6d

Happy 3rd birthday to the Mystery AI Hype Theater 3000 podcast! 🎂

To celebrate, @emilymbender.bsky.social and @alexhanna.bsky.social recorded a special episode to answer listener questions about the show, & reflect on the past & future of AI hype.

Listen here: www.buzzsprout.com/2126417/epis...

1 1 1

Reposted by The Distributed AI Research (DAIR) Institute

Brinkman @meeshdreams.bsky.social · 5d

Heart breaking story, great blog attached.

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 5d

In this possible futures entry, @meron.bsky.social imagines a future where the families of the disappeared find closure.
www.dair-institute.org/blog/eritrea...

Closure for Families of the Missing

Behind every name is a family torn apart, a void where a loved one should be.

www.dair-institute.org

1 1

Reposted by The Distributed AI Research (DAIR) Institute

Lauren Woolsey @cgsunit.bsky.social · 5d

Today's the day for my anti-AI zine volume 2: "Human Perspectives on the Latest AI Hype Cycle" 🎉

Enjoy the fruits of my focus these past few months and learn from many great people!

Scanned zine to print your own and the full text and references are available at padlet.com/laurenUU/antiAI

Front and back cover of the Zine sitting among Japanese maple leaves. Front cover has the title "Human Perspectives on the Latest AI Hype Cycle" with subtitle "AI Sucks and You Should Not Use It, Volume 2"
along with the date of October 2025 and author Lauren Woolsey.

Back cover has the text "References available on the back of this unfolded sheet and at padlet.com/laurenUU/antiAI" along with a QR code to that link. Then it has the text "Share with a friend, light the world! Connect w/ me: @cgsunit.bsky.social"

Pages 2 and 3 of the Zine, open among tree leaves.

Page 2 starts with handwritten "First...some backstory!" and then the text reads as follows: "Version Volume 1 of this zine, (June 2025), is called “Why GenAI Sucks and you should not use it.” I gave copies to my friends, did swaps at Grand Rapids Zine Fest, and shared the digital scan with hundreds of folks. It’s been great to connect with a community of humans who also think AI sucks! Since June, more great folks have added to the conversation. Let me introduce a few here..."

Page 3 is titled Anthony Moser and has the following text: "“I am an AI hater. This is considered rude, but I do not care, because I am a hater.” So opens this most excellent essay (posted August 2025).
You absolutely need to read it. Also, it has 24 linked resources, if my Zine v1.1 list wasn’t enough to get you started being a hater."

Pages 4 and 5 of the Zine, open among tree leaves.

Page 4 is titled Olivia Guest and has the text: "1. Look at Guest’s incredible collection promoting Critical AI Literacy (CAIL): olivia.science/ai . 2. Discover a framework to define AI in “What Does 'Human-Centred AI' Mean?” (July 2025). 3. Share with educator friends Guest et al: “Against the Uncritical Adoption of 'AI' Technologies in Academia” (September 2025). Such a helpful paper for advocacy!"

Page 5 is titled Ali Alkhatib and has the following text: "“AI is an ideological project to shift authority and autonomy away from individuals, towards centralized structures of power.” -from his essay Defining AI. Ali is on my recent radar because he’s starting “AI Skeptics Reading Group” the same month that this Zine launches (October 2025)! If you're a reader, check out the book list on p. 7 here!"

Pages 6 and 7 of the Zine, in partial shadow from tree leaves and surrounded by Japanese maple leaves.

Page 6 is titled Distributed AI Research (DAIR) Institute and has the text: "Great projects DAIR supports: Data Workers Inquiry (work led by Dr. Milagros Miceli), Mystery AI Hype Theater 3000 (by E. Bender and A. Hanna), Possible Futures workshop and Zine series. Timnit Gebru is founder and executive director of DAIR and co-author of the “TESCREAL Bundle” research paper. (Read it!)

Page 7 is titled Further Reading and has a drawn stack of books with the following titles and publication months: Resisting AI (08/22), Blood in the Machine (09/23), The AI Mirror (06/24), Taming Silicon Valley (09/24), Why We Fear AI (03/25), More Everything Forever (04/25), The AI Con (05/25), Empire of AI (05/25). There are notes for The AI Con that the authors run the podcast mentioned on page 6 and that it is the book that the Reading Group from page 5 started on 10/13/25. The page ends with the text "Authors and full titles in reference list!" and a signature from Lauren "Double U."

6 34 95

Reposted by The Distributed AI Research (DAIR) Institute

Olivia Guest · Ολίβια Γκεστ @olivia.science · 4d

New preprint 🌟 Psychology is core to cognitive science, and so it is vital we preserve it from harmful frames. @irisvanrooij.bsky.social & I use our psych and computer science expertise to analyse and craft:

Critical Artificial Intelligence Literacy for Psychologists. doi.org/10.31234/osf...

