Lightnews — Scholar-powered news

Reposted by klaudia jaźwińska

Karl Bode @karlbode.com · 7h

twice a year the entire U.S. press becomes a marketing extension of a single billionaire-owned retailer, and nobody in any position of editorial power thinks it's weird or gross

photo of Google News search results for Amazon Prime Day (the second this year)

190 2.5K 7.9K

klaudia jaźwińska @klaudia.bsky.social · 6d

Today for @columjournreview.bsky.social I wrote about how Reddit's role as a vital supplier of data for AI companies has made it a consequential player in the relationship between platforms and publishers www.cjr.org/analysis/red...

(cc: @towcenter.bsky.social)

Reddit is winning the AI game.

Since the site struck licensing deals with Google and OpenAI, its traffic—and its importance to legacy media outlets—has surged.

www.cjr.org

1 5

Reposted by klaudia jaźwińska

Columbia Journalism Review @columjournreview.bsky.social · 6d

Since Reddit struck licensing deals with Google and OpenAI, its traffic—and its importance to legacy media outlets—has surged, by @klaudia.bsky.social of the @towcenter.bsky.social. www.cjr.org/analysis/red...

Reddit is winning the AI game.

Since the site struck licensing deals with Google and OpenAI, its traffic—and its importance to legacy media outlets—has surged.

www.cjr.org

6 9

klaudia jaźwińska @klaudia.bsky.social · 12d

damn there's a fake, likely AI-generated quote attributed to me in an article about AI. I've never used the phrase "Faustian bargain" in my life

1 3

Reposted by klaudia jaźwińska

Josh Sternberg @joshsternberg.com · 18d

It’s not lost on me that the only profession mentioned in the Constitution (the press) is a constant target by those who profess their love for the founding document, who loudly say that we can only have a society based on rules set 250 yrs ago.

Carl Quintanilla @carlquintanilla.bsky.social · 18d

“.. Those who refuse to sign the new policy will be stripped of or denied permanent credentials to cover the Pentagon.”

@washingtonpost.com
www.washingtonpost.com/business/202...

8 120 400

Reposted by klaudia jaźwińska

klaudia jaźwińska @klaudia.bsky.social · 20d

In our latest piece for CJR, @aisvarya17.bsky.social and I argue that third-party evaluation of AI models is not just a technical matter; it’s a matter of accountability. www.cjr.org/analysis/jou...

Journalists need their own benchmark tests for AI tools.

The performance tests used by AI companies don’t measure what matters in the newsroom.

www.cjr.org

4 5

Reposted by klaudia jaźwińska

Laura Anne Gilman @lauraannegilman.bsky.social · 20d

www.nytimes.com/1939/02/04/a...

Headline from the New York Times, 1939: “Goebbels ends careers of five “Aryan“ actors who made witticisms about the Nazi regime.”

55 2.6K 5.6K

Reposted by klaudia jaźwińska

Emily Bell @emilybell.bsky.social · 20d

In a shock development - AI companies are failing to benchmark what matters in newsrooms. More excellent work from our Tow Center research team @klaudia.bsky.social and @aisvarya17.bsky.social

Columbia Journalism Review @columjournreview.bsky.social · 20d

The performance tests used by AI companies don’t measure what matters in the newsroom. By @klaudia.bsky.social and @aisvarya17.bsky.social for @towcenter.bsky.social. www.cjr.org/analysis/jou...

1 5 12

klaudia jaźwińska @klaudia.bsky.social · 20d

In our latest piece for CJR, @aisvarya17.bsky.social and I argue that third-party evaluation of AI models is not just a technical matter; it’s a matter of accountability. www.cjr.org/analysis/jou...

Journalists need their own benchmark tests for AI tools.

The performance tests used by AI companies don’t measure what matters in the newsroom.

www.cjr.org

4 5

Reposted by klaudia jaźwińska

Columbia Journalism Review @columjournreview.bsky.social · 20d

The performance tests used by AI companies don’t measure what matters in the newsroom. By @klaudia.bsky.social and @aisvarya17.bsky.social for @towcenter.bsky.social. www.cjr.org/analysis/jou...

1 5 14

Reposted by klaudia jaźwińska

Natalie Wynn @contrapoints.bsky.social · 23d

stop trying to make civil war happen

40 75 1.1K

Reposted by klaudia jaźwińska

Tuttle Tuttle Toil & Trouble 🎃 @katekilla.bsky.social · 25d

I shouldn't be shocked, but I'm still always brought up short to realize how many people think "civility" means "white people speaking in relatively calm voices" no matter what vile shit they say.