🧵 1/

Cover page of Guest, O., & van Rooij, I. (2025, October 4). Critical Artificial Intelligence Literacy for Psychologists. https://doi.org/10.31234/osf.io/dkrgj_v1

Table 1 Guest, O., & van Rooij, I. (2025, October 4). Critical Artificial Intelligence Literacy for Psychologists. https://doi.org/10.31234/osf.io/dkrgj_v1

Table 2 Guest, O., & van Rooij, I. (2025, October 4). Critical Artificial Intelligence Literacy for Psychologists. https://doi.org/10.31234/osf.io/dkrgj_v1

8 97 250

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 5d

In this possible futures entry, @meron.bsky.social imagines a future where the families of the disappeared find closure.
www.dair-institute.org/blog/eritrea...

Closure for Families of the Missing

Behind every name is a family torn apart, a void where a loved one should be.

www.dair-institute.org

3 3

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 5d

Today is the 12th anniversary of 369 Eritrean refugees who drowned in Lampedusa exactly 12 years ago. As our researcher refugee advocate Meron Estefanos wrote, one of them was 22 years old and was giving birth as she was drowning. Her body was later found with an umbilical cord attached to her baby.

2 8 29

Reposted by The Distributed AI Research (DAIR) Institute

HumanityExists @humanityexists.bsky.social · 5d

I highly recommend following ALL of em’ critical voices of our time…✌🏼

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 6d

Happy 3rd birthday to the Mystery AI Hype Theater 3000 podcast! 🎂

To celebrate, @emilymbender.bsky.social and @alexhanna.bsky.social recorded a special episode to answer listener questions about the show, & reflect on the past & future of AI hype.

Listen here: www.buzzsprout.com/2126417/epis...

1 2

Reposted by The Distributed AI Research (DAIR) Institute

Teresa Heffernan @tjheffernan.bsky.social · 6d

Infected by AI...we need immunity to this BS. Not a "future" for anyone.

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 6d

Happy 3rd birthday to the Mystery AI Hype Theater 3000 podcast! 🎂

To celebrate, @emilymbender.bsky.social and @alexhanna.bsky.social recorded a special episode to answer listener questions about the show, & reflect on the past & future of AI hype.

Listen here: www.buzzsprout.com/2126417/epis...

2 4

Reposted by The Distributed AI Research (DAIR) Institute

Data & Society @datasociety.bsky.social · 5d

“AI isn’t magic; it’s a pyramid scheme of human labor,” said @adiod.bsky.social, a researcher at @dairinstitute.bsky.social. “These raters are the middle rung: invisible, essential and expendable.” www.theguardian.com/technology/2...

How thousands of ‘overworked, underpaid’ humans train Google’s AI to seem smart

Contracted AI raters describe grueling deadlines, poor pay and opacity around work to make chatbots intelligent

www.theguardian.com

24 48

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 6d

Happy 3rd birthday to the Mystery AI Hype Theater 3000 podcast! 🎂

To celebrate, @emilymbender.bsky.social and @alexhanna.bsky.social recorded a special episode to answer listener questions about the show, & reflect on the past & future of AI hype.

Listen here: www.buzzsprout.com/2126417/epis...

7 30

Reposted by The Distributed AI Research (DAIR) Institute

Emily M. Bender @emilymbender.bsky.social · 6d

As a birthday bonus, here’s one more audiogram for Mystery AI Hype Theater 3000 Ep 63

www.buzzsprout.com/2126417/epis...

w/ @alexhanna.bsky.social
production by Ozzy Llinas Goodman

1 5 23

Reposted by The Distributed AI Research (DAIR) Institute

Peter Tarras @petertarras.bsky.social · 6d

If the current 'AI' hype seems ideologically motivated to you, it's because there's a whole bundle of ideologies operating in the background. Timnit Gebru @timnitgebru.bsky.social and Émile Torres @xriskology.bsky.social call it the 'TESCREAL bundle'. You can read their fundamental paper here:

The TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence | First Monday

firstmonday.org

17 69

Reposted by The Distributed AI Research (DAIR) Institute

AGF 🎧 poemproducer @poemproducer.bsky.social · 7d

for "ai" 👂🏻 @dairinstitute.bsky.social

www.dair-institute.org

DAIR Institute

The Distributed AI Research Institute is a globally distributed organization of academics, activists, and engineers conducting community-rooted research.

www.dair-institute.org

2 1

Reposted by The Distributed AI Research (DAIR) Institute

dylan @dylnbkr.bsky.social · 8d

$Across the top of the image is a fractured black rectangle where the left half is labeled "Frame: F is happening!" and the right side is labeled "Negation: F is not true!" left side of the box: "AI*" is so useful, it's revolutionizing everything. *massive, generative commercial "AI" models. right side of the box: "AI" capabilities are overblown, it's mostly slop. Below the box is a large image of Kirby, a pink blob inhaling and the "Frame F" fragment. In a box next to Kirby is the text: Kirby. K is happening. That's why they're saying F. In the pink box is the text: "AI" is functioning to fuel mass surveillance, exploitation of workers & the environment, and the consolidation of wealth and power on a massive scale.$

2 40 95

Reposted by The Distributed AI Research (DAIR) Institute

TalesofanAlfa @talesofanalfa.bsky.social · 8d

History repeats itself. Nothing ever changes.

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 12d

"“AI isn’t magic; it’s a pyramid scheme of human labor,” said @adiod.bsky.social , a researcher at the Distributed AI Research Institute based in Bremen, Germany. “These raters are the middle rung: invisible, essential and expendable.”

www.theguardian.com/technology/2...

How thousands of ‘overworked, underpaid’ humans train Google’s AI to seem smart

Contracted AI raters describe grueling deadlines, poor pay and opacity around work to make chatbots intelligent

www.theguardian.com

3 11

Reposted by The Distributed AI Research (DAIR) Institute

Ruha Benjamin @ruha9.bsky.social · 8d

Yall please go watch @merriam-webster.com’s new AI ad😅

youtu.be/RvQnl3O0D3c?...

LLM

YouTube video by Merriam-Webster

youtu.be

4 76 160

Reposted by The Distributed AI Research (DAIR) Institute

UC Davis School of Law @ucdavislaw.bsky.social · 8d

Join the Center for Innovation, Law, and Society (CILS) for a talk with Dr. @timnitgebru.bsky.social about the history of the Artificial General Intelligence (AGI) movement and its link to the 20th-century eugenics movement.

This in-person event will also be livestreamed: bit.ly/42kxaQV

#AGI #AI

14 30

Reposted by The Distributed AI Research (DAIR) Institute

Jase @jasonbell.bsky.social · 10d

If you really knew how your AI was being designed, filtered and trained.... you wouldn't want to use it.

This has been going on for years with the large tech companies, not just in AI.

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 12d

"“AI isn’t magic; it’s a pyramid scheme of human labor,” said @adiod.bsky.social , a researcher at the Distributed AI Research Institute based in Bremen, Germany. “These raters are the middle rung: invisible, essential and expendable.”

www.theguardian.com/technology/2...

How thousands of ‘overworked, underpaid’ humans train Google’s AI to seem smart

Contracted AI raters describe grueling deadlines, poor pay and opacity around work to make chatbots intelligent

www.theguardian.com

2 4

Reposted by The Distributed AI Research (DAIR) Institute

Preston Rudy @pordy24.bsky.social · 11d

Human labor powers machine ‘intelligence’

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · Jul 23

Indeed, @adiod.bsky.social writes:

“The choice facing Africa is...continue as a data colony providing cheap labour for surveillance tools that serve foreign interests, or assert digital sovereignty by developing governance frameworks that prioritise African lives”

africainfact.com/ai-surveilla...

AI surveillance and data colonialism shape African conflicts - Africa in Fact

On 3 November 2021, Professor Meareg Amare, a respected chemistry academic at Bahir Dar University in Ethiopia’s Amhara region, was gunned down outside his home by men wearing special forces uniforms....

africainfact.com

1

Reposted by The Distributed AI Research (DAIR) Institute

Andrew Clews @crackedwindscreen.bsky.social · 10d

Want an example of that? The autonomous vehicle industry. Who have somehow conned people into believing that teleops is suitable (which is insane if you actually think about it) and they kick in every mile or so.

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 12d

"“AI isn’t magic; it’s a pyramid scheme of human labor,” said @adiod.bsky.social , a researcher at the Distributed AI Research Institute based in Bremen, Germany. “These raters are the middle rung: invisible, essential and expendable.”

www.theguardian.com/technology/2...

How thousands of ‘overworked, underpaid’ humans train Google’s AI to seem smart

Contracted AI raters describe grueling deadlines, poor pay and opacity around work to make chatbots intelligent

www.theguardian.com

1 3

The Distributed AI Research (DAIR) Institute @dairinstitute.bsky.social · 9d

Our own Kathleen Siminyu writes: "most parallel data [...] was from the religious domain, either translations of the Bible or other religious texts [...] done by religious organizations whose primary aim was evangelization, both in pre-colonial & post-colonial times."
akademie.dw.com/en/mind-the-...

Mind the gap: Building inclusive AI for African languages

Poor quality and domain bias in training data hinders language tool development. That's why language diversity must be a priority, says Kathleen Siminyu from the Distributed Artificial Intelligence Re...

akademie.dw.com

5 7