270 4.6K 19K

Reposted by klaudia jaźwińska

Anna Merlan @annamerlan.bsky.social · 26d

Each alleged mass shooter is even more On the Computer All the Time than the previous one. Probably nothing to worry about there

170 2.2K 13K

Reposted by klaudia jaźwińska

klaudia jaźwińska @klaudia.bsky.social · Aug 26

NEW in @columjournreview.bsky.social: Why AI models are bad at fact-checking photos www.cjr.org/tow_center/w...

1 12 20

klaudia jaźwińska @klaudia.bsky.social · Aug 29

You can view them at the github repo linked in the article github.com/TowCenter/fa...

GitHub - TowCenter/factchecking-ai

Contribute to TowCenter/factchecking-ai development by creating an account on GitHub.

github.com

1 1 3

Reposted by klaudia jaźwińska

Tow Center for Digital Journalism @towcenter.bsky.social · Aug 28

We asked seven different AI models to identify the location, date and source of photos taken by photojournalists, they collectively only answered 14 out of 280 queries completely correctly.

@aisvarya17.bsky.social @klaudia.bsky.social @columjournreview.bsky.social

www.cjr.org/tow_center/w...

Why AI Models Are Bad at Verifying Photos

“You don't know when it's just making stuff up.”

www.cjr.org

1 8 6

klaudia jaźwińska @klaudia.bsky.social · Aug 28

DeepSeek doesn't have the option of searching with images. We have tested DeepSeek in our other research www.cjr.org/tow_center/w...

AI Search Has A Citation Problem

We Compared Eight AI Search Engines. They’re All Bad at Citing News.

www.cjr.org

1 1 4

Reposted by klaudia jaźwińska

Aisvarya Chandrasekar @aisvarya17.bsky.social · Aug 28

New Research: Can AI models be used to fact-check images?

www.cjr.org/tow_center/w...

I cannot summarize this piece better than @klaudia.bsky.social did in this thread. The methodology and data from the experiment are linked in the article. Reach out with any questions, comments, or feedback!

1 3

Reposted by klaudia jaźwińska

Foeke Postma @foeke.bsky.social · Aug 28

More testing of LLMs for journalism, this time on identifying image location, date, and source.
By @aisvarya17.bsky.social and @klaudia.bsky.social with some of my own commentary.
www.cjr.org/tow_center/w... @columjournreview.bsky.social

3 21 51

Reposted by klaudia jaźwińska

Sarah Scott Scire @sarahscire.com · Aug 27

"@towcenter.bsky.social systematically tested the ability of seven AI chatbots—ChatGPT, Perplexity, Grok, Gemini, Claude, and Copilot—on a basic fact-checking task ... We found that all seven models failed to consistently correctly identify the provenance of the photos."

klaudia jaźwińska @klaudia.bsky.social · Aug 26

NEW in @columjournreview.bsky.social: Why AI models are bad at fact-checking photos www.cjr.org/tow_center/w...

1 2

Reposted by klaudia jaźwińska

klaudia jaźwińska @klaudia.bsky.social · Aug 26

As AI-generated content floods our social media feeds, it's tempting to turn to LLMs to try to figure out whether photos we see are real, or to learn more about when and where they were taken. Unfortunately, these tools can't be relied upon for visual fact-checking tasks. www.cjr.org/tow_center/w...

Why AI models are bad at verifying photos

Many of the latest large language models, including OpenAI’s recently released GPT-5, tout an impressive capability: reasoning about images. We have seen viral trends of AI models geolocating obscure ...

www.cjr.org

1 1 1

Reposted by klaudia jaźwińska

Dhrumil Mehta @dmil.bsky.social · Aug 26

Great work by @aisvarya17.bsky.social and @klaudia.bsky.social. Read more below!

www.cjr.org/tow_center/w...

Why AI models are bad at verifying photos

Many of the latest large language models, including OpenAI’s recently released GPT-5, tout an impressive capability: reasoning about images. We have seen viral trends of AI models geolocating obscure ...

www.cjr.org

2 1

Reposted by klaudia jaźwińska

The Ass is the father of the legs @internethippo.bsky.social · Aug 26

The mental health benefits of not knowing any history right now must be massive. Moving through your day with total equanimity. What happens next? Who knows? Maybe something good

100 2.3K 12K

Reposted by klaudia jaźwińska

Kashana @kashana.blacksky.app · Aug 26

Well, AI drinks our water, raises our electric bills, and tells kids to off themselves, but at least it’s making us dumber

10 290 1.1K

klaudia jaźwińska @klaudia.bsky.social · Aug 26

Why AI models are bad at verifying photos

Many of the latest large language models, including OpenAI’s recently released GPT-5, tout an impressive capability: reasoning about images. We have seen viral trends of AI models geolocating obscure ...

www.cjr.org

1 